Contribution of Intestinal Cytochrome P450-Mediated Metabolism to Drug-Drug Inhibition and Induction Interactions

Drug Metab. Pharmacokinet. 25 (1): 28–47 (2010).

Review Contribution of Intestinal Cytochrome P450-Mediated Metabolism to Drug-Drug Inhibition and Induction Interactions Aleksandra GALETIN*, Michael GERTZ and J. Brian HOUSTON School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Manchester, UK Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk

Summary: Current drug-drug interaction (DDI) prediction models incorporate intestinal interaction as the ratio of the intestinal availability in the presence and absence of an inhibitor/inducer (FG? and FG, respectively). The incorporation of the gut is commonly associated with a reduced number of false negative predictions; however, in some instances a trend for over-prediction is apparent. This differential success is highly dependent on the initial model assumptions and parameter estimates used (often inconsistent between the datasets) and cannot be associated exclusively with the incorporation of the intestine. The current review provides an assessment of the contribution of intestinal inhibition and induction in conjunction with different perpetrator and victim drug-related properties, focusing in particular on victim drugs with high intestinal first-pass extraction (À75%). Recommendations are given in order to avoid significant over-estimation of true positives and increased number of false positive predictions. This review discusses advantages and limitations of different in vitro and in vivo methods for assessing intestinal availability and associated inter-individual variability, due to the sensitivity of the DDI prediction models to the FG. Keywords: CYP3A4; drug-drug interactions; intestinal first-pass; inhibition; induction

to 9-fold for CYP2C19, CYP2D6 and CYP2C9).1) A number of different UDP-glucuronosyltransferase (UGT) conjugation enzymes are also expressed to some extent in the human small intestine, including UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9 and UGT1A10 as well as UGT2B4 and UGT2B6;9,10,17) UGT1A8 and UGT1A10 expression is exclusive to the small intestine.18) Extensive conjugation by intestinal UGTs has been suggested to contribute to the low bioavailability of raloxifene19,20) and troglitazone.21,22) In addition to UGTs, intestine also expresses a number of sulfotransferases (SULTs); in contrast to CYPs and UGTs, these conjugation enzymes are not located in the endoplasmic reticulum but in the cytosol.13,23,24) Recent analysis of the expression levels of the major human SULTs in a range of human tissues has shown that the small intestine contains the largest overall amount of SULT. Of the 5 principal human SULTs investigated, SULT1B1 and SULT1A3 were predominantly expressed

Metabolic enzymes and transporters in the human small intestine Human small intestine expresses a range of oxidation and conjugation metabolic enzymes.1–13) Cytochrome P450 (CYP) enzymes have been characterised in most detail, with CYP3A accounting for approximately 80% of total P450s in the gut (59–96%), followed by CYP2C9 (¿14%). A number of other P450 enzymes (CYP2C19, CYP2J2 and CYP2D6) are expressed, but expression levels are º2% of total P450s (Fig. 1); these estimates are based on the analysis of CYP abundance data from 31 individuals.1) The total amount of CYP3A expressed in the human small intestine (65.7–70.5 nmol) represents approximately 1% of the estimated hepatic levels.14–16) Although intestinal CYP3A4 content was reported to be highly variable (17-fold), variability in total CYP3A (18–151 pmol/mg protein) was on average comparable to variability observed for other CYP enzymes in the gut (7-

Received; September 21, 2009, Accepted; December 10, 2009 *To whom correspondence should be addressed: Dr Aleksandra GALETIN, School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Stopford Building, Oxford Road, Manchester, M13 9PT, UK. Tel. ＋44 161 275 6886, Fax. ＋44 161 275 8349, E-mail: Aleksandra.Galetin＠manchester.ac.uk The work was funded by a consortium of pharmaceutical companies (GlaxoSmithKline, Lilly, Novartis, Pfizer and Servier) within the Centre for Applied Pharmacokinetic Research at the University of Manchester. Financial support for AG's academic position in the School of Pharmacy and Pharmaceutical Sciences, University of Manchester and MG's PhD studentship was kindly provided by Pfizer, Sandwich, UK.

28

Role of Intestine in the Drug-Drug Interaction Prediction

29

Fig. 1. Cytochrome P450 abundance in human liver (A) and small intestine (B)122) The estimates of the CYP abundance in the liver are based on meta-analysis of 42 to 241 liver samples.67,97) In contrast, the intestinal data are based on 31 individuals.1)

(36 and 31% of total SULTs, respectively), whereas SULTs 1A1, 1E1 and 2A1 represent minor forms (6–19% of total).25) The expression of metabolic enzymes along the small intestine is not uniform, with the highest levels of P450 and UGT protein found in the proximal region of the intestine, declining distally.5,7,14,26–28) Consistent with these findings is the declining activity of CYP3A from the duodenum to the ileum, with the maximum activity shown in the upper jejunum for erythromycin N-demethylation and 1?-midazolam hydroxylation.7,14) This heterogeneous expression of metabolic enzymes is also observed within the small intestinal villi, where the highest levels are found in mature enterocytes lining the villus tips.29,30) The distribution of UGTs along the small intestine is regarded as being analogous to the regional distribution of CYPs.5,28) A recent study by Cao et al. indicated a 3-fold greater UGT gene expression relative to CYP3A4 in the human duodenum, but it is questionable whether this estimate will reflect the UGT:P450 protein abundance ratio along the whole length of the gut.31) The most common method of assessing the relative importance of intestinal metabolism is based on the comparison of hepatic and intestinal activity (expressed per mg of microsomal protein); differential catalytic activity in comparison to the liver has been reported across studies.14,15,32–35) Comparison of intestinal and hepatic catalytic activities normalised for the mean population relative abundance of P450 enzymes in specific organs (i.e., expressed per pmol P450) results in a very good agreement in the metabolic activities between liver and intestine and across the range of different P450s.32) However, normalisation for the enzyme abundance is challenging in the case of UGTs, as the absolute UGT expression levels are currently not available in the liver and in the small intestine. In order to allow valid comparison between

hepatic and intestinal glucuronidation, Cubitt et al. have expressed glucuronidation clearance data per gram of tissue. Scaling of intestinal and hepatic glucuronidation CLint was performed using the intestinal microsomal recovery of 20.6 mg/g intestine.21) This weighted scaling factor was estimated from mucosal microsomal protein and mucosal CYP3A4 in each of the intestinal sections from the original data reported by Paine et al.14) Hepatic data were scaled using standard hepatic microsomal recovery of 40 mg protein/g liver.36,37) The age of the donors has been shown to affect the estimate of hepatic microsomal recovery with a maximum value (40 mg protein/g liver) obtained in approximately 28-year-old donors, followed by a gradual decrease in older age.38) In contrast, the impact of age on intestinal recovery is currently unknown. In addition, it should be noted that intestinal microsomal recovery was estimated from samples obtained using the mucosal scraping method for the preparation of intestinal microsomes. The application of this method has been reported to give lower metabolic activity in intestinal microsomes than the enterocyte elution method;32,39) however, it is as yet unknown whether the choice of enterocyte preparation method will affect the microsomal protein yield. In addition to human in vitro systems, a number of animal models have been used to assess intestinal metabolic activity.5,31,33,40) Komura and co-workers have noted differential intestinal metabolic activities for midazolam and nisoldipine with a rank order of monkey§humanÀ rat.33,40) Despite different CYP3A expression levels in the intestine between monkey and human and differential parameter estimates for established CYP3A probe midazolam, the authors have proposed the use of monkey as an appropriate animal model for evaluating small intestinal first-pass metabolism of CYP3A substrates; although this has been questioned.41) Recent stu-

30

Aleksandra GALETIN, et al.

dies have also used transgenic Cyp3a-/- mice expressing human CYP3A4 in either intestine or liver in order to determine the relative importance of intestinal versus hepatic CYP3A.42) Studies with tissue-specific CYP3A4 transgenic mice revealed that intestinal CYP3A4 has a major impact on oral triazolam exposure, with minimal contribution of the hepatic CYP3A4.43) The same authors investigated further the utility of knockout and CYP3A4 transgenic mice for drug-drug interaction assessment in the liver and intestine using triazolam interactions with ketoconazole and gefitinib as examples. Based on these preliminary studies, the authors have proposed the use of humanised mouse models as useful animal models in early drug development. However, the extrapolation to human has to be cautious (as with all animal models mentioned above), as illustrated by the study in Cyp3a knockout mice using common CYP3A4 probe midazolam.44) In contrast to expected, the study has shown marked midazolam metabolism in Cyp3a knockout mice due to up-regulation of CYP2C enzymes. The presence of efflux transporters, in particular Pglycoprotein (P-gp), on the apical membrane of enterocytes can modulate intestinal CYP3A first-pass metabolism by decreasing the intracellular concentration of drugs and metabolites via active efflux. This proposed interplay between the two proteins allows recirculation of the drug and in conjunction with the variable abundance of CYP3A4 and P-gp contributes to the inter-individual variability in drug absorption and bioavailability.45–48) A number of studies have suggested that the expression of P-gp increases progressively from the stomach,27,49,50) whereas other studies report that the efflux transporters have a similar distribution pattern to CYP3A, with higher expression levels in jejunum in comparison to ileum and colon.51,52) In addition to P-gp, MRP1 and MRP2 transporters are located in the intestine at the basolateral and apical membrane, respectively,53–55) and are primarily involved in the efflux of conjugative metabolites. In contrast to efflux transporters, uptake transporters and their potential interplay with metabolic enzymes and efflux transporters have received less attention. Recent studies have reported mRNA levels of a number of organic anion transporting polypeptide (OATP) uptake transporters on the apical membrane of enterocytes, including OATP1A2, OATP2B1, OATP1B3 and OATP1B1;56) the latter two were previously thought to be liver-specific.57–59) Considering that only mRNA levels of OATP1B1 and OATP1B3 were detectable in intestine, their role in intestinal uptake is most likely negligible.

Clinical evidence of intestinal metabolism CYP3A4 is susceptible to a number of inhibition and induction metabolic drug-drug interactions (DDI)60–66) due to its high abundance in both liver and intestine.1,67,68)

In addition, the relatively low blood flow to the intestinal mucosa means that the outflow of the drugs from enterocytes is not as high as that from hepatocytes and hence lower abundance of CYP3A can act several fold more efficiently than in the liver. A number of clinical studies have indicated a more pronounced inhibition and induction of intestinal CYP3A in comparison to the liver.69–75) For mutual P-gp and CYP3A4 substrates (e.g., cyclosporine, tacrolimus) the interaction is complex and a result of interactive nature and differing inter-relationship of these two proteins in the intestine and liver.76) However, certain studies suggest that for high permeability substrates (e.g., verapamil) the role of P-gp in the in vivo pharmacokinetics after oral administration and its contribution to the overall inhibition effect may have been overestimated.77,78) Numerous drug-drug interaction studies where victim drugs were administered both i.v. and orally in the absence and presence of an inhibitor/inducer provide indirect clinical evidence for the contribution of the intestine to the magnitude of DDI, although this interpretation has been challenged.5) In all the cases shown in Table 1 and discussed throughout this review, the term intestine refers to intestinal CYP3A, as equivalent information on any UGT-mediated DDIs is very limited.79–81) Table 1 shows up to 2.8-fold greater increase in the AUC ratio observed after oral dose of the victim drug in comparison to the i.v. administration, in the presence of either reversible or time-dependent inhibitors, suggesting a contribution of first-pass metabolism in the small intestine. In studies where grapefruit juice (GFJ) was used as a perpetrator, selective inhibition of intestinal CYP3A4 metabolism was observed,82–85) allowing the delineation of the importance of the small intestine;86,87) this approach is discussed in more detail below. Differential change in the AUC and potential contribution of the small intestine was also shown in the case of induction studies,73,74,88) resulting in up to 8.5-fold greater changes in the AUC after oral administration of nifedipine in the presence of rifampicin (Table 1).

Prediction of metabolic drug-drug interactions In recent years numerous efforts have been made to predict the magnitude of in vivo DDI from in vitro data, with varying degrees of success.60–62,64–66,89–92) The accurate determination of perpetrator concentration in vivo is problematic because a direct measurement is not possible and there is no generally accepted consensus for the extrapolation of inhibitor/inducer concentration in plasma to that at the enzyme site.60–62,89,93–95) DDI prediction models have been refined over the years to incorporate the contribution of multiple inhibitors (or metabolites) and/or the consequence of multiple inhibition mechanisms.63,96–100) Additionally, enzyme properties,63,101,102) characteristics of the victim drugs, namely incorporation

31

Role of Intestine in the Drug-Drug Interaction Prediction

Table 1. Increase in AUC of a range of victim drugs administered i.v. and orally in the presence and absence of CYP3A4 inhibitors and inducers Inhibitor

Fold increase in AUC (i.v.)

Fold increase in AUC (oral)

Alfentanil

Grapefruit juice88)

1.1

1.6

1.5

Alfentanil

Troleandomycin88)

7.3

19

2.6

Cyclosporine

Grapefruit juice82)

1.1

1.6

1.5

Cyclosporine

Ketoconazole174)

1.9

5.3

2.8

Felodipine

Grapefruit juice85)

0.9

1.7

1.9

Midazolam

Clarithromycin174)

2.7

7.0

2.6

Midazolam

Grapefruit juice84)

1.0

1.5

1.5

Midazolam

Grapefruit juice88)

1.3

1.7

1.4

Midazolam

Itraconazole175)

3.2

6.6

2.1

Midazolam

Fluconazole175)

2.0

3.6

1.8

Midazolam

Erythromycin176)

2.2

4.4

2.0

Midazolam

Saquinavir70)

2.5

5.2

2.1

Midazolam

Voriconazole71)

3.5

9.4

2.7

Midazolam

Ketoconazole72)

5.0

14

2.7

Midazolam

Diltiazem100)

1.7

4.0

2.3

Nifedipine

Grapefruit juice177)

1.1

1.6

1.4

Saquinavir

Grapefruit juice178)

0.9

1.5

1.7

Tacrolimus

Ketoconazole135)

1.3

2.9

2.3

Fold decrease in AUC (i.v.)

Fold decrease in AUC (oral)

Fold difference (oral/i.v.)

Victim drug

Inducer

Fold difference (oral/i.v.)

Alfentanil

Rifampicin88)

2.7

22

8.2

Cyclosporine

Rifampicin139)

1.4

3.7

2.6

Midazolam

Rifampicin75)

2.2

9.7

4.4

Tacrolimus

Rifampicin138)

1.5

3.1

2.0

Nifedipine

Rifampicin73)

1.4

12

8.5

of parallel elimination pathways,63,65,103,104) in vitro CYP3A4 kinetic complexities,62,105) contribution of the intestinal inhibition and recently also induction have been considered.61,63–65,91,100,106–108) There is an interest in the potential impact of metabolic intestinal interactions and this is increasingly becoming a part of the prediction strategy. This review will discuss the ability and limitations of the current approaches to estimate the extent of intestinal inhibition and induction DDI, focusing in particular on the most abundant intestinal P450, CYP3A4. Considering the multifactorial nature of the prediction models, the assessment will consider different perpetrator and victim drug-related properties, focusing in particular on the extent of intestinal first-pass extraction of victim drugs. Drug-drug interaction prediction models: The most common in vivo metric used to assess DDIs is the change in area under the plasma concentration time curve (AUC) of the victim drug following multiple dosing of a second interacting drug relative to the control state.60,109–114) Under the assumption that the change in P450 mediated clearance is not associated with any effect on the intestinal absorption or plasma protein binding of

the victim drug, comparison of AUC under control conditions and in the presence of the inhibitor/inducer depends on both hepatic and intestinal events, as illustrated in Eq. (1): AUC? FG? 1 ＝ ･n (1) n fmCYPi AUC FG Si CLint/CLint?＋ 1－Si fmCYPi where CLint? and CLint represent intestinal intrinsic clearance in the presence and absence of the perpetrator drug, fmCYPi denotes the fraction of the victim drug clearance attributed to the particular P450 subject to the inhibition/induction effect, and the term i indicates the existence of multiple enzymes (n).104,115) The (1－SfmCYPi) term in the denominator accounts for other clearance pathways of the victim drug (either other metabolic enzymes, biliary or renal excretion), assuming that the other pathways are not subject to inhibition.62,103,104,116) The first term in the equation comprises of the FG?/FG ratio (intestinal availability in the presence and absence of the perpetrator drug, respectively), which allows incorporation of the changes in intestinal metabolism; the remaining part of the equation relates to changes in

Ø

»

32

Aleksandra GALETIN, et al.

Table 2. Mechanistic equations describing changes in hepatic and intestinal CLint

Mechanistic equation for CLint ratio Mechanism Hepatic

CLint CLint?

Intestinal

m

Reversible inhibitiona

m

1＋S [I]j /Kij

1＋S [IG]j /Kij

j

m

Irreversible inhibitionb

1＋S j

Induction

c

CLint G CLint G?

j

kinact, j [I] kdeg, H (KI, j＋[I])

m

1＋S j

1 Emax j [I] 1＋ S j [I]＋ EC50j

kinact, j [IG] kdeg, G (KI, j＋[IG])

1 Emax j [IG] 1＋ S j [IG]＋ EC50j

m

m

CLint G? and CLint G represent intestinal intrinsic clearance in the presence and absence of the perpetrator drug. a [I] is the inhibitor concentration available to the enzyme and Ki is the inhibition constant.60–62) [IG] refers to the concentration in the intestine during absorption phase.122) b kinact is the maximum inactivation rate constant, KI is the inhibitor concentration at 50% of kinact and kdeg, H and kdeg, G represent the endogenous degradation rate constants of the enzyme in the liver and intestine, respectively.63,107,114,124,179) c Emax is the maximum induction effect and EC50 the concentration at 50% of Emax.66,91,117). In each case the possibility of m inhibitory/induction species is accounted for by the subscript j.

hepatic metabolism. Interaction with intestinal CYP3A enzymes is accommodated in an analogous way to the liver and the FG?/FG ratio can be expanded as shown in Eq. (2). This approach is applicable for both inhibition (reversible and irreversible) and induction interactions,63,65,91,97,107,114,117) as shown in Table 2. FG? 1 ＝ (2) CLint G? FG FG＋(1－FG) CLint G Reversible, time-dependent inhibition and induction mechanisms can all be defined by a specific mechanistic equation for the CLint ratio which can be substituted within the generic equations (Eqs. (1) and (2)) for the hepatic and intestinal components (Table 2). In these equations several surrogate concentration terms are used for the inhibitor/inducer concentration at the enzyme/effector site ([I])—the average systemic plasma concentration after repeated oral administration or the maximum hepatic input concentration60,93,118) for the liver and the estimated inhibitor concentration in the intestine during the absorption phase.65,95,106) In contrast to the most commonly used time-averaged value,62,89,103) certain simulation programmes (e.g., Simcyp}) can incorporate the time course of the inhibitor concentration.61,68,119,120) Prediction principles in the static and time course (dynamic) models are the same, but the time course can be important in the assessment of reversible inhibition DDIs and in particular for inhibitors with active metabolites contributing to the inhibitory effect (e.g., itraconazole).98) The time course should be less critical for time-dependent inhibition or induction where the altered state of ac-

tivity is relatively stable during the dosing interval of study in line with the recovery half-life of the affected enzyme. However, it can provide useful information on the potential variation in the magnitude of induction, depending on the changes in the inducer concentration relative to EC50 associated with the time course. In contrast, the static model assumes a constant extent of induction. The generic equations can be adapted to simultaneously accommodate different inhibition mechanisms (reversible and time-dependent) or even inhibition and induction mechanisms for one DDI (Eq. (3)),91,96,97) where the term k indicates the existence of multiple mechanisms p. AUC? 1 ＝ (3) fmCYP AUC ＋(1－fmCYP) p CL CL ? / k int int P k

Limitations and accuracy of different methods to estimate FG?/FG ratio: The FG?/FG ratio can be estimated in three different ways,106) as outlined below; in each case the FG control values can be obtained from either in vivo or in vitro data (discussed in detail in section ``In vivo and in vitro methods for estimation of FG''). 1) In vivo FG?/FG ratio—obtained from i.v. and oral data in the presence of an inhibitor or inducer (limited availability of such datasets). The assumption behind this approach is that the perpetrator drug does not affect the fraction absorbed (Fa) or the plasma protein binding of a victim drug and does not alter hepatic blood flow. 2) Maximum FG?/FG ratio—assuming the `worst case' scenario, i.e., maximum inhibition of intestinal CYP3A4

Role of Intestine in the Drug-Drug Interaction Prediction

resulting in FG?＝1, and therefore the maximum ratio as 1/FG. However, there is no equivalent scenario suitable for induction DDIs. 3) Model predicted FG?/FG ratio—obtained from the change in the intestinal intrinsic clearance in the presence of an inhibitor/inducer (CLint G?) using the in vitro obtained parameters (e.g., Ki in the case of reversible inhibition) and the estimated inhibitor concentration in the intestinal wall during the absorption phase (IG) (Eq. (4)):97,106,108,121,122) DkaFa IG＝ (4) Qent where Fa is the fraction absorbed, ka the absorption rate constant, D the dose, and Qent the enterocytic blood flow (average 0.3 L/min).65) Decrease in the intestinal intrinsic clearance of the victim drug in the presence of timedependent inhibitor is obtained from corresponding in vitro and enzyme parameters (kinact, KI and kdeg, G) and IG;64,107) in the case of induction, corresponding in vitro parameters (Emax, EC50) are used to estimate the extent of change in the intestinal clearance (Table 2). Intestinal CYP3A4 kdeg, G has been estimated from the reported changes in the AUC of a victim drug at different times after ingestion of grapefruit juice. The rationale behind this approach is selective and irreversible inhibition of intestinal CYP3A4 by components of grapefruit juice, with no effect on hepatic enzymes.87,123) The CYP3A4 recovery t1/2 estimated by this method ranges from 13 to 33 h across different studies (mean value of 23±10 h), resulting in the corresponding intestinal kdeg of 5×10－4 min－1.64,86,101) The estimated intestinal CYP3A4 t1/2deg is shorter in comparison to hepatic estimates,63,124) probably reflecting the turnover of intestinal mucosal cells.5,125) The accurate estimation of this enzyme property is very important for the assessment of the contribution of the intestinal time-dependent inhibition and induction interactions. Impact of inter-individual variability in hepatic CYP3A4 t1/2deg (20–146 h, kdeg, H of 0.8–5×10－4 min－1) on the assessment of time-dependent inhibition potential and DDI prediction accuracy was investigated previously.63) The study recommended the use of average hepatic CYP3A4 t1/2deg of 3 days (i.e., kdeg, H of 1.6×10－4 min－1) in the assessment of time-dependent interaction potential and this approach yielded a high prediction success rate (defined by 2-fold of observed in vivo values criteria). Estimation of intestinal concentration of the perpetrator drug. Analogous to the liver, a direct estimate of the concentration of the perpetrator drug at the enzyme site in the intestine is not possible. The current approach to estimate IG (Eq. (4)) accounts for some of the drug/formulation dependent (Fa, ka and D) and physiological properties (Qent). In a recent analysis of 36 reversible and time-dependent inhibitors,106) estimated inhibitor concentrations during the absorption phase span three

33

orders of magnitude, with the concentrations of digoxin and azithromycin among the lowest (º0.1 mM) and fluconazole and gemfibrozil among the highest (À100 mM). Although informative, this approach has certain limitations. Drug solubility, i.e., aqueous at pH 7.4 or in biorelevant media (e.g., fasted-state simulated intestinal fluid, FaSSIF),126,127) is not taken into consideration when interpreting IG estimates. One of the assumptions in this approach is that the inhibitor/inducer is not subject to extensive first-pass metabolism itself. This may not be correct for certain inhibitors for which extensive metabolism has been reported, for example diltiazem and itraconazole.98,100,128) A recent study in rats129) suggests that hydroxy-itraconazole formed by intestinal CYP3A controls the time course of hepatic CYP3A inhibition and is mainly responsible for the observed increase in hepatic availability of itraconazole. In addition, currently there is no agreement on the estimate of any potential binding within the enterocytes and hence this can lead to potential over-estimation of IG. Physiological variability in enterocytic blood flow (0.1–0.5 L/min, 2–10% cardiac output)130,131) results in a range of IG after a standard dose (e.g., 8–42 mM ketoconazole after 200 mg dose and 23–117 mM rifampicin after 600 mg dose),106) contributing potentially to the inter-individual differences in the magnitude of inhibition/induction effect in conjunction with other factors (dose of the perpetrator drug, intake of food, CYP3A4/3A5 intestinal abundance and activity). Current FDA regulations for the classification of potential P-gp inhibitors132) propose estimation of the intestinal lumen concentration by dividing the dose by a volume of 250 mL.132,133) The concentrations estimated by this approach are generally two orders of magnitude higher compared to estimates based on Fa, ka, D and Qent.

Impact of incorporation of intestinal interaction in the prediction strategy Intestinal inhibition: High concentrations of the putative inhibitor/inducer achieved during the absorption phase imply a high potential for DDIs at the level of the intestine. The potential for an interaction in the intestine is determined by the ratio of intestinal concentration of the inhibitor/inducer to its estimated potency, i.e., IG/Ki or IG/EC50 for reversible inhibition and induction interactions, respectively. In addition, significant intestinal firstpass metabolism of the victim drug may contribute to the inter-individual variability in the magnitude of DDI, as seen for a number of CYP3A4 substrates.72,74,85,134,135) The impact of both inhibitor- (IG, Ki) and substraterelated (FG) characteristics on the FG ratio was addressed recently over the IG/Ki range from 0.01–100 and for victim drugs with differential levels of intestinal extraction, from low to high.106) A relative IG/Ki ratio of À10 indicates a potential for interaction in the intestine. Similarly, in the case of time-dependent inhibitors, tight binders

34

Aleksandra GALETIN, et al.

Table 3. List of bioavailability and intestinal availability values for 25 CYP3A4 substrates. The FG values were estimated either from i.v. and oral data122) or from grapefruit juice interaction data.86) In case of multiple studies, numbers in brackets represent 95% confidence interval. F

FG (i.v./oral)122)

FGGFJ86)

Alfentanil

0.42

0.6

0.61

Alprazolam

0.84

0.94

0.89

Atorvastatin

0.14

0.24

0.56 (0.36, 0.76)

Buspirone

0.05

0.21

0.11a

Cisapride

0.4–0.5

—

0.55 (0.32, 0.83)

0.22–0.36

0.44 (0.35, 0.54)

0.65 (0.59, 0.70)

Felodipine

0.14

0.45

0.53 (0.49, 0.57)

Lovastatin

Ã0.05

—

0.07

Maraviroc

0.23

0.75–0.93

—

Methadone

0.92

0.78

0.85 (0.83, 0.86)

Midazolam

0.24–0.41

0.51 (0.47, 0.56)

0.56 (0.49, 0.65)

Nifedipine

0.41

0.78

0.62 (0.52, 0.73)

Nisoldipine

0.05–0.08

0.11

0.44 (0.12, 0.76)

Quinidine

0.78

0.9

0.92

Repaglinide

0.56

—

0.89

Rifabutin

0.2

0.21

—

Saquinavir

º0.04

0.18

0.54 (0.33, 0.81)

Sildenafil

0.38

0.54

0.82

Simvastatin

º0.05

—

0.14 (0.0, 0.28)

Tacrolimus

0.14

0.14

—

Terfenadine

º0.01

—

0.40

Trazodone

0.77

0.83

—

Triazolam

0.55

0.75

0.64 (0.60, 0.68)

Verapamil

0.22

0.59

0.71 (0.67, 0.75)

Zolpidem

0.72

0.79

—

Cyclosporine (healthy)

a

Not recommended as GFJ interaction study changed buspirone t1/2 suggesting inhibition of hepatic metabolism.

(KIº1 mM) or fast inactivating compounds (kinactÀ 0.01 min－1) have a potential for interaction in the intestine when IGÀ1 mM. The greatest sensitivity is observed for interactions involving victim drugs with moderate to high intestinal extraction (FGº0.5). In contrast, the effect is minimal (maximum 2-fold FG ratio) for drugs with an intestinal extraction Ã50% (e.g., midazolam, triazolam), irrespective of the potency of the inhibitor.106) Comparison of the relative ratios of the plasma and intestinal concentrations of the perpetrator drug to its potency could rationalise the contribution of each organ to the magnitude of the DDI. For example, low plasma concen-

trations (for example, if fuº0.01) can lead to low I/Ki ratios and negligible interaction potential at the level of the liver. At the same time, the analogous ratio for the intestine could be high, indicating a more pronounced effect on the level of intestine. The incorporation of the intestinal interaction into the DDI prediction strategy in the form of the FG?/FG ratio has resulted in variable prediction success.61,64,65,89,91,106,107) This differential success cannot be associated exclusively with the incorporation of the intestine, as other parameters were not consistent between the datasets, e.g., value of fmCYP3A4, use of different inhibitor concentrations in the prediction model, different enzyme degradation rate constants and number of victim drugs investigated. Incorporation of the maximum intestinal inhibition (as 1/FG in the generic prediction model) resulted in the successful prediction of time-dependent CYP3A4 DDIs with verapamil and midazolam.107) The same approach was used for the prediction of 28 time-dependent DDIs by Galetin et al.;63) this database was extended to 50 DDIs and is discussed here. Details of the clinical studies of this dataset are given in Burt et al. (submitted for publication). In contrast to studies that focus on a single victim drug (in most cases midazolam),107,108) this dataset included 14 victim drugs with differential extent of intestinal extraction, with in vivo FG values ranging from 0.07 to 0.94 for lovastatin and alprazolam, respectively. Different methods to estimate FG were recently reviewed86,122) and updated FG values were used in the analysis (values shown in Table 3). Contribution of the intestine was incorporated as the predicted FG ratio from in vitro data, IG and intestinal kdeg, G (as discussed in the section ``Limitations and Accuracy of Different Methods to Estimate the Extent of Intestinal Interaction''). Comparison of predicted/observed AUC ratios with or without the intestinal interaction incorporated and the prediction success associated are shown in Figure 2A and Table 4. Overall prediction success of the time-dependent DDIs was high regardless of inclusion of the intestinal interaction, with median predicted/observed AUC ratios ranging from 89–143%; the upper interquartile of the predicted/observed AUC ratios across scenarios did not exceed 250%. The inclusion of the intestine minimised the number of false negative predictions and resulted in 70% of studies predicted within the 2-fold range; approximately half of the studies were predicted within the 1.5fold of the in vivo value. No particular trend was apparent between the magnitude of DDI (weak, medium or potent interaction as defined by FDA,111)) and prediction accuracy related to the inclusion of the intestinal interactions. Incorporation of intestinal inhibition resulted in overpredictions of some of the interactions, in particular lovastatin interactions with diltiazem and atorvastatin interactions with clarithromycin or erythromycin, where

35

Role of Intestine in the Drug-Drug Interaction Prediction

Fig. 2. Comparison of predicted/observed AUC ratios for time-dependent inhibition (A) and induction DDIs (B) related to FG values of the victim drugs associated with the interactions  represents predictions without intestinal contribution and  with intestinal interaction incorporated as predicted FG?/FG. Outliers identified represent: 1, lovastatin; 2, simvastatin; 3, buspirone; 4, atorvastatin; 5, terfenadine; 6, cyclosporine; 7, felodipine; 8, midazolam; 9, tacrolimus; 10, saquinavir and 11, nifedipine. Dashed lines represent 2-fold from the line of unity.

Table 4. Comparison between predicted/observed AUC ratio and different prediction scenarios for 50 time-dependent and 36 induction drug-drug interactions Time-dependent drug-drug interactions (n＝50) Median (%)

% within 2-fold

afe

rmse

No intestinal interaction

89 (68–111)1

90

1.15

1.23

Predicted FG?/FG2

143 (102–213)

70

1.43

4.41

Table 5. Accuracy of AUC ratio predictions for 50 time-dependent and 35 induction DDIs with or without the intestinal component in the prediction model. Predictions were categorised into three different groups according to the FG of the victim drug Time-dependent drug-drug interactions FG of the victim drug Number of DDIs

Induction drug-drug interactions

º0.25

º0.5

À0.5

º0.25

º0.5

À0.5

10

19

31

7

8

27

No intestinal interaction

Induction drug-drug interactions (n＝35) No intestinal interaction

68 (36–130)1

51

1.56

21

Predicted FG?/FG2

281 (176–612)

34

3.10

343

3

－ － Interquartile range; 2kdeg, G of 5×10 4 min 1 used; 3Based on the fold change in the AUC in the presence of rifampicin in order to allow direct comparison to inhibition interactions

Within 1.5-fold (%)

40

68

74

43

38

30

Outside 2-fold (%)

22

11

10

29

25

56

afe

1.37

1.19

1.14

1.07

1.12

1.84

rmse

2.05

1.50

1.02

3.59

3.43

24

1

the magnitude of interactions was over-predicted by À4.5-fold (Fig. 2A). These findings were in agreement with over-predictions of reversible CYP3A4-mediated DDIs of atorvastatin and tacrolimus, reported previously.89) However, incorporation of the intestinal interaction improved the prediction success of DDIs with midazolam, triazolam and nifedipine as victim drugs. This might be explained by the greater confidence in FG estimates for the latter three drugs in comparison to atorvastatin and lovastatin (Table 3), as it will be addressed in section ``In vivo and in vitro methods for estimation of FG.'' Table 5 shows the DDI prediction accuracy for the scenarios investigated (＋/－ intestinal component) in relation to the FG of the victim drugs in the dataset. For vic-

With intestinal interaction (predicted FG?/FG) Within 1.5-fold (%)

30

16

68

0

0

19

Outside 2-fold (%)

70

68

6

100

100

56

afe

2.8

2.5

1.14

10.3

10.4

2.17

rmse

9.3

6.9

1.3

62

59

22

tim drugs with moderate to low intestinal extraction (FGÀ0.5), highly accurate DDI predictions were observed, as Æ90% studies were predicted within 2-fold regardless of whether intestinal component was included in the model or not (Table 5, Fig. 3). This subset included a representative range of DDIs, as the observed magnitude spanned from 1.2- to 5.1-fold increase in AUC for quinidine-erythromycin and triazolam-clarithromycin studies, respectively. The largest discrepancy in predic-

36

Aleksandra GALETIN, et al.

Fig. 3. Prediction accuracy of 50 time-dependent inhibition and 36 induction DDIs in the absence and presence of the intestinal contribution Drug-drug interactions are classified into groups FGº0.5 or FGÀ0.5 according to the extent of intestinal extraction of the victim drug involved. First bar in each case represents the prediction without the intestinal interaction incorporated in the Eq. (1), whereas second bar refers to the combined hepatic and intestinal model. When incorporated, intestinal interaction is obtained as the predicted FG ratio. The box-and-whisker plots illustrate the distribution in the prediction success; the upper and lower whisker represent highest and lowest value that is not an outlier (represented by ). The black line represents the median value.

tion success between the hepatic and combined hepatic and intestinal model was observed for victim drugs with high intestinal extraction (FGº0.25, e.g., buspirone). For this subset, incorporation of the intestine resulted in 70% of the studies outside 2-fold, in contrast to 22% when only hepatic interaction was considered. The use of 1/FG rather than the predicted FG ratio increased further the prediction inaccuracy and the extent of over-prediction (data not shown). Over the entire range of FG increased bias was observed when the intestine was incorporated in the model; this was particularly apparent for the FGº0.25 subset. Other groups have also incorporated intestinal inhibition in the form of the predicted FG ratio rather than assuming maximum inhibition of intestinal enzymes;61,65) however, the number of victim drugs investigated in those datasets was smaller. Most studies tend to include DDIs with midazolam, triazolam and alprazolam (as most commonly reported DDIs), whereas Obach and coworkers have also investigated nifedipine, buspirone and simvastatin.65) The same authors have reported an improvement in the prediction bias and accuracy of a range of reversible and also time-dependent CYP3A4 DDIs when intestinal interaction was not ignored; in both studies use of unbound hepatic input concentration was advocated.64,65) Consistent with findings in the time-dependent DDI dataset discussed above, buspirone interactions with erythromycin and diltiazem were over-predicted when intestinal component was included; the fold difference (4 to 9) depended on the choice of inhibitor concentration used in the model.

Intestinal induction: In contrast to inhibition, the number of studies investigating prediction of induction DDIs has been small. In developing prediction strategies, advantage has been taken of experiences with inhibition and modelling approaches that integrate both induction and inhibition are available.108,136,137) Recently, a comprehensive set of 103 DDI studies involving 6 inducers of CYP3A4 and 38 victim drugs (fmCYP3A4 ranging from 0.05 to 0.94 for ropivacaine and midazolam, respectively) was reported.66) Generic DDI prediction model with the induction mechanistic term (Table 2) was applied; however, intestinal interaction was not incorporated in the analysis. A subset of 35 rifampicin DDIs with victim drugs with known FG values was selected for the current analysis. The dataset included 18 victim drugs with a differential extent of intestinal first-pass (FG＝0.14–0.94 in the case of tacrolimus and alprazolam, respectively). In vitro data from human hepatocytes (single donor) based on testosterone activity as an induction endpoint were used for the predictions. Intestinal interaction was incorporated as the predicted FG ratio obtained from the in vitro induction parameters and estimated IG concentration of rifampicin (39 mM after 600 mg dose assuming average Qent of 0.3 L/min). Analogous to the time-dependent DDIs, prediction accuracy was related to the FG of the victim drugs in the actual studies investigated (Fig. 2B, Table 5). Consistent with the inhibition database, significant outliers included interactions with atorvastatin, buspirone, simvastatin, saquinavir and tacrolimus, where in some instances the predicted/observed AUC ratio exceeded 2000% (saquinavir, tacrolimus).

Role of Intestine in the Drug-Drug Interaction Prediction

In contrast to inhibition studies, significant inaccuracy was observed even when hepatic induction was considered in isolation, with 49% of the studies predicted outside of 2-fold of unity. Under-prediction trend was apparent, with a median predicted/observed AUC ratio of 68%, ranging from 6–272% (Fig. 3). Only 31% of the studies were predicted within 1.5-fold of in vivo, in contrast to 72% of time-dependent inhibition studies predicted within this range when no intestinal interaction was incorporated. Considering very poor prediction accuracy of hepatic induction, it is not surprising that the incorporation of the intestinal component resulted in 66% of the studies predicted outside 2-fold of the in vivo value and an increased numbers of over-predictions (Fig. 3). In the case of time-dependent DDIs, introduction of the intestinal interaction increased the bias by 33%; in contrast, a 99% increase was observed when intestine was incorporated in the induction DDI model. However, the interpretation of this over-prediction trend has to be cautious and should not be associated entirely with the incorporation of the intestine. For a limited number of studies, induction data were available after both i.v. and oral administration of the victim drug in the presence of rifampicin, which allowed the delineation of the induction effect between liver and intestine. These studies also allowed us to assess the accuracy of the predicted extent of intestinal induction from in vitro data in the form of FG?/FG ratio. In the case of nifedipine, alfentanil and midazolam, DDI model predicted À70% reduction in FG due to induction (73, 81 and 84%, respectively), which was consistent with the observed in vivo changes (79, 75 and 72%, respectively).73,75,88) Despite the accuracy of FG?/FG ratio prediction, more than half of midazolam studies were over-predicted with the combined model (predicted/observed AUC ratio ranged up to 621%); similar trend was seen in the case of nifedipine (predicted/observed AUC ratio À700%) (Fig. 2B). In the case of cyclosporine and tacrolimus, model predicted 85–90% decrease in FG, in contrast to clinical data where only 50–57% increase in the intestinal extraction was observed in the presence of rifampicin.74,138,139) Analysis of the in vivo data after i.v. and oral administration of tacrolimus (one of the most pronounced outliers, Fig. 2B) indicated that rifampicin administration had a minor effect on tacrolimus FH (decrease by 1%) consistent with low clearance of tacrolimus relative to hepatic blood flow. However, predicted change in the AUC assuming only hepatic interaction resulted in 5-fold overestimation of the actual magnitude of DDI reported after i.v. administration in the study by Hebert et al. (1.5-fold change in AUC).138) Combined over-estimation of both hepatic and intestinal interaction (as discussed above) leads to an overall 24-fold over-prediction of the observed change in AUC of orally administered tacrolimus. Unfortunately, i.v. data were not available for other in-

37

duction pairs to increase our confidence in the predicted changes in FG in the presence of rifampicin using the current model. Limitations and accuracy of different methods to estimate the extent of intestinal interaction: As illustrated above, differential methods can be employed in order to incorporate the contribution of intestinal inhibition, induction or combined effect in the DDI prediction strategy. Recently, a comprehensive analysis of various approaches to assess the FG ratio has been performed using a dataset of 36 reversible and time-dependent inhibitors of CYP3A4 in the assessment.106) For a limited number of studies, data were available after both i.v. and oral administration of the victim drug in the presence of inhibitor, which allowed the assessment of the validity of the maximum inhibition assumption. The analysis indicated a very good agreement between maximum and observed FG ratios in vivo for interactions involving potent inhibitors (regardless of the mechanism) and substrates predominantly metabolised with no transporter issues (e.g., midazolam). Midazolam studies were the most abundant, with data available in the presence of 9 reversible and irreversible inhibitors (Table 1). For each of the inhibitors, FG? approached 1 as expected (range from 0.83 to 1 in the interactions with erythromycin and itraconazole, respectively), confirming the application of the maximum FG approach. Analogous findings were observed for alfentanil, felodipine and nifedipine in the presence of grapefruit juice as a perpetrator drug.86) However, the assumption of complete intestinal inhibition (1/FG) over-predicted the FG ratio observed in vivo for dual P-gp-CYP3A4 substrates cyclosporine and tacrolimus up to 3.4-fold.106) This is not surprising considering that the estimated cyclosporine FG did not approach 1 in the presence of ketoconazole or grapefruit juice (0.69 and 0.66, respectively);69,86) a similar result was apparent in the case of tacrolimus, where the presence of ketoconazole increased its FG from 14 to only 30%.135) The lack of availability of i.v. data for other victim drugs in the presence of either ketoconazole or other inhibitors prevents a more extensive comparison. Therefore, the assumption of maximal intestinal inhibition may result in an over-estimation of the importance of intestinal inhibition in the case of moderate to weak inhibitors or in the interactions involving victim drugs with differential contribution of both metabolic enzymes and transporters to their disposition, as illustrated above. Analysis of both inhibition and induction DDIs highlighted that prediction accuracy was lowest for the DDIs involving victim drugs with FGº0.5 (Fig. 3), and in particular for the subset of highly extracted drugs with FGº0.25 (Table 5). Use of the refined predicted FG approach and appropriate intestinal CYP3A4 kdeg, G (5× 10－4 min－1) is recommended for DDIs with these victim drugs in order to avoid significant over-estimation of true

38

Aleksandra GALETIN, et al.

positives and increased number of false positive predictions. For drugs with intestinal extraction Ã50% (e.g., midazolam), the maximum fold change in FG is 2, even in the cases of potent inhibition, i.e., when the perpetrator drug changes CLint G of a victim drug by À90%. Although this may suggest that the contribution of intestinal interaction is likely to be relatively minor, for some DDIs this 2-fold change may represent a significant improvement in DDI prediction and result in elimination of a falsenegative result, or can result in over-prediction. Therefore, accurate estimate of the FG control is essential for the assessment of the interaction magnitude, in addition to the inhibitor/inducer and enzyme properties, in particular for victim drugs where FGº0.25. The following section will address different methods to estimate the intestinal extraction from both in vivo and in vitro data and highlight their potential limitations.

In vivo and in vitro methods for estimation of FG Direct measurement of FG is rarely performed in human for practical and ethical reasons, with the exception of studies in anhepatic patients during liver transplant operations or in patients whose portal blood circulation bypassed the liver.140,141) However, the severe state of illness in these patients makes the FG estimate an unlikely representation of the FG in healthy individuals. Different indirect methods have been proposed to estimate the FG, namely use of plasma concentration-time profiles after oral and i.v. administration and interaction data in the presence of grapefruit juice; these methods have recently been reviewed and the FG estimates obtained by both methods have been revisited.86,122) The estimation of FG from i.v. and oral data is based on the assumption of negligible metabolism in enterocytes after i.v. administration and that systemic clearance of a drug after i.v. dose (corrected for renal excretion) reflects only hepatic elimination.106,142) The extent of intestinal extraction of 25 drugs estimated by this method is shown in Table 3, ranging from 6 to 93% in the case

of alprazolam and lovastatin, respectively.122) In all the cases CYP3A4 contributes predominantly to elimination (fmCYP3A4À0.7), with the exception of trazodone and zolpidem.63,95,103) In vivo FG estimates obtained by this method are of limited availability and represent the measure of the net result of the transporter-enzyme interplay. The i.v./oral approach is based on several assumptions that may lead to inaccuracies in FG estimates. The assumption that the extent of intestinal metabolism after i.v. administration is negligible may not necessarily be valid as an average 8% extraction ratio after an i.v. dosed midazolam has been reported in anhepatic patients (up to 26% in one patient).140) Variation in drug-dependent parameters such as CLi.v, F, Fa and blood:plasma ratio contributes to differential FG estimates and these parameters are not generally reported in the clinical studies used to estimate the FG. For example, in the case of tacrolimus, the blood:plasma ratio has been reported to range from 12 to 67 (average value used in this analysis was 35).143,144) Although these parameters are generally available in the literature, they are associated with considerable inter-individual variability. In the absence of accurate estimates of Fa, the indirect method assumes complete absorption (i.e., Fa＝1) when estimating the hepatic extraction ratio (EH). This can result in an under-estimation of FG and subsequently over-estimation of the FG ratio in the DDI prediction, in particular if the latter is obtained assuming maximum intestinal inhibition. Additional errors may occur due to the use of average QH, despite its range from 17.1 to 25 mL/min/kg,145) which may result in biased estimates of EH and consequently FG for drugs with blood clearance (CLb) approaching hepatic blood flow. A number of drugs in the current analysis have CLb values À70% of the average QH, namely atorvastatin, buspirone, felodipine, repaglinide, saquinavir and verapamil. DDIs associated with some of these drugs represent significant outliers in the case of both induction and inhibition interactions (Figs. 2A and 2B). This can be rationalised by the sensitivity of i.v./oral FG esti-

Fig. 4. Impact of QH on the estimation of FG (grey) and FH (white) for three Ca2＋ -channel antagonists reported to increase QH A: FH and FG estimates based on the mean population QH (20.7 mL/min/kg); error bars indicate the physiological range of QH (17.1–25.5 mL/min/kg). B: FG and FH estimates based on the reported increase in QH by 71, 47 and 36% for felodipine, verapamil and nifedipine, respectively.85,146,147)

Role of Intestine in the Drug-Drug Interaction Prediction

mates on QH and consequently sensitivity of the DDI models on the FG parameter when estimated to be in the range º0.25, as is the case for buspirone and atorvastatin. In addition, certain drugs, for example felodipine, verapamil and nifedipine, were previously reported to temporarily increase QH by 36 to 71%85,146,147) and this is generally not taken into account when estimating FG from i.v./oral data. Figure 4 highlights the potential bias in FG and FH estimates for these drugs associated with alterations in QH. The FH and subsequent FG estimates of felodipine and verapamil are highly dependent on QH, whereas the effect of QH is minor for the estimation of nifedipine FH and FG. This can be explained by the generally higher CLb values of felodipine and verapamil (approximately 75% of average QH in comparison to 50% in the case of nifedipine). Assuming that the bioavailability remains unaffected, the increasing QH in the presence of the Ca2＋-channel antagonists decreases the FG estimate of felodipine, nifedipine and verapamil by 67%, 16% and 86%, respectively, as shown in Figure 4. Therefore, it must be emphasized, that especially for drugs with high hepatic extraction, an accurate assessment of QH is of pivotal importance for the estimate of FG, and as long as QH is not reported in studies aiming to determine intestinal availability, FG estimates for these drugs should be regarded with caution. Alternatively, FG can be estimated from the large number of interaction studies reported with grapefruit juice.87,123,148–150) This approach is based on the assumption of complete and irreversible inhibition of CYP3Amediated metabolism in the intestine but no effect on hepatic enzymes or the fraction absorbed of the investigated drug. The application of this method was addressed in detail recently,86) highlighting the limitation of this method for the estimation of FG for drugs whose disposition is co-dependent on efflux/uptake transporters and metabolic enzymes. The authors have reported the highest discrepancy between the i.v./oral and GFJ method in the case of atorvastatin, cyclosporine and saquinavir, all substrates for P-gp and OATP transporters.58,151,152) For example, analysis of i.v./oral data for atorvastatin estimated an FG value of 0.24,153) whereas grapefruit juice interaction data suggested a far less extensive intestinal contribution to atorvastatin first-pass metabolism (FG, GFJ＝ 0.56).86) In contrast, good agreement between the GFJ and i.v./oral FG estimates was observed for purely metabolised drugs, supporting the interchangeable use of these methods for drugs where transporter mediated uptake/efflux is insignificant. One key assumption in the estimation of FG from GFJ data is the complete and selective inhibition of CYP3A4 metabolism in the intestine. Assessment of this assumption was undertaken for drugs where i.v./oral data in the absence and presence of GFJ were reported in the same individuals.82–85,88,154) In the case of complete intestinal in-

39

hibition, the FG of the drug in the presence of GFJ should equal 1 and this was confirmed for drugs subject solely to metabolism (e.g., midazolam, alfentanil, felodipine and nifedipine). In contrast, GFJ was unable to completely inhibit the intestinal metabolism of cyclosporine and saquinavir, suggesting biased FG estimates from GFJ data. These findings support the discrepancy observed between the methods and higher FG estimates for cyclosporine and saquinavir observed from the GFJ data in comparison to the i.v./oral data (Table 3). In the current analysis FG values for simvastatin and lovastatin based on GFJ data were used in the prediction models due to lack of i.v. data. Both victim drugs fall into the category of high intestinal extraction; therefore, inaccuracy in FG estimates associated with limitations of the GFJ method will affect DDI prediction success with these victim drugs (Fig. 3). Inter-study and inter-individual variability in FG: The strongest clinical evidence of the extent of intestinal metabolism is available for midazolam, as data in anhepatic patients were reported together with a number of i.v./oral (n＝16) and GFJ (n＝6) studies.86,122) The mean (weighted for population size) FG estimate obtained for midazolam obtained from i.v./oral data of 315 individuals was 51±10%, ranging from 40 to 79%. Interindividual variability was also observed in the oral bioavailability (29±5%) and FH estimates (58±9%). The variable extent of midazolam FG values results in a maximum FG ratio (and hence increase in the hepatic prediction) of up to 2.5-fold. This is in good agreement with the observed 1.8- to 2.7-fold increase in AUC ratio after oral administration of midazolam in comparison to i.v. administration in the presence of 9 reversible and timedependent inhibitors (Table 1). While an extensive number of i.v./oral or GFJ studies were also reported for cyclosporine, felodipine, nifedipine, cisapride and triazolam, FG estimates of other drugs relied on data from a relatively small number of studies and individuals (n＝ 6–20).86,122) An additional limitation apparent for both i.v./oral and GFJ method is the fact that studies are mainly conducted in men. In a study with a large number of volunteers (58 males and 41 females), Kharasch et al. have shown that FG displayed gender-specific differences for midazolam and alfentanil.155) Therefore, FG values estimated in male populations for other drugs might not be applicable for mixed populations. Additionally, most FG estimates are generally obtained in healthy volunteers. In the case of cyclosporine, comparable availability of the GFJ interaction data in both healthy (6 studies, 74 individuals) and kidney transplant patient population (8 studies, 64 individuals) allowed the assessment of the impact of health status on FG estimates.86) Significantly lower FG values (pº0.01) were observed in the healthy group compared to stable kidney transplant patients, i.e., 0.64 and 0.77, respectively. The

40

Aleksandra GALETIN, et al.

availability of GFJ data allowed the assessment of both inter-individual and inter-study variation in FG. The interindividual variation in cyclosporine FG estimates (33% and 27% for healthy and kidney transplant patients, respectively) was more pronounced in comparison to the inter-study variability (9.2% and 5.7% for healthy and kidney transplant patients, respectively). A smilar trend was seen for felodipine, with inter-individual and interstudy variability of 24–54% and 13%, respectively.86) Reduction in intestinal extraction was observed in studies investigating cirrhotic patients, as reported for verapamil, nifedipine and midazolam.156–158) Increase in splanchnic blood flow and consequently decreased contact with intestinal CYP3A in the case of cirrhotic patients with portosystemic shunts has been suggested as a rationale for increased FG observed for midazolam in this group of patients.159) In contrast, Chalasani and coworkers proposed that a decrease in intestinal CYP3A4 activity in cirrhotic patients causes the increase in FG, as changes in splanchnic blood flow alone would be insufficient to explain the findings.157) Recently, McConn et al. confirmed that duodenal CYP3A expression and total midazolam hydroxylation were lower by 47 and 34%, respectively in patients with cirrhosis; the severity of the condition correlated with the decrease in intestinal expression levels and activity.160) Prediction of FG from in vitro data: In silico approaches to estimate FG using physiologically-based pharmacokinetic models with differential ability to incorporate in vitro metabolic and transporter data are becoming increasingly important in the current integrative and mechanistic systems biology approaches. Modelling of intestinal first-pass requires incorporation of drug absorption, zonal and cellular heterogeneous distribution of metabolic enzyme and efflux/uptake transporters along the length of the intestine and enterocytic rather than total organ blood flow. A number of physiologicallybased models have been reported with different levels of complexity and integration of passive permeability with the activity of metabolic enzymes and transporters.161–165) In contrast to complex physiologically-based models, `minimal' QGut model is also proposed to predict FG from in vitro clearance and permeability data by overcoming the inadequacy of the perfusion limited approaches adapted from the corresponding liver model.166,167) The QGut model and models adapted from the original Compartmental Absorption Transit (CAT)168) and Advanced CAT model165,169) have now been incorporated in the commercially available softwares (GastroPlus and Simcyp}). These integrated dynamic models for the prediction of oral drug absorption and metabolism have recently been discussed.165,170,171) The major disadvantage of some of the models proposed is their complexity and the limited availability of all the parameters required for the prediction of FG. Therefore, addressing all the com-

plexities of the intestinal first-pass metabolism adequately and yet retaining a relatively practical model remains a challenge. In addition, comprehensive analysis of the utility of the FG estimates obtained from in vitro clearance and permeability data for the prediction of inhibition and induction DDIs is lacking.

Summary and Conclusions The contribution of intestinal interaction is incorporated into the prediction equation based on hepatic enzyme interaction data in the form of the ratio of FG in the presence and absence of the perpetrator; this approach is applicable for both inhibition and induction interactions. The variable success of current prediction strategies might pose the question of the benefit of incorporation of the intestinal component in the prediction models. Specificities of intestinal physiology and prevalence of clinical evidence support the importance of intestinal CYP3A; therefore, its contribution to the overall magnitude of inhibition or induction DDIs should not be ignored. Differences observed in the prediction success cannot be associated exclusively with the incorporation of the intestine in the prediction strategy, as values of certain key parameters used in conjunction with the in vitro data are often inconsistent between the datasets (e.g., fmCYP3A4, kdeg). In some instances, use of unrealistically high hepatic input concentrations may compensate for ignoring the contribution of the gut and result in comparable prediction outcome to the scenarios where the inhibitor concentration corrected for the plasma binding is used in conjunction with intestinal inhibition. This review emphasizes the importance of a valid mechanistic model and appropriate interpretation of clinical observations to increase our confidence in the quantitative prediction of the intestinal contribution to the overall pharmacokinetics in human. As the current analysis has shown, incorporation of the intestinal component reduces the number of false negatives; yet, in some instances, depending on the initial model assumptions, it can lead to increased numbers of false-positives, as illustrated in particular in the case of induction DDIs. However, over-prediction of hepatic induction interaction is evident for certain drugs (e.g., tacrolimus) warranting further refinement of the hepatic component in the induction model per se to increase our confidence in overall prediction of these DDIs. In the case of inhibition, prediction models provide more options for the incorporation of the intestinal contribution, using either the pragmatic approach assuming complete inhibition (1/FG) or the more refined predicted FG ratio. A very good agreement between maximum and observed FG ratios in vivo for interactions involving potent inhibitors (regardless of the mechanism) and substrates predominantly metabolised with no transporter issues (e.g., midazolam, nifedipine) supports the utility of this

41

Role of Intestine in the Drug-Drug Interaction Prediction

approach. However, limitations exist for the assessment of DDIs with moderate to weak inhibitors or victim drugs with contribution from both metabolic enzymes and transporters in their disposition (e.g., cyclosporine, tacrolimus). Use of the predicted FG ratio with appropriate intestinal CYP3A4 kdeg is recommended for timedependent DDIs, in particular for the interactions involving victim drugs with high intestinal first-pass (FGº0.25). In addition, current estimates of the predicted FG ratio for induction DDIs are based on the use of induction in vitro data obtained in hepatocytes due to limited availability of induction data in intestinal in vitro systems.172,173) Consideration of the physiological specificities associated with the intestine and the use of corresponding in vitro data may improve our prediction success for induction DDIs. Due to the sensitivity of the DDI prediction models to the accuracy of the FG estimates, the advantages and limitations of different in vivo and in vitro methods to assess FG and inter-individual variability associated with this parameter were discussed. The FG estimates obtained via different approaches in vivo (anhepatic patients, i.v./oral or from grapefruit juice interactions) have been shown to be comparable in the case of midazolam. However, it is evident that this is not always the case, as a large discrepancy between grapefruit juice and i.v./oral FG estimates was seen for drugs with FGº0.25 (e.g., saquinavir). Considering the range of assumptions in both indirect methods and the generally limited number of studies reported for highly extracted drugs (e.g., buspirone), it is advisable to consider the available in vivo data very critically. A number of recent reports have assessed differential prediction strategies (e.g., inclusion of the time course in the model) using midazolam as the only CYP3A4 victim drug in the dataset. Alternatively, midazolam DDIs constituted a large proportion in the database driving the overall prediction success and general recommendations. As illustrated here, midazolam DDIs show good prediction success regardless of the method applied to implement the contribution of intestine (maximum inhibition or predicted FG ratio). Although this is reassuring, the prospective prediction of CYP3A4 DDI for new molecular entities should not entirely rely on the midazolam scenario. A number of drugdependent properties discussed here and variability associated with these parameters need to be considered in order to avoid under-estimation of the FG and subsequent over-estimation of the FG ratio in the DDI prediction. In addition, variability observed in FG estimates due to either health status or gender raises concern in the extrapolation across different populations. The incorporation of the inter-individual variability in FG as a result of differential enzyme/transporter abundance, contribution of polymorphic variants and disease- and age-specific differences in demography is essential in order to

progress further in DDI prediction. In addition, broader application of current physiologically-based pharmacokinetic models as tools to rationalise the alterations in the intestinal clearance in the presence of inhibitor/inducers of either transporters or enzymes is crucial, as well as consideration of any spatial disparities in the location of absorption of victim and perpetrator drugs.

References 1)

2)

3)

4)

5)

6)

7)

8)

9)

10)

11)

12)

13)

Paine, M. F., Hart, H. L., Ludington, S. S., Haining, R. L., Rettie, A. E. and Zeldin, D. C.: The human intestinal cytochrome P450 ``pie''. Drug Metab. Dispos., 34: 880–886 (2006). Lown, K. S., Kolars, J. C., Thummel, K. E., Barnett, J. L., Kunze, K. L., Wrighton, S. A. and Watkins, P.B.: Interpatient heterogeneity in expression of CYP3A4 and CYP3A5 in small bowel. Lack of prediction by the erythromycin breath test. Drug Metab. Dispos., 22: 947–955 (1994). Madani, S., Paine, M. F., Lewis, L., Thummel, K. E. and Shen, D. D.: Comparison of CYP2D6 content and metoprolol oxidation between microsomes isolated from human livers and small intestines. Pharm. Res., 16: 1199–1205 (1999). Glaeser, H., Drescher, S., van der Kuip, H., Behrens, C., Geick, A., Burk, O., Dent, J., Somogyi, A., Von Richter, O., Griese, E. U., Eichelbaum, M. and Fromm, M. F.: Shed human enterocytes as a tool for the study of expression and function of intestinal drug-metabolizing enzymes and transporters. Clin. Pharmacol. Ther., 71: 131–140 (2002). Lin, J. H., Chiba, M. and Baillie, T. A.: Is the role of the small intestine in first-pass metabolism overemphasized? Pharmacol. Rev., 51: 135–158 (1999). Matsumoto, S., Hirama, T., Matsubara, T., Nagata, K. and Yamazoe, Y.: Involvement of CYP2J2 on the intestinal firstpass metabolism of antihistamine drug, astemizole. Drug Metab. Dispos., 30: 1240–1245 (2002). Zhang, Q. Y., Dunbar, D., Ostrowska, A., Zeisloft, S., Yang, J. and Kaminsky, L. S.: Characterization of human small intestinal cytochromes P-450. Drug Metab. Dispos., 27: 804–809 (1999). Wojnowski, L. and Kamdem, L. K.: Clinical implications of CYP3A polymorphisms. Expert. Opin. Drug Metab. Toxicol., 2: 171–182 (2006). Ritter, J. K.: Intestinal UGTs as potential modifiers of pharmacokinetics and biological responses to drugs and xenobiotics. Expert Opin. Drug Metab. Toxicol., 3: 93–107 (2007). Ohno, S. and Nakajin, S.: Determination of mRNA expression of human UDP-glucuronosyltransferases and application for localization in various human tissues by real-time reverse transcriptase-polymerase chain reaction. Drug Metab. Dispos., 37: 32–40 (2009). Tukey, R. H. and Strassburg, C. P.: Human UDPglucuronosyltransferases: metabolism, expression, and disease. Annu. Rev. Pharmacol. Toxicol., 40: 581–616 (2000). Chen, G., Zhang, D., Jing, N., Yin, S., Falany, C. N. and Radominska-Pandya, A.: Human gastrointestinal sulfotransferases: identification and distribution. Toxicol. Appl. Pharmacol., 187: 186–197 (2003). Lindsay, J., Wang, L. L., Li, Y. and Zhou, S. F.: Structure, function and polymorphism of human cytosolic sulfotransferases. Curr. Drug Metab., 9: 99–105 (2008).

42

14)

15)

16)

17)

18)

19)

20)

21)

22)

23)

24)

25)

26)

27)

28)

Aleksandra GALETIN, et al.

Paine, M. F., Khalighi, M., Fisher, J. M., Shen, D. D., Kunze, K. L., Marsh, C. L., Perkins, J. D. and Thummel, K. E.: Characterization of interintestinal and intraintestinal variations in human CYP3A-dependent metabolism. J. Pharmacol. Exp. Ther., 283: 1552–1562 (1997). von Richter, O., Burk, O., Fromm, M. F., Thon, K. P., Eichelbaum, M. and Kivisto, K. T.: Cytochrome P450 3A4 and Pglycoprotein expression in human small intestinal enterocytes and hepatocytes: a comparative analysis in paired tissue specimens. Clin. Pharmacol. Ther., 75: 172–183 (2004). Yang, J., Tucker, G. T. and Rostami-Hodjegan, A.: Cytochrome P450 3A expression and activity in the human small intestine. Clin. Pharmacol. Ther., 76: 391 (2004). Fisher, M. B., Paine, M. F., Strelevitz, T. J. and Wrighton, S. A.: The role of hepatic and extrahepatic UDP-glucuronosyltransferases in human drug metabolism. Drug Metab. Rev., 33: 273–297 (2001). Tukey, R. H. and Strassburg, C. P.: Genetic multiplicity of the human UDP-glucuronosyltransferases and regulation in the gastrointestinal tract. Mol. Pharmacol., 59: 405–414 (2001). Dalvie, D., Obach, R. S., Kang, P., Prakash, C., Loi, C. M., Hurst, S., Nedderman, A., Goulet, L., Smith, E., Bu, H.Z. and Smith, D. A.: Assessment of three human in vitro systems in the generation of major human excretory and circulating metabolites. Chem. Res. Toxicol., 22: 357–368 (2009). Kemp, D. C., Fan, P. W. and Stevens, J. C.: Characterization of raloxifene glucuronidation in vitro: contribution of intestinal metabolism to presystemic clearance. Drug Metab. Dispos., 30: 694–700 (2002). Cubitt, H. E., Houston, J. B. and Galetin, A.: Relative importance of intestinal and hepatic glucuronidation—impact on the prediction of drug clearance. Pharm. Res., 26: 1073–1083 (2009). Watanabe, Y., Nakajima, M. and Yokoi, T.: Troglitazone glucuronidation in human liver and intestine microsomes: high catalytic activity of UGT1A8 and UGT1A10. Drug Metab. Dispos., 30: 1462–1469 (2002). Glatt, H., Boeing, H., Engelke, C. E., Ma, L., Kuhlow, A., Pabel, U., Pomplun, D., Teubner, W. and Meinl, W.: Human cytosolic sulphotransferases: genetics, characteristics, toxicological aspects. Mutat. Res., 482: 27–40 (2001). Coughtrie, M. W.: Sulfation through the looking glass—recent advances in sulfotransferase research for the curious. Pharmacogenomics J., 2: 297–308 (2002). Riches, Z., Stanley, E. L., Bloomer, J. C. and Coughtrie, M. W.: Quantitative evaluation of the expression and activity of five major sulfotransferases (SULTs) in human tissues: the SULT ``pie''. Drug Metab. Dispos., 37: 2255–2261 (2009). Lapple, F., von Richter, O., Fromm, M. F., Richter, T., Thon, K. P., Wisser, H., Griese, E. U., Eichelbaum, M. and Kivisto, K. T.: Differential expression and function of CYP2C isoforms in human intestine and liver. Pharmacogenetics, 13: 565–575 (2003). Thorn, M., Finnstrom, N., Lundgren, S., Rane, A. and Loof, L.: Cytochromes P450 and MDR1 mRNA expression along the human gastrointestinal tract. Br. J. Clin. Pharmacol., 60: 54–60 (2005). Strassburg, C. P., Kneip, S., Topp, J., Obermayer-Straub, P., Barut, A., Tukey, R. H. and Manns, M. P.: Polymorphic gene

29)

30)

31)

32)

33)

34)

35)

36)

37)

38)

39)

40)

41)

42)

regulation and interindividual variation of UDPglucuronosyltransferase activity in human small intestine. J. Biol. Chem., 275: 36164–36171 (2000). Kolars, J. C., Lown, K. S., Schmiedlin-Ren, P., Ghosh, M., Fang, C., Wrighton, S. A., Merion, R. M. and Watkins, P. B.: CYP3A gene expression in human gut epithelium. Pharmacogenetics, 4: 247–259 (1994). Watkins, P. B.: The barrier function of CYP3A4 and Pglycoprotein in the small bowel. Adv. Drug Deliv. Rev., 27: 161–170 (1997). Cao, X., Gibbs, S. T., Fang, L., Miller, H. A., Landowski, C. P., Shin, H. C., Lennernas, H., Zhong, Y., Amidon, G. L., Yu, L. X. and Sun, D.: Why is it challenging to predict intestinal drug absorption and oral bioavailability in human using rat model. Pharm. Res., 23: 1675–1686 (2006). Galetin, A. and Houston, J. B.: Intestinal and hepatic metabolic activity of five cytochrome P450 enzymes: impact on prediction of first-pass metabolism. J. Pharmacol. Exp. Ther., 318: 1220–1229 (2006). Komura, H. and Iwaki, M.: Species differences in in vitro and in vivo small intestinal metabolism of CYP3A substrates. J. Pharm. Sci., 97: 1775–1800 (2007). Obach, R. S., Zhang, Q. Y., Dunbar, D. and Kaminsky, L. S.: Metabolic characterization of the major human small intestinal cytochrome p450s. Drug Metab. Dispos, 29: 347–352 (2001). van de Kerkhof, E. G., de Graaf, I. A. and Groothuis, G. M.: In vitro methods to study intestinal drug metabolism. Curr. Drug Metab., 8: 658–675 (2007). Ito, K. and Houston, J. B.: Prediction of human drug clearance from in vitro and preclinical data using physiologically based and empirical approaches. Pharm. Res., 22: 103–112 (2005). Barter, Z. E., Bayliss, M. K., Beaune, P. H., Boobis, A. R., Carlile, D. J., Edwards, R. J., Houston, J. B., Lake, B. G., Lipscomb, J. C., Pelkonen, O. R., Tucker, G. T. and Rostami-Hodjegan, A.: Scaling factors for the extrapolation of in vivo metabolic drug clearance from in vitro data: reaching a consensus on values of human microsomal protein and hepatocellularity per gram of liver. Curr. Drug Metab., 8: 33–45 (2007). Barter, Z. E., Chowdry, J. E., Harlow, J. R., Snawder, J. E., Lipscomb, J. C. and Rostami-Hodjegan, A.: Covariation of human microsomal protein per gram of liver with age: absence of influence of operator and sample storage may justify interlaboratory data pooling. Drug Metab. Dispos., 36: 2405–2409 (2008). Mohri, K. and Uesawa, Y.: Enzymatic activities in the microsomes prepared from rat small intestinal epithelial cells by differential procedures. Pharm. Res., 18: 1232–1236 (2001). Komura, H., Yasuda, M., Yoshida, N. H. and Sugiyama, Y.: Species difference in nisoldipine oxidation activity in the small intestine. Drug Metab. Pharmacokinet, 17: 427–436 (2002). Akabane, T., Tabata, K., Kadono, K., Sakuda, S., Terashita, S. and Teramura, T.: A comparison of pharmacokinetics between humans and monkeys. Drug Metab. Dispos., in press (2009). van Herwaarden, A. E., Wagenaar, E., van der Kruijssen, C. M., van Waterschoot, R. A., Smit, J. W., Song, J. Y., van der Valk, M. A., van Tellingen, O., van der Hoorn, J. W., Rosing, H., Beijnen, J. H. and Schinkel, A. H.: Knockout of cytochrome P450 3A yields new mouse models for understanding xenobiotic metabolism. J. Clin. Invest., 117: 3583–3592 (2007).

Role of Intestine in the Drug-Drug Interaction Prediction

43)

44)

45)

46)

47)

48)

49)

50)

51)

52)

53)

54)

55)

56)

van Waterschoot, R. A., Rooswinkel, R. W., Sparidans, R. W., van Herwaarden, A. E., Beijnen, J. H. and Schinkel, A. H.: Inhibition and stimulation of intestinal and hepatic CYP3A activity: studies in humanized CYP3A4 transgenic mice using triazolam. Drug Metab. Dispos., 37: 2305–2315 (2009). van Waterschoot, R. A., van Herwaarden, A. E., Lagas, J. S., Sparidans, R. W., Wagenaar, E., van der Kruijssen, C.M., Goldstein, J. A., Zeldin, D. C., Beijnen, J. H. and Schinkel, A. H.: Midazolam metabolism in cytochrome P450 3A knockout mice can be attributed to up-regulated CYP2C enzymes. Mol. Pharmacol., 73: 1029–1036 (2008). Kivisto, K. T., Niemi, M. and Fromm, M. F.: Functional interaction of intestinal CYP3A4 and P-glycoprotein. Fundam. Clin. Pharmacol, 18: 621–626 (2004). Lin, J. H. and Yamazaki, M.: Role of P-glycoprotein in pharmacokinetics: clinical implications. Clin. Pharmacokinet., 42: 59–98 (2003). Wu, C.Y. and Benet, L. Z.: Predicting drug disposition via application of BCS: transport/absorption/elimination interplay and development of a biopharmaceutics drug disposition classification system. Pharm. Res., 22: 11–23 (2005). Paine, M. F., Criss, A. B. and Watkins, P. B.: Two major grapefruit juice components differ in time to onset of intestinal CYP3A4 inhibition. J. Pharmacol. Exp. Ther., 312: 1151–1160 (2005). Fojo, A. T., Ueda, K., Slamon, D. J., Poplack, D. G., Gottesman, M. M. and Pastan, I.: Expression of a multidrugresistance gene in human tumors and tissues. Proc. Natl. Acad. Sci., 84: 265–269 (1987). Mouly, S. and Paine, M.F.: P-glycoprotein increases from proximal to distal regions of human small intestine. Pharm. Res., 20: 1595–1599 (2003). Berggren, S., Gall, C., Wollnitz, N., Ekelund, M., Karlbom, U., Hoogstraate, J., Schrenk, D. and Lennernas, H.: Gene and protein expression of P-glycoprotein, MRP1, MRP2, and CYP3A4 in the small and large human intestine. Mol. Pharm., 4: 252–257 (2007). Nakamura, T., Sakaeda, T., Ohmoto, N., Tamura, T., Aoyama, N., Shirakawa, T., Kamigaki, T., Nakamura, T., Kim, K.I., Kim, S.R., Kuroda, Y., Matsuo, M., Kasuga, M. and Okumura, K.: Real-time quantitative polymerase chain reaction for MDR1, MRP1, MRP2, and CYP3A-mRNA levels in Caco-2 cell lines, human duodenal enterocytes, normal colorectal tissues, and colorectal adenocarcinomas. Drug Metab. Dispos., 30: 4–6 (2002). Endres, C. J., Hsiao, P., Chung, F. S. and Unadkat, J. D.: The role of transporters in drug interactions. Eur. J. Pharm. Sci., 27: 501–517 (2006). Ieiri, I., Higuchi, S. and Sugiyama, Y.: Genetic polymorphisms of uptake (OATP1B1, 1B3) and efflux (MRP2, BCRP) transporters: implications for inter-individual differences in the pharmacokinetics and pharmacodynamics of statins and other clinically relevant drugs. Expert Opin. Drug Metab. Toxicol., 5: 703–729 (2009). Maeda, K. and Sugiyama, Y.: Impact of genetic polymorphisms of transporters on the pharmacokinetic, pharmacodynamic and toxicological properties of anionic drugs. Drug Metab. Pharmacokinet., 23: 223–235 (2008). Glaeser, H., Bailey, D. G., Dresser, G. K., Gregor, J. C.,

57)

58) 59)

60)

61)

62)

63)

64)

65)

66)

67)

68)

69)

70)

71)

72)

43

Schwarz, U. I., McGrath, J. S., Jolicoeur, E., Lee, W., Leake, B. F., Tirona, R. G. and Kim, R. B.: Intestinal drug transporter expression and the impact of grapefruit juice in humans. Clin. Pharmacol. Ther., 81: 362–370 (2007). Ho, R. H. and Kim, R. B.: Transporters and drug therapy: implications for drug disposition and disease. Clin. Pharmacol. Ther., 78: 260–277 (2005). Niemi, M.: Role of OATP transporters in the disposition of drugs. Pharmacogenomics, 8: 787–802 (2007). Shitara, Y., Horie, T. and Sugiyama, Y.: Transporters as a determinant of drug clearance and tissue distribution. Eur. J. Pharm. Sci., 27: 425–446 (2006). Ito, K., Brown, H. S. and Houston, J. B.: Database analyses for the prediction of in vivo drug-drug interactions from in vitro data. Br. J. Clin. Pharmacol., 57: 473–486 (2004). Einolf, H. J.: Comparison of different approaches to predict metabolic drug-drug interactions. Xenobiotica, 37: 1257–1294 (2007). Galetin, A., Ito, K., Hallifax, D. and Houston, J. B.: CYP3A4 substrate selection and substitution in the prediction of potential drug-drug interactions. J. Pharmacol. Exp. Ther., 314: 180–190 (2005). Galetin, A., Burt, H., Gibbons, L. and Houston, J. B.: Prediction of time-dependent CYP3A4 drug-drug interactions: impact of enzyme degradation, parallel elimination pathways, and intestinal inhibition. Drug. Metab. Dispos., 34: 166–175 (2006). Obach, R. S., Walsky, R. L. and Venkatakrishnan, K.: Mechanism-based inactivation of human cytochrome p450 enzymes and the prediction of drug-drug interactions. Drug Metab. Dispos., 35: 246–255 (2007). Obach, R. S., Walsky, R. L., Venkatakrishnan, K., Gaman, E. A., Houston, J. B. and Tremaine, L. M.: The utility of in vitro cytochrome P450 inhibition data in the prediction of drugdrug interactions. J. Pharmacol. Exp. Ther., 316: 336–348 (2006). Shou, M., Hayashi, M., Pan, Y., Xu, Y., Morrissey, K., Xu, L. and Skiles, G. L.: Modeling, prediction, and in vitro in vivo correlation of CYP3A4 induction. Drug Metab. Dispos., 36: 2355–2370 (2008). Rowland-Yeo, Rostami-Hodjegan, A. and Tucker, G. T.: Abundance of cytochrome P450 in human liver: a meta-analysis. Br. J. Clin. Pharmacol., 57: 687–688 (2004). Rostami-Hodjegan, A. and Tucker, G. T.: Simulation and prediction of in vivo drug metabolism in human populations from in vitro data. Nat. Rev. Drug Discov., 6: 140 (2007). Gomez, D. Y., Wacher, V. J., Tomlanovich, S. J., Hebert, M. F. and Benet, L. Z.: The effects of ketoconazole on the intestinal metabolism and bioavailability of cyclosporine. Clin. Pharmacol. Ther., 58: 15–19 (1995). Palkama, V. J., Ahonen, J., Neuvonen, P. J. and Olkkola, K. T.: Effect of saquinavir on the pharmacokinetics and pharmacodynamics of oral and intravenous midazolam. Clin. Pharmacol. Ther., 66: 33–39 (1999). Saari, T. I., Laine, K., Leino, K., Valtonen, M., Neuvonen, P. J. and Olkkola, K. T.: Effect of voriconazole on the pharmacokinetics and pharmacodynamics of intravenous and oral midazolam. Clin. Pharmacol. Ther., 79: 362–370 (2006). Tsunoda, S. M., Velez, R. L., von Moltke, L. L. and Greenblatt,

44

73)

74)

75)

76)

77)

78)

79)

80)

81)

82)

83)

84)

85)

Aleksandra GALETIN, et al.

D. J.: Differentiation of intestinal and hepatic cytochrome P450 3A activity with use of midazolam as an in vivo probe: effect of ketoconazole. Clin. Pharmacol. Ther., 66: 461–471 (1999). Holtbecker, N., Fromm, M. F., Kroemer, H. K., Ohnhaus, E. E. and Heidemann, H.: The nifedipine-rifampin interaction. Evidence for induction of gut wall metabolism. Drug Metab. Dispos., 24: 1121–1123 (1996). Wu, C. Y., Benet, L. Z., Hebert, M. F., Gupta, S. K., Rowland, M., Gomez, D. Y. and Wacher, V. J.: Differentiation of absorption and first-pass gut and hepatic metabolism in humans: studies with cyclosporine. Clin. Pharmacol. Ther., 58: 492–497 (1995). Gorski, J. C., Vannaprasaht, S., Hamman, M. A., Ambrosius, W. T., Bruce, M. A., Haehner-Daniels, B. and Hall, S.D.: The effect of age, sex, and rifampin administration on intestinal and hepatic cytochrome P450 3A activity. Clin. Pharmacol. Ther., 74: 275–287 (2003). Benet, L. Z., Cummins, C. L. and Wu, C. Y.: Unmasking the dynamic interplay between efflux transporters and metabolic enzymes. Int. J. Pharm., 277: 3–9 (2004). Sandstrom, R., Knutson, T. W., Knutson, L., Jansson, B. and Lennernas, H.: The effect of ketoconazole on the jejunal permeability and CYP3A metabolism of (R/S)-verapamil in humans. Br. J. Clin. Pharmacol., 48: 180–189 (1999). Lin, J.H.: Drug-drug interaction mediated by inhibition and induction of P-glycoprotein. Adv. Drug Deliv. Rev., 55: 53–81 (2003). Uchaipichat, V., Winner, L. K., Mackenzie, P. I., Elliot, D. J., Williams, J. A. and Miners, J. O.: Quantitative prediction of in vivo inhibitory interactions involving glucuronidated drugs from in vitro data: the effect of fluconazole on zidovudine glucuronidation. Br. J. Clin. Pharmacol., 61: 427–439 (2006). Williams, J. A., Hyland, R., Jones, B. C., Smith, D. A., Hurst, S., Goosen, T. C., Peterkin, V., Koup, J. R. and Ball, S. E.: Drugdrug interactions for UDP-glucuronosyltransferase substrates: a pharmacokinetic explanation for typically observed low exposure (AUCi/AUC) ratios. Drug Metab. Dispos., 32: 1201–1208 (2004). Miners, J. O., Mackenzie, P. I. and Knights, K. M.: The prediction of drug-glucuronidation parameters in humans: UDPglucuronosyltransferase enzyme-selective substrate and inhibitor probes for reaction phenotyping and in vitro-in vivo extrapolation of drug clearance and drug-drug interaction potential. Drug Metab. Rev., 42: 189–201 (2009). Ducharme, M. P., Warbasse, L. H. and Edwards, D. J.: Disposition of intravenous and oral cyclosporine after administration with grapefruit juice. Clin. Pharmacol. Ther., 57: 485–491 (1995). Ku, Y. M., Min, D. I. and Flanigan, M.: Effect of grapefruit juice on the pharmacokinetics of microemulsion cyclosporine and its metabolite in healthy volunteers: does the formulation difference matter? J. Clin. Pharmacol., 38: 959–965 (1998). Kupferschmidt, H. H., Ha, H. R., Ziegler, W. H., Meier, P. J. and Krahenbuhl, S.: Interaction between grapefruit juice and midazolam in humans. Clin. Pharmacol. Ther., 58: 20–28 (1995). Lundahl, J., Regardh, C. G., Edgar, B. and Johnsson, G.: Effects of grapefruit juice ingestion—pharmacokinetics and hae-

86)

87)

88)

89)

90)

91)

92)

93)

94)

95)

96)

97)

98)

99)

modynamics of intravenously and orally administered felodipine in healthy men. Eur. J. Clin. Pharmacol., 52: 139–145 (1997). Gertz, M., Davis, J. D., Harrison, A., Houston, J. B. and Galetin, A.: Grapefruit juice-drug interaction studies as a method to assess the extent of intestinal availability: utility and limitations. Curr. Drug Metab., 9: 785–795 (2008). Saito, M., Hirata-Koizumi, M., Matsumoto, M., Urano, T. and Hasegawa, R.: Undesirable effects of citrus juice on the pharmacokinetics of drugs: focus on recent studies. Drug Saf., 28: 677–694 (2005). Kharasch, E.D., Walker, A., Hoffer, C. and Sheffels, P.: Intravenous and oral alfentanil as in vivo probes for hepatic and first-pass cytochrome P450 3A activity: noninvasive assessment by use of pupillary miosis. Clin. Pharmacol. Ther., 76: 452–466 (2004). Brown, H. S., Galetin, A., Hallifax, D. and Houston, J. B.: Prediction of in vivo drug-drug interactions from in vitro data: factors affecting prototypic drug-drug interactions involving CYP2C9, CYP2D6 and CYP3A4. Clin. Pharmacokinet., 45: 1035–1050 (2006). Youdim, K. A., Zayed, A., Dickins, M., Phipps, A., Griffiths, M., Darekar, A., Hyland, R., Fahmi, O., Hurst, S., Plowchalk, D. R., Cook, J., Guo, F. and Obach, R. S.: Application of CYP3A4 in vitro data to predict clinical drug-drug interactions; predictions of compounds as objects of interaction. Br. J. Clin. Pharmacol., 65: 680–692 (2008). Fahmi, O. A., Maurer, T. S., Kish, M., Cardenas, E., Boldt, S. and Nettleton, D.: A combined model for predicting CYP3A4 clinical net drug-drug interaction based on CYP3A4 inhibition, inactivation, and induction determined in vitro. Drug Metab. Dispos., 36: 1698–1708 (2008). Obach, R. S.: Predicting drug-drug interactions from in vitro drug metabolism data: challenges and recent advances. Curr. Opin. Drug Discov. Devel., 12: 81–89 (2009). Kanamitsu, S., Ito, K. and Sugiyama, Y.: Quantitative prediction of in vivo drug-drug interactions from in vitro data based on physiological pharmacokinetics: use of maximum unbound concentration of inhibitor at the inlet to the liver. Pharm. Res., 17: 336–343 (2000). Houston, J. B. and Galetin, A.: Progress towards prediction of human pharmacokinetic parameters from in vitro technologies. Drug Metab. Rev., 35: 393–415 (2003). Houston, J. B. and Galetin, A.: Methods for predicting in vivo pharmacokinetics using data from in vitro assays. Curr. Drug Metab., 9: 940–951 (2008). Hinton, L. K., Galetin, A. and Houston, J. B.: Multiple inhibition mechanisms and prediction of drug-drug interactions: status of metabolism and transporter models as exemplified by gemfibrozil-drug interactions. Pharm. Res., 25: 1063–1074 (2008). Rostami-Hodjegan, A. and Tucker, G.: `In silico' simulations to assess the `in vivo' consequences of `in vitro' metabolic drugdrug interactions. Drug Discovery Today: Technologies, 1: 441 (2004). Isoherranen, N., Kunze, K. L., Allen, K. E., Nelson, W. L. and Thummel, K. E.: Role of itraconazole metabolites in CYP3A4 inhibition. Drug Metab. Dispos., 32: 1121–1131 (2004). Zhang, X., Jones, D. R. and Hall, S. D.: Prediction of the effect

Role of Intestine in the Drug-Drug Interaction Prediction

100)

101)

102)

103)

104)

105)

106)

107)

108)

109)

110)

111)

112)

of erythromycin, diltiazem, and their metabolites, alone and in combination, on CYP3A4 inhibition. Drug Metab. Dispos., 37: 150–160 (2009). Zhang, X., Quinney, S. K., Gorski, J. C., Jones, D. R. and Hall, S. D.: Semiphysiologically based pharmacokinetic models for the inhibition of midazolam clearance by diltiazem and its major metabolite. Drug Metab. Dispos., 37: 1587–1597 (2009). Yang, J., Liao, M., Shou, M., Jamei, M., Yeo, K. R., Tucker, G. T. and Rostami-Hodjegan, A.: Cytochrome p450 turnover: regulation of synthesis and degradation, methods for determining rates, and implications for the prediction of drug interactions. Curr. Drug Metab., 9: 384–394 (2008). Venkatakrishnan, K. and Obach, R. S.: In vitro-in vivo extrapolation of CYP2D6 inactivation by paroxetine: prediction of nonstationary pharmacokinetics and drug interaction magnitude. Drug Metab. Dispos., 33: 845–852 (2005). Brown, H.S., Ito, K., Galetin, A. and 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: 508–518 (2005). Ito, K., Hallifax, D., Obach, R. S. and 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: 837–844 (2005). Houston, J.B. and Galetin, A.: Modelling atypical CYP3A4 kinetics: principles and pragmatism. Arch. Biochem. Biophys., 433: 351–360 (2005). Galetin, A., Hinton, L. K., Burt, H., Obach, R. S. and Houston, J. B.: Maximal inhibition of intestinal first-pass metabolism as a pragmatic indicator of intestinal contribution to the drug-drug interactions for CYP3A4 cleared drugs. Curr. Drug Metab., 8: 685–693 (2007). Wang, Y. H., Jones, D. R. and Hall, S. D.: Prediction of cytochrome P450 3A inhibition by verapamil enantiomers and their metabolites. Drug Metab. Dispos., 32: 259–266 (2004). Fahmi, O. A., Hurst, S., Plowchalk, D., Cook, J., Guo, F., Youdim, K., Dickins, M., Phipps, A., Darekar, A., Hyland, R. and Obach, R. S.: Investigation of different algorithms for predicting clinical drug-drug interactions, based on the use of CYP3A4 in vitro data; predictions of compounds as precipitants of interaction. Drug Metab. Dispos., 37: 1658–1666 (2009). Tucker, G. T., Houston, J. B. and Huang, S. M.: Optimizing drug development: strategies to assess drug metabolism/transporter interaction potential—toward a consensus. Clin. Pharmacol. Ther., 70: 103–114 (2001). Yao, C. and Levy, R. H.: Inhibition-based metabolic drug-drug interactions: predictions from in vitro data. J. Pharm. Sci., 91: 1923–1935 (2002). Bjornsson, T. D., Callaghan, J. T., Einolf, H. J., Fischer, V., Gan, L., Grimm, S., Kao, J., King, S. P., Miwa, G., Ni, L., Kumar, G., McLeod, J., Obach, R. S., Roberts, S., Roe, A., Shah, A., Snikeris, F., Sullivan, J. T., Tweedie, D., Vega, J. M., Walsh, J. and Wrighton, S. A.: The conduct of in vitro and in vivo drug-drug interaction studies: a Pharmaceutical Research and Manufacturers of America (PhRMA) perspective. Drug Metab. Dispos., 31: 815–832 (2003). Huang, S. M., Strong, J. M., Zhang, L., Reynolds, K. S., Nallani,

113)

114)

115) 116)

117)

118)

119)

120)

121)

122)

123)

124)

45

S., Temple, R., Abraham, S., Habet, S. A., Baweja, R. K., Burckart, G. J., Chung, S., Colangelo, P., Frucht, D., Green, M. D., Hepp, P., Karnaukhova, E., Ko, H. S., Lee, J. I., Marroum, P. J., Norden, J. M., Qiu, W., Rahman, A., Sobel, S., Stifano, T., Thummel, K., Wei, X. X., Yasuda, S., Zheng, J. H., Zhao, H. and Lesko, L. J.: New era in drug interaction evaluation: US Food and Drug Administration update on CYP enzymes, transporters, and the guidance process. J. Clin. Pharmacol., 48: 662–670 (2008). Huang, S. M., Temple, R., Throckmorton, D. C. and Lesko, L. J.: Drug interaction studies: study design, data analysis, and implications for dosing and labeling. Clin. Pharmacol. Ther., 81: 298–304 (2007). Grimm, S. W., Einolf, H. J., Hall, S. D., He, K., Lim, H. K., Ling, K. H., Lu, C., Nomeir, A. A., Seibert, E., Skordos, K. W., Tonn, G. R., Van Horn, R., Wang, R. W., Wong, Y. N., Yang, T. J. and Obach, R. S.: The conduct of in vitro studies to address time-dependent inhibition of drug-metabolizing enzymes: a perspective of the pharmaceutical research and manufacturers of America. Drug Metab. Dispos., 37: 1355–1370 (2009). Rowland, M. and Matin, S. B.: Kinetics of drug-drug interactions. J. Pharmacokinet. Biopharm., 1: 553–567 (1973). Chien, J. Y., Lucksiri, A., Ernest, C. S., 2nd, Gorski, J. C., Wrighton, S. A. and Hall, S. D.: Stochastic prediction of CYP3A-mediated inhibition of midazolam clearance by ketoconazole. Drug Metab. Dispos., 34: 1208–1219 (2006). Chu, V., Einolf, H. J., Evers, R., Kumar, G., Moore, D., Ripp, S., Silva, J., Sinha, V., Sinz, M. and Skerjanec, A.: In vitro and in vivo induction of cytochrome p450: a survey of the current practices and recommendations: a pharmaceutical research and manufacturers of america perspective. Drug Metab. Dispos., 37: 1339–1354 (2009). Ito, K., Iwatsubo, T., Kanamitsu, S., Ueda, K., Suzuki, H. and Sugiyama, Y.: Prediction of pharmacokinetic alterations caused by drug-drug interactions: metabolic interaction in the liver. Pharmacol. Rev., 50: 387–412 (1998). Jamei, M., Marciniak, S., Feng, K., Barnett, A., Tucker, G. and Rostami-Hodjegan, A.: The Simcyp}) population-based ADME simulator. Expert Opin. Drug Metab. Toxicol., 5: 211–223 (2009). Almond, L. M., Yang, J., Jamei, M., Tucker, G. T. and RostamiHodjegan, A.: Towards a quantitative framework for the prediction of DDIs arising from cytochrome P450 induction. Curr. Drug Metab., 10: 420–432 (2009). Obach, R. S., Walsky, R. L., Venkatakrishnan, K., Houston, J. B. and Tremaine, L. M.: In vitro cytochrome P450 inhibition data and the prediction of drug-drug interactions: qualitative relationships, quantitative predictions, and the rank-order approach. Clin. Pharmacol. Ther., 78: 582–592 (2005). Galetin, A., Gertz, M. and Houston, J. B.: Potential role of intestinal first-pass metabolism in the prediction of drug-drug interactions. Expert Opin. Drug Metab. Toxicol., 4: 909–922 (2008). Bailey, D. G., Malcolm, J., Arnold, O. and Spence, J. D.: Grapefruit juice-drug interactions. Br. J. Clin. Pharmacol., 46: 101–110 (1998). Ghanbari, F., Rowland-Yeo, K., Bloomer, J. C., Clarke, S. E., Lennard, M. S., Tucker, G. T. and Rostami-Hodjegan, A.: A

46

125)

126)

127)

128)

129)

130)

131)

132)

133)

134)

135)

136)

137)

138)

Aleksandra GALETIN, et al.

critical evaluation of the experimental design of studies of mechanism based enzyme inhibition, with implications for in vitro-in vivo extrapolation. Curr. Drug Metab., 7: 315–334 (2006). Borm, P. J., Koster, A. S., Frankhuijzen-Sierevogel, A. and Noordhoek, J.: Comparison of two cell isolation procedures to study in vitro intestinal wall biotransformation in control and 3-methyl-cholanthrene pretreated rats. Cell Biochem. Funct., 1: 161–167 (1983). Dressman, J. B., Thelen, K. and Jantratid, E.: Towards quantitative prediction of oral drug absorption. Clin. Pharmacokinet., 47: 655–667 (2008). Dressman, J. B., Vertzoni, M., Goumas, K. and Reppas, C.: Estimating drug solubility in the gastrointestinal tract. Adv. Drug Deliv. Rev., 59: 591–602 (2007). Templeton, I. E., Thummel, K. E., Kharasch, E. D., Kunze, K. L., Hoffer, C., Nelson, W. L. and Isoherranen, N.: Contribution of itraconazole metabolites to inhibition of CYP3A4 in vivo. Clin. Pharmacol. Ther., 83: 77–85 (2008). Quinney, S. K., Galinsky, R. E., Jiyamapa-Serna, V. A., Chen, Y., Hamman, M. A., Hall, S. D. and Kimura, R. E.: Hydroxyitraconazole, formed during intestinal first-pass metabolism of itraconazole, controls the time course of hepatic CYP3A inhibition and the bioavailability of itraconazole in rats. Drug Metab. Dispos., 36: 1097–1101 (2008). Granger, D. N., Richardson, P. D., Kvietys, P. R. and Mortillaro, N. A.: Intestinal blood flow. Gastroenterology, 78: 837–863 (1980). Dregelid, E., Haukaas, S., Amundsen, S., Eide, G. E., Soreide, O., Lekven, J. and Svanes, K.: Microsphere method in measurement of blood flow to wall layers of small intestine. Am. J. Physiol., 250: G670–678 (1986). Zhang, L., Zhang, Y. D., Strong, J. M., Reynolds, K. S. and Huang, S. M.: A regulatory viewpoint on transporter-based drug interactions. Xenobiotica, 38: 709–724 (2008). Fenner, K. S., Troutman, M. D., Kempshall, S., Cook, J. A., Ware, J. A., Smith, D. A. and Lee, C. A.: Drug-drug interactions mediated through P-glycoprotein: clinical relevance and in vitro-in vivo correlation using digoxin as a probe drug. Clin. Pharmacol. Ther., 85: 173–181 (2009). Quinney, S. K., Haehner, B. D., Rhoades, M. B., Lin, Z., Gorski, J. C. and Hall, S. D.: Interaction between midazolam and clarithromycin in the elderly. Br. J. Clin. Pharmacol., 65: 98–109 (2008). Floren, L. C., Bekersky, I., Benet, L. Z., Mekki, Q., Dressler, D., Lee, J. W., Roberts, J. P. and Hebert, M. F.: Tacrolimus oral bioavailability doubles with coadministration of ketoconazole. Clin. Pharmacol. Ther., 62: 41–49 (1997). Kato, M., Chiba, K., Horikawa, M. and Sugiyama, Y.: The quantitative prediction of in vivo enzyme-induction caused by drug exposure from in vitro information on human hepatocytes. Drug Metab. Pharmacokinet., 20: 236–243 (2005). Fahmi, O. A., Boldt, S., Kish, M., Obach, R. S. and Tremaine, L. M.: Prediction of drug-drug interactions from in vitro induction data: application of the relative induction score approach using cryopreserved human hepatocytes. Drug Metab. Dispos., 36: 1971–1974 (2008). Hebert, M. F., Fisher, R. M., Marsh, C. L., Dressler, D. and Bekersky, I.: Effects of rifampin on tacrolimus pharmacokinetics

139)

140)

141)

142)

143)

144)

145)

146)

147) 148)

149) 150)

151)

152)

153)

in healthy volunteers. J. Clin. Pharmacol., 39: 91–96 (1999). Hebert, M. F., Roberts, J.P., Prueksaritanont, T. and Benet, L. Z.: Bioavailability of cyclosporine with concomitant rifampin administration is markedly less than predicted by hepatic enzyme induction. Clin. Pharmacol. Ther., 52: 453–457 (1992). Paine, M. F., Shen, D. D., Kunze, K. L., Perkins, J. D., Marsh, C. L., McVicar, J. P., Barr, D. M., Gillies, B. S. and Thummel, K. E.: First-pass metabolism of midazolam by the human intestine. Clin. Pharmacol‚ Ther‚, 60: 14–24 (1996). Thummel, K. E., O'Shea, D., Paine, M. F., Shen, D. D., Kunze, K. L., Perkins, J. D. and Wilkinson, G. R.: Oral first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism. Clin. Pharmacol. Ther., 59: 491–502 (1996). Hall, S. D., Thummel, K. E., Watkins, P. B., Lown, K. S., Benet, L. Z., Paine, M. F., Mayo, R. R., Turgeon, D. K., Bailey, D. G., Fontana, R. J. and Wrighton, S. A.: Molecular and physical mechanisms of first-pass extraction. Drug Metab. Dispos., 27: 161–166 (1999). Benet, L. Z., Oie, S. and JB, S.: Appendix II design and optimization of dosage regimens: pharmacokinetic data. In The Pharmacological Basis of Therapeutics, Ninth Edition. Hardman, J. G., Limbird, L. E., et al. (eds.), The McGraw Companies, Inc., (1996). Jusko, W. J. and D'Ambrosio, R.: Monitoring FK 506 concentrations in plasma and whole blood. Transplant. Proc., 23: 2732–2735 (1991). Kato, M., Chiba, K., Hisaka, A., Ishigami, M., Kayama, M., Mizuno, N., Nagata, Y., Takakuwa, S., Tsukamoto, Y., Ueda, K., Kusuhara, H., Ito, K. and Sugiyama, Y.: The intestinal firstpass metabolism of substrates of CYP3A4 and Pglycoprotein—quantitative analysis based on information from the literature. Drug Metab. Pharmacokinet., 18: 365–372 (2003). Bengtsson-Hasselgren, B., Ronn, O., Blychert, L. O., Edgar, B. and Raner, S.: Acute effects of felodipine and nifedipine on hepatic and forearm blood flow in healthy men. Eur. J. Clin. Pharmacol., 38: 529–533 (1990). Meredith, P. A., Elliott, H. L. and Reid, J. L.: Verapamil pharmacokinetics. Br. J. Clin. Pharmacol., 21: 554–555 (1986). Fuhr, U.: Drug interactions with grapefruit juice. Extent, probable mechanism and clinical relevance. Drug Saf., 18: 251–272 (1998). Kirby, B. J. and Unadkat, J. D.: Grapefruit juice, a glass full of drug interactions? Clin. Pharmacol. Ther., 81: 631–633 (2007). Dresser, G. K. and Bailey, D. G.: The effects of fruit juices on drug disposition: a new model for drug interactions. Eur. J. Clin. Invest., 33 Suppl 2: 10–16 (2003). Lau, Y. Y., Huang, Y., Frassetto, L. and Benet, L. Z.: Effect of OATP1B transporter inhibition on the pharmacokinetics of atorvastatin in healthy volunteers. Clin. Pharmacol. Ther., 81: 194–204 (2007). Shitara, Y. and Sugiyama, Y.: Pharmacokinetic and pharmacodynamic alterations of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors: drug-drug interactions and interindividual differences in transporter and metabolic enzyme functions. Pharmacol. Ther., 112: 71–105 (2006). Lennernas, H.: Clinical pharmacokinetics of atorvastatin. Clin.

Role of Intestine in the Drug-Drug Interaction Prediction

154)

155)

156)

157)

158)

159)

160)

161)

162)

163)

164)

165)

166)

167)

Pharmacokinet., 42: 1141–1160 (2003). Farkas, D., Oleson, L. E., Zhao, Y., Harmatz, J. S., Zinny, M. A., Court, M. H. and Greenblatt, D. J.: Pomegranate juice does not impair clearance of oral or intravenous midazolam, a probe for cytochrome P450-3A activity: comparison with grapefruit juice. J. Clin. Pharmacol., 47: 286–294 (2007). Kharasch, E. D., Walker, A., Isoherranen, N., Hoffer, C., Sheffels, P., Thummel, K., Whittington, D. and Ensign, D.: Influence of CYP3A5 genotype on the pharmacokinetics and pharmacodynamics of the cytochrome P4503A probes alfentanil and midazolam. Clin. Pharmacol. Ther., 82: 410–426 (2007). Eichelbaum, M., Albrecht, M., Kliems, G., Schafer, K. and Somogyi, A.: Influence of meso-caval shunt surgery on verapamil kinetics, bioavailability and response. Br. J. Clin. Pharmacol., 10: 527–530 (1980). Chalasani, N., Gorski, J. C., Patel, N. H., Hall, S. D. and Galinsky, R. E.: Hepatic and intestinal cytochrome P450 3A activity in cirrhosis: effects of transjugular intrahepatic portosystemic shunts. Hepatology, 34: 1103–1108 (2001). Kleinbloesem, C. H., van Harten, J., Wilson, J. P., Danhof, M., van Brummelen, P. and Breimer, D. D.: Nifedipine: kinetics and hemodynamic effects in patients with liver cirrhosis after intravenous and oral administration. Clin. Pharmacol. Ther., 40: 21–28 (1986). Rostami-Hodjegan, A. and Tucker, G. T.: The effects of portal shunts on intestinal cytochrome P450 3A activity. Hepatology, 35: 1549–1550; author reply 1550–1551 (2002). McConn, D. J., 2nd, Lin, Y. S., Mathisen, T. L., Blough, D. K., Xu, Y., Hashizume, T., Taylor, S. L., Thummel, K. E. and Shuhart, M. C.: Reduced duodenal cytochrome P450 3A protein expression and catalytic activity in patients with cirrhosis. Clin. Pharmacol. Ther., 85: 387–393 (2009). Ito, K., Kusuhara, H. and Sugiyama, Y.: Effects of intestinal CYP3A4 and P-glycoprotein on oral drug absorption—theoretical approach. Pharm. Res., 16: 225–231 (1999). Pang, K. S.: Modeling of intestinal drug absorption: roles of transporters and metabolic enzymes (for the Gillette Review Series). Drug Metab. Dispos., 31: 1507–1519 (2003). Tam, D., Tirona, R. G. and Pang, K. S.: Segmental intestinal transporters and metabolic enzymes on intestinal drug absorption. Drug Metab. Dispos., 31: 373–383 (2003). Badhan, R., Penny, J., Galetin, A. and Houston, J. B.: Methodology for development of a physiological model incorporating CYP3A and P-glycoprotein for the prediction of intestinal drug absorption. J. Pharm. Sci., 98: 2180–2197 (2009). Jamei, M., Turner, D., Yang, J., Neuhoff, S., Polak, S., RostamiHodjegan, A. and Tucker, G.: Population-based mechanistic prediction of oral drug absorption. AAPS. J., 11: 225–237 (2009). Rowland, M., Benet, L. Z. and Graham, G. G.: Clearance concepts in pharmacokinetics. J. Pharmacokinet. Biopharm., 1: 123–136 (1973). Wilkinson, G. R. and Shand, D. G.: Commentary: a physiologi-

47

cal approach to hepatic drug clearance. Clin. Pharmacol. Ther., 18: 377–390 (1975). 168) Yu, L. X. and Amidon, G. L.: A compartmental absorption and transit model for estimating oral drug absorption. Int. J. Pharm., 186: 119–125 (1999). 169) Agoram, B., Woltosz, W. S. and Bolger, M. B.: Predicting the impact of physiological and biochemical processes on oral drug bioavailability. Adv. Drug Deliv. Rev., 50 Suppl 1: S41–67 (2001). 170) Dokoumetzidis, A., Kalantzi, L. and Fotaki, N.: Predictive models for oral drug absorption: from in silico methods to integrated dynamical models. Expert Opin. Drug Metab. Toxicol., 3: 491–505 (2007). 171) Pang, K. C., Maeng, H. J. and Fan, J.: Interplay of transporters and enzymes in drug and metabolite processing. Mol. Pharm., (2009). 172) Gupta, A., Mugundu, G. M., Desai, P. B., Thummel, K. E. and Unadkat, J. D.: Intestinal human colon adenocarcinoma cell line LS180 is an excellent model to study pregnane X receptor, but not constitutive androstane receptor, mediated CYP3A4 and multidrug resistance transporter 1 induction: studies with anti-human immunodeficiency virus protease inhibitors. Drug Metab. Dispos., 36: 1172–1180 (2008). 173) van de Kerkhof, E. G., de Graaf, I. A., Ungell, A. L. and Groothuis, G. M.: Induction of metabolism and transport in human intestine: validation of precision-cut slices as a tool to study induction of drug metabolism in human intestine in vitro. Drug Metab. Dispos., 36: 604–613 (2008). 174) Gorski, J. C., Jones, D. R., Haehner-Daniels, B. D., Hamman, M. A., O'Mara, E. M., Jr. and Hall, S. D.: The contribution of intestinal and hepatic CYP3A to the interaction between midazolam and clarithromycin. Clin. Pharmacol. Ther., 64: 133–143 (1998). 175) Olkkola, K. T., Ahonen, J. and Neuvonen, P. J.: The effects of the systemic antimycotics, itraconazole and fluconazole, on the pharmacokinetics and pharmacodynamics of intravenous and oral midazolam. Anesth. Analg., 82: 511–516 (1996). 176) Olkkola, K. T., Aranko, K., Luurila, H., Hiller, A., Saarnivaara, L., Himberg, J. J. and Neuvonen, P. J.: A potentially hazardous interaction between erythromycin and midazolam. Clin. Pharmacol. Ther., 53: 298–305 (1993). 177) Rashid, T. J., Martin, U., Clarke, H., Waller, D. G., Renwick, A. G. and George, C. F.: Factors affecting the absolute bioavailability of nifedipine. Br. J. Clin. Pharmacol., 40: 51–58 (1995). 178) Kupferschmidt, H. H., Fattinger, K. E., Ha, H. R., Follath, F. and Krahenbuhl, S.: Grapefruit juice enhances the bioavailability of the HIV protease inhibitor saquinavir in man. Br. J. Clin. Pharmacol., 45: 355–359 (1998). 179) Venkatakrishnan, K. and Obach, R. S.: Drug-drug interactions via mechanism-based cytochrome P450 inactivation: points to consider for risk assessment from in vitro data and clinical pharmacologic evaluation. Curr. Drug Metab., 8: 449–462 (2007).