Supplemental material to this article can be found at: http://dmd.aspetjournals.org/content/suppl/2016/03/29/dmd.115.066845.DC1 1521-009X/44/6/821–832$25.00 http://dx.doi.org/10.1124/dmd.115.066845 DRUG METABOLISM AND DISPOSITION Drug Metab Dispos 44:821–832, June 2016 Copyright ª 2016 The Author(s) This is an open access article distributed under the CC-BY Attribution 4.0 International license. Prediction of Drug-Drug Interactions Arising from CYP3A induction Using a Physiologically Based Dynamic Model s

Lisa M. Almond, Sophie Mukadam, Iain Gardner, Krystle Okialda, Susan Wong, Oliver Hatley, Suzanne Tay, Karen Rowland-Yeo, Masoud Jamei, Amin Rostami-Hodjegan, and Jane R. Kenny

Simcyp (a Certara Company), Sheffield, United Kingdom (L.M.A., I.G., O.H., K.R.-Y., M.J., A.R.-H.); DMPK, Genentech Inc., South San Francisco, California (S.M., K.O., S.W., S.T., J.R.K.); and Manchester Pharmacy School, University of Manchester, United Kingdom (A.R.-H.)

Received August 18, 2015; accepted March 28, 2016

ABSTRACT

Using physiologically based pharmacokinetic modeling, we pre- carbamazepine, phenytoin, and phenobarbital in hepatocytes from Downloaded from dicted the magnitude of drug-drug interactions (DDIs) for studies with four donors, were then used to evaluate use of the refined rifampicin rifampicin and seven CYP3A4 probe substrates administered i.v. (10 model for calibration. Calibration of mRNA and activity data for other studies) or orally (19 studies). The results showed a tendency to inducers using the refined rifampicin model led to more accurate underpredict the DDI magnitude when the victim drug was adminis- DDI predictions compared with the initial model (activity GMFE 1.49 vs. tered orally. Possible sources of inaccuracy were investigated sys- 1.68; mRNA GMFE 1.35 vs. 1.46), suggesting that robust in vivo refer- tematically to determine the most appropriate model refinement. When ence values can be used to overcome interdonor and laboratory-to- dmd.aspetjournals.org the maximal fold induction (Indmax) for rifampicin was increased (from 8 laboratory variability. Use of uncalibrated data also performed well to 16) in both the liver and the gut, or when the Indmax was increased in (GMFE 1.39 and 1.44 for activity and mRNA). As a result of experimental the gut but not in liver, there was a decrease in bias and increased variability (i.e., in donors and protocols), it is prudent to fully characterize precision compared with the base model (Indmax = 8) [geometric in vitro induction with prototypical inducers to give an understanding of mean fold error (GMFE) 2.12 vs. 1.48 and 1.77, respectively]. how that particular system extrapolates to the in vivo situation when Induction parameters (mRNA and activity), determined for rifampicin, using an uncalibrated approach. at ASPET Journals on October 1, 2021

Introduction whereas dynamic models account for changes in perpetrator concentration Over recent years, the use of in vitro-in vivo extrapolation linked with with time (Einolf, 2007; Almond et al., 2009; EMA, 2012; FDA, 2012). physiologically based pharmacokinetic (IVIVE-PBPK) models that The concentration used as the input (driving) concentration for the integrate key in vitro drug parameters with human system parameters prediction drug interactions [e.g., inlet (portal vein) vs. outlet (liver) vs. (e.g., demography, physiology, genetics) to predict pharmacokinetics Cmax (systemic)] and whether the total or unbound concentrations that are and drug-drug interactions (DDIs) and to assist in decision making has used can vary across static methods (Almond et al., 2009), with some become increasingly common (EMA, 2012; Rostami-Hodjegan et al., regulatory guidance favoring more cautious approaches using total 2012; Huang et al., 2013). More recently, these approaches have also concentrations in the basic models but unbound concentrations in the been used to inform the wording of product information labels (Janssen mechanistic static models (FDA, 2012). Although the overall effect of Biotech, 2013a,b; Imbruvica: Highlights of Prescribing Information, time-dependent inhibition and induction at the new enzyme steady-state http://www.imbruvica.com/downloads/Prescribing_Information.pdf’ and level can be simulated only by using static approaches, investigation of Olysio: Highlights of Prescribing Information, http://www.olysio.com/ the time course can be simulated using dynamic models that factor in the shared/product/olysio/prescribing-information.pdf). In particular, the changing concentrations of substrate, perpetrator, as well as enzyme. An benefits of adopting mechanistic approaches (including information on additional advantage of the dynamic models (particularly in the case of both the perpetrator and victim drug, e.g., fraction metabolized (fm)and competitive inhibition) is to enable evaluation of the dosing schedule dependence of the DDI and possible strategies to minimize such effects. fraction metabolized in the gut (FG) over purely pragmatic approaches have been recognized (Einolf, 2007; Almond et al., 2009; FDA, 2012). Although dynamic approaches have increased complexity compared Mechanistic models can be further classified as either dynamic or static. with static approaches, they make fewer assumptions and are necessary Static models assume a constant perpetrator concentration throughout the if the intention is to account for phenomena such as autoinduction, where full dosing interval and ignore temporal changes in concentrations, the perpetrator induces enzyme levels, in turn increasing its own metabolism and thereby altering concentrations achieved with sub- sequent doses. This in turn impacts the level of enzyme achieved when dx.doi.org/10.1124/dmd.115.066845. the system reaches steady state. Here, our focus is the dynamic s This article has supplemental material available at dmd.aspetjournals.org. prediction of induction potential of a new drug using IVIVE-PBPK, as

ABBREVIATIONS: AUC, area under the curve; CBZ, carbamazepine; DDI, drug-drug interaction, FG, fraction metabolized in the gut; fu, fraction unbound; fu,gut, fraction unbound in the gut; fm, fraction metabolized; GMFE, geometric mean fold error; Indmax, maximal fold induction; IVIVE, in vitro-in vivo extrapolation; kdef,, rate of enzyme degradation); MDZ, midazolam; PBPK, physiologically based pharmacokinetic; PHB, phenobarbital; PHY, phenytoin; RMSE, root mean square error.

821 822 Almond et al. implemented in the Simcyp simulator (Almond et al., 2009), where in In Vitro Data Analysis. Data for mRNA and activity were plotted as fold vitro data for a new drug is calibrated against in vitro data for a increase over vehicle control versus the concentration of the inducer. Curve fitting compound with known induction potential as a positive control (e.g., was carried out on data from each hepatocyte donor individually and then mean rifampicin). The effect of the unknown drug in vivo can then be Indmax (maximum fold induction, Emax +1)andIndC50 (the concentration that predicted based on the difference in potency of the new compound yields half of the Emax) were calculated. Both three-parameter (assuming the Hill exponent is equal to 1) and four-parameter sigmoidal models were fitted to the in compared with rifampicin and the plasma levels achieved after dosing in vitro data (mRNA and activity) using GraphPad Prism (version 5). Parameters vivo in humans. derived from these two models were not significantly different; therefore, the values Numerous independent publications have described the dynamic from the simpler model (three-parameter fit) were used for subsequent analysis induction model within the Simcyp simulator as being successfully (Table 2). It should be noted that Indmax is the maximum fold induction and as such applied for the quantitative prediction of CYP3A4 induction (Gandelman is not corrected for baseline (i.e., is equal to Emax + 1). Values are entered as Indmax, et al., 2011; Xu et al., 2011; Dhuria et al., 2013; Greupink et al., 2013; and this correction is handled within the Simcyp Simulator (see eq. 3 and eq. 4). Einolf et al., 2014); however, we have noted cases of under prediction in Clinical Pharmacokinetic Data for the Assessment of Prediction the interaction between rifampicin and orally dosed midazolam (MDZ). Accuracy. PubMed and The Metabolism & Transport Drug Interaction Database The success of IVIVE approaches to predict enzyme induction depends (http://www.druginteractioninfo.org/applications/metabolism-transport-drug- on a number of factors, including the type (induction of mRNA vs. interaction-database/) were used to identify relevant clinical DDI data arising from induction in white subjects. DDI studies involving the CYP3A4 inducer enzyme activity) and quality of in vitro data, the methods used to analyze rifampicin with the CYP3A4 substrates MDZ, alfentanil, alprazolam, nifedipine, the in vitro data, the approach taken to scale the in vitro data to the in vivo simvastatin, and zolpidem were identified. In vivo studies were included in the situation (use of calibrators for in vitro and in vivo induction data), as well analysis if the report included sufficient details of the dosage regimen to allow Downloaded from as variability in the data from the clinical studies against which the accurate replication of the trial design as well as the fold-change in the plasma area predictions are compared. In this study, a systematic evaluation of an under the curve (AUC). Where concentration-time profiles were available in the IVIVE-PBPK approach to predict the interactions between rifampicin references, these data were digitized (GetData software http://getdata-graph-digitizer. and CYP3A substrates with ranging fm3A4 (the fractional contribution of com/index.php) and compared with the predicted concentration-time profiles. Fifteen clinical studies describing the disposition of MDZ, before and after CYP3A4 to systemic clearance) and FG (the fraction escaping gut wall multiple dosing with rifampicin, were identified. Of these studies, one study was

metabolism) was carried out. Model refinements to improve the pre- dmd.aspetjournals.org diction accuracy were investigated and then applied to predict the excluded because the data were from subjects of mixed ethnicity, only one-third of whom were white (Adams et al., 2005), and the data were not stratified in a way interaction with other independent CYP3A inducers, using rich in vitro that allowed simulation of the different ethnic groups independently. Similarly, data generated using multiple human hepatocyte donors within a single data from the i.v. MDZ arm from the study by Floyd et al. (2003) could not be laboratory and standardized protocols. used, although data from female white subjects after an oral dose were described and hence were included (Floyd et al., 2003). In the study by Eap et al. (2004), CYP3A4 induction was assessed with 7.5 and 0.075 mg of orally administered Materials and Methods MDZ on consecutive days. The magnitude of interaction with the 0.075-mg dose Materials. Cryopreserved human hepatocytes from four donors (Hu1206, was much lower than for the 7.5-mg dose (AUC ratio 2.3- vs. 19.1-fold), which at ASPET Journals on October 1, 2021 Hu1191, Hu1198, Hu4193), cryopreserved hepatocytes recovery media, and may be due to issues with the limit of detection after induction of CYP3A, and so AlamarBlue cell viability reagent were purchased from Life Technologies (Grand only the 7.5-mg data from this study have been included. All other studies were Island, NY). InvitroGro culture media (CP and HI) and Torpedo antibiotic mix included to assess the prediction accuracy of the model. Information describing were purchased from BioreclamationIVT (Baltimore, MD). QuantiGene Plex 2.0 the dosing regimen, the route of administration of MDZ, and the study size is assay kits (panel no. 11477) were purchased from Affymetrix (Santa Clara, CA). provided for the remaining studies in Table 3. Dimethyl sulfoxide, rifampicin, testosterone, phenobarbital (PHB), carbamaze- As none of the DDI studies identified above described the concentration-time pine (CBZ), and phenytoin (PHY) were purchased from Sigma-Aldrich (St. profiles of rifampicin, independent studies were identified for the performance Louis, MO). verification of rifampicin exposure. Of these, two studies were carried out in white Generation of Induction Parameters In Vitro. The changes in mRNA and healthy volunteers (Acocella et al., 1971; Drusano et al., 1986) and were used to enzyme activity were assessed in parallel in cryopreserved human hepatocytes evaluate the simulated concentration-time profiles of rifampicin. from four donors using previously described methods (Halladay et al., 2012). In brief, hepatocytes were incubated with varying concentrations of prototypical inducers (serial dilutions of inducers in dimethyl sulfoxide were prepared daily) TABLE 2 b before the assessment of activity (measurement of 6 -hydroxytestosterone In vitro induction parameters (Indmax and IndC50) for rifampicin, carbamazepine, formation measured by liquid chromatography-tandem mass spectroscopy (LC- phenobarbital, and phenytoin generated using mRNA and activity data MS/MS)] and mRNA levels (QuantiGene Plex 2.0 Affymettrix assay kit). Cell Data are shown as the mean and standard deviation from four human hepatocyte donors. toxicity and cell viability were monitored using lactate dehydrogenase leakage Where Indmax is the maximum fold induction (equal to Emax +1) and IndC50 is the concentration and AlamarBlue assays (Halladay et al., 2012). The concentration ranges that gives half maximal fold induction (analogous to EC50). (Table 1) were selected for each inducer based on previous published studies Activity mRNA with the aim of determining a robust Indmax and IndC50. Indmax IndC50 Indmax IndC50

fold mMfoldmM TABLE 1 Rifampicin Mean 22.7 0.30 29.9 0.71 Final concentrations of inducer in culture medium with 0.1% dimethyl sulfoxide (v/v) S.D. 7.8 0.10 7.0 0.35 Carbamazepine Mean 16.6 59.1 21.9 58.7 Inducer Concentrations S.D. 6.1 37.3 12.4 18.0 Phenobarbital Mean 21.1 473 44.2 743 m M S.D. 11.5 245 25.9 334 Rifampicin 0.03, 0.1, 0.3, 1, 3, 10, 30 Phenytoin Mean 13.6 51.3 24.5 123 Carbamazepine 1, 3, 10, 30, 100, 300, 1000 S.D. 3.7 29.4 7.6 120 Phenytoin 1, 3, 10, 30, 100, 300, 1000 Efavirenz Mean 13.5 4.9 18.1 8.4 Phenobarbital 10, 30, 100, 300, 1000, 2000, 3000 S.D. 4.2 1.7 5.4 5.1 Efavirenz 0.1, 0.3, 1, 2, 3, 10, 30 Nifedipine Mean 15.6 4.0 30.0 13.0 Nifedipine 0.03, 1, 2, 3, 10, 30, 100 S.D. 11.3 1.9 22.0 9.5 Prediction of CYP3A4 Induction Using PBPK 823

TABLE 3 Rifampicin-mediated drug-drug interaction studies reported in the literature Details of the exposure of CYP3A4 probe substrate in the before and after multiple dosing of rifampicin are shown. A negative dose stagger indicates that the victim was dosed before the perpetrator. Data are expressed as mean (coefficient of variation) with the exception of those given.

Study Rifampicin Victim (Dose) Dose Stagger n AUC AUCi 1/AUC Ratio i.v. administration of victim drugs ng/mL.h ng/mL.h Link et al., 2008 600 mg daily for 6 days MDZ (2 mg) 24 8 126 (84–269)a 82.4 (58.8–102)a 1.53 Kharasch et al., 2004 600 mg daily for 5 days MDZ (1 mg) 12c 10 28.4 (14.1) 14.8 (18.2) 1.92 Gorski et al., 2003 600 mg daily for 7 days MDZ (0.05 mg/kg) 12 52 118 (35.4) 52.8 (29.7) 2.23 Phimmasone and Kharasch, 2001 600 mg daily for 5 days MDZ (1 mg) 12 6 53.0 (26.4) 25.5 (19.0) 2.08 Szalat et al., 2007h 600 mg daily for 7 days MDZ (0.05 mg/kg) 12c 3 89.5 (18.3) 51.8 (13.5) 1.73 Kharasch et al., 1997 600 mg daily for 5 days MDZ (1 mg) 24 9 72.2d (n/a) 27.4d (n/a) 2.64 Holtbecker et al., 1996 600 mg daily for 7 days NIF (0.02 mg/kg) 0 6 38.1 (12.6) 26.7 (44.9) 1.43 Phimmasone and Kharasch, 2001 600 mg daily for 5 days ALF (0.015 mg/kg)) 13c 6 111 (52.1) 48.2 (19.7) 2.31 Kharasch et al., 2004 600 mg daily for 5 days ALF (0.015 mg/kg) 13c 10 64.8 (41.0) 24.3 (26.7) 2.67 Kharasch et al., 2011 600 mg daily for 6 days ALF (1 mg) 9c 6 59.0 (45.8) 21.0 (38.1) 2.81 Oral administration of victim drugs Backman et al., 1996 600 mg daily for 5 days MDZ (15 mg) 17 10 170 (23.4) 7.00 (40.6) 24.3 Backman et al., 1998 600 mg daily for 5 days MDZ (15 mg) 17 9 277 (78.0) 4.40 (68.2) 63.0 Chung et al., 2006 600 mg daily for 9 days MDZ (0.075 mg/kg) 22 18 49.0 (22–103)b 6.10 (125–371)b 8.03 c Eap et al., 2004 450 mg daily for 5 days MDZ (7.5 mg) 12 4 67.0 (44.8) 3.50 (5.70) 19.1 Downloaded from Gurley et al., 2006 300 mg twice a day for 7 days MDZ (8 mg) 0c 19 79.6 (29.1) 4.55 (49.2) 17.5 Gurley et al., 2008 300 mg twice a day for 7 days MDZ (8 mg) 2 16 107 (38.0) 6.46 (54.3) 16.6 Link et al., 2008 600 mg daily for 6 days MDZ (7.5mg) 24 8 103 (64–164)a 1.60 (1–7.2)a 64.3 Reitman et al., 2011 600 mg daily for 28 days MDZ (2 mg) 0 11 21.4 (33.6) 2.64 (45.3) 8.11 Kharasch et al., 2004 600 mg daily for 6 days MDZ (3 mg) 12c 10 20.9 (20.1) 1.10 (45.5) 19.0 Floyd et al., 2003 600 mg daily for 16 days MDZ (2 mg; 25 mg)e 012g 27.1 (n/a) 19.9 (n/a) 17.0 f Gorski et al., 2003 600 mg daily for 7 days MDZ (4 mg; 6 mg)e 12 52 35.8 (58.1) 3.70 (75.7) 25.6f c Schmider et al., 1999 450 mg daily for 4 days APZ (1 mg) 0 4 242 (31.3) 28.4 (23.9) 8.53 dmd.aspetjournals.org Chung et al., 2006 600 mg daily for 9 days SMV (40 mg) 22 18 29.0 (8–56)b 2.60 (0.8–26)b 11.2 Kyrklund et al., 2000 600 mg daily for 28 days SMV (40 mg) 0 10 17.3 (57.2) 2.40 (75.4) 7.21 Holtbecker et al., 1996 600 mg daily for 7 days NIF (20 mg) 0 6 230 (14.7) 18.8 (45.7) 12.2 Villikka et al., 1997b 600 mg daily for 5 days ZOL (20 mg) 17 8 1110 (36.9) 332 (56.4) 3.34 Kharasch et al., 2004 600 mg daily for 6 days ALF (0.06 mg/kg) 13c 10 103 (29.1) 4.70 (97.9) 21.9 Kharasch et al., 2011 600 mg daily for 5 days ALF (4 mg) 12c 6 108 (63.0) 6.40 (50.0) 16.9 Villikka et al., 1997a 600 mg daily for 5 days TZM (0.5 mg) 17 10 14.8 (21.4) 0.74 (59.8) 20.0

ALF, alfentanil; APZ, alprazolam; AUC, area under the curve; MDZ, midazolam; n/a, not available; NIF, nifedipine; ROA, root of administration; SMV, simvastatin; TZM, triazolam; ZOL, zolpidem. at ASPET Journals on October 1, 2021 aMedian and range. bGeometric mean and range. cAmbiguous. dCalculated assuming a body weight of 70 kg in both control and rifampicin arms of the study. eDose escalated for the RIF arm of the study to give equivalent MDZ concentrations as at baseline; fThe ratio of clearance due to dose escalation. gMidazolam AUC in the absence and presence of rifampicin were calculated from oral clearances provided for 12 white subjects (all women) of the 57 subjects studied in total. Data for white men were not provided. hCerebrotendinous xanthomatosis (CTX) patients.

Further literature searching was carried out to identify DDI investigations of victim, perpetrator, and levels of the active CYP3A4 enzyme (in the liver and other inducers (CBZ, PHY, and PHB) with the CYP3A substrates. A total of six gut) of each virtual subject. The effect of autoinduction was automatically studies were identified as summarized in Table 4. considered where the metabolism of the inducer is adequately defined (e.g., for PBPK Modeling. Populations of virtual human subjects were generated in the CBZ and PHY). Simcyp Population-based Simulator using a correlated Monte Carlo approach The differential equations describing the kinetics of victim and perpetrator (Jamei et al., 2009). A minimal (or lumped) PBPK model of distribution was drugs and enzyme dynamics for inhibition have been reported in full previously assumed for all compounds, where all organs other than the intestine and liver are (Rowland Yeo et al., 2010). Here, we focus only on the equations describing time combined (Rowland Yeo et al., 2010). variant intrinsic clearance of the victim in the presence of a perpetrator compound With the exception of data describing the induction efficacy and potency, in in the liver and gut (eq. 1 and eq. 2). The effect of competitive inhibition between vitro and pharmacokinetic data for substrates (Supplemental Table 1) and inducers substrate and perpetrator is described by the terms ILiv, Ipv, and Kiu-e, effects due to (Supplemental Table 2) were taken from the literature. In cases where data were enzyme induction or mechanism based inhibition are incorporated by time- available from more than one independent source for the same parameter, they dependent changes in the levels of active enzyme (ENZact,h) (eq. 3 and eq. 4). were combined to give weighted means based on the number of observations. Finally, the time-dependent value of intrinsic clearance is used in the differential With the exception of alfentanil and PHB, the compound files were taken from equations used to calculate the plasma concentration time profile and AUC those released in version 12 Release 2 of the Simcyp simulator with any (Rowland Yeo et al., 2010): subsequent updates highlighted (Supplemental Material). n m Vmax H 2 pe Enzact; H For each of the CYP3A substrates used in this study and for two of the four CL9int uH ¼ + + fu 2 ðI =ðKp =B : P ÞÞ p¼1 e¼1 þ B IN Liv IN IN þ ð =ð = : ÞÞ perpetrators (CBZ and PHY) sufficient in vitro metabolism information was Kmu 2 pe 1 fuB CLiv Kp B P Kiu 2 e available to simulate the contribution of different enzymes to the overall ð1Þ elimination of the compound. These data were used as input data to the Simcyp simulator and extrapolated to predict the intrinsic clearance in the whole liver and gut in both the absence and presence of an inducer. For the other n m VmaxG 2 pe Enz act; G compounds (rifampicin and PHB) assessed as drug-interaction perpetrators, CL9int uG ¼ + + ð2Þ p¼1 e¼1 fugut 2 IN Ipv CL was defined from in vivo estimates of systemic and oral clearance, K 2 þ þ fu C mu pe 1 K 2 gut pv respectively. The PBPK model was then used to simulate the time course of iu e 824 Almond et al.

TABLE 4 Summary of the clinical drug-drug interactions studies available within the literature The exposure of CYP3A4 probe substrate before and after multiple dosing of carbamazepine, phenytoin, and phenobarbital are shown Data are expressed as mean (coefficient of variation) with the exception of those where the individual data are provided (n = 2).

Dose Study Inducer Victim (Dose) n AUC (mg/L.h) AUCi (mg/L.h) 1/AUC Ratio Stagger Carbamazepine Ucar et al., 2004 CBZ (200 mg daily for 2 days; 300 mg twice SMV (80 mg) 0 12 0.089 (58.1) 0.023 (56.7) 3.93 a day for 12 days) Andreasen et al., 2007 CBZ (200 mg twice a day for 2 days; 400 mg QND (200 mg) 0a 10 5.12 (n/a) 1.98 (n/a) 2.57 twice a day for 14 days) Vlase et al., 2011 CBZ (400 mg daily for 16 days) ZOL (5 mg) 0a 18 0.235 (70.4) 0.102 (58.1) 2.31 Phenytoin Data et al., 1976 Dose adjusted to maintain plasma PHY conc QND (300 mg) 0a 2 12.6 (10.3, 15.0) 5.53 (4.24, 6.82) 2.28 10–20 mg/ml Phenobarbital Schellens et al., 1989 PHB (100 mg daily for 8 days) NIF (20 mg) 12a 15 0.343 (36.4) 0.135 (57.8) 2.54 Data et al., 1976 Dose adjusted to maintain plasma PHB conc. QND (300 mg) 0a 2 12.0 (9.33, 14.6) 4.10 (3.19, 5.00) 2.92 10–20 mg/ml

AUC, area under the curve; CBZ, carbamazepine; n/a, not available; PHB, phenobarbital; PHY, phenytoin; QND, quinidine; SMV, simvastatin; ZOL, zolpidem. Downloaded from aAmbiguous.

Where CL9intuH and CL9intuG are the unbound intrinsic clearance of substrate the change in metabolic ratio of 6b-hydroxycortisol to cortisol following multiple per whole liver and gut, respectively, in the presence of a perpetrator dosing of rifampicin (600 mg daily 14 days) (Tran et al., 1999) in conjunction with n m compound. + and + refer to the total number of pathways and enzymes concentration-time profile data (Acocella et al., 1971). These in vivo values for p¼1 e¼1 rifampicin are then used to calibrate the in vitro Indmax and IndC50 values of other dmd.aspetjournals.org involved in metabolism of the substrate, respectively. B:P and B:PIN are the blood inducers/test compounds against in vitro values of rifampicin from the same to plasma ratios of substrate and perpetrator fuB and fuB 2 IN are the unbound experiment as shown in eq. 5 and eq. 6: fraction in plasma to the blood to plasma ratio (fu / B:P) of the substrate and " ! # perpetrator, respectively. fugut-IN is the fraction unbound in the gut. VmaxH 2 pe and ðIndmax;test 2 1Þ Indmax;cal ¼ ðIndmax; RIF in vivo 2 1Þ þ 1 ð5Þ VmaxG 2 pe are the maximum metabolic reaction velocity of substrate (victim) per ðIndmax;RIF 2 1Þ whole liver and gut, respectively, Kmu 2 pe is the Michaelis constant (corrected for nonspecific binding); Enzact,H and Enzact,G is the amount of active enzyme, in this IndC50;test case CYP3A at any given time in the liver and gut, respectively; and ILiv and CLiv IndC50;cal ¼ IndC50; RIF in vivo ð6Þ IndC ; at ASPET Journals on October 1, 2021 are the time varying liver concentrations of inhibitor and substrate, respectively, 50 RIF KpandKpIN are the tissue to plasma partition coefficients of substrate and where cal, test, RIF, and RIF in vivo indicate whether the induction parameters are perpetrator. For compounds that show no competitive inhibition, the inhibition calibrated, the in vitro values of the test compound in a given assay, the in vitro fu 2 ðI =ðKp =B : P ÞÞ fu 2 I terms 1 þ B IN Liv IN IN and 1 þ gut IN pv for the values for rifampicin in a given assay and the reference in vivo values for Kiu 2 e Kiu 2 e rifampicin, respectively. liver and gut, respectively, equal to one and hence no inhibition is simulated: Design of Virtual Studies. To ensure that the characteristics of virtual subjects reflected those of the subjects studied in vivo, the age range, proportion of males dEnz act; H3A4 ¼ kdegH3A4 Enz 0;H3A4 and females, and the number of subjects were matched to the information on dt ! individual clinical trials presented in the publications. The simulations were also ð 2 Þð ð : ÞÞ Indmax 1 fuB 2 IN It;Liv KpIN B PIN matched to each published study in terms of dose, as well as the time, frequency, 1 þ 2 Enzact;H3A4 IndC50 þðfuB 2 IN It;Liv ðKpIN B : PIN ÞÞ duration, and route of dosing for both the perpetrator (in this case an inducer of !! CYP3A4) and victim (a substrate of CYP3A4). For each simulation, 10 separate ðkinactÞðfuB 2 IN It;Liv ðKpIN B : PIN ÞÞ k 2 þ trials were generated to assess variability across groups. Although some of the degH 3A4 þð ð : ÞÞ KI fuB 2 IN It;Liv KpIN B PIN victim drugs are metabolized by CYP3A5 in addition to CYP3A4, only CYP3A4 ð3Þ was considered as CYP3A5 induction is less well characterized and generally accepted as less significant compared with CYP3A4 (Williamson et al., 2011). dEnz act; G3A4 ¼ kdegG3A4 Enz0;G3A4 The accuracy of simulations that were run using in vivo reference values dt (Indmax =8;IndC50 = 0.32) for rifampicin itself and for calibration of other ðIndmax 2 1ÞIt; Gut 1 þ 2 Enzact;G3A4 inducers was assessed (model A). The simulated plasma rifampicin concentra- IndC50 þ It; Gut tions and the simulated fm3A4 and FG for the CYP3A4 substrates were verified ðk ÞI ; þ inact t Gut ; ð Þ against observed data. Parameters with uncertainty were identified, and sensitivity kdegG 2 3A4 þ 4 KI It; Gut analysis was then used to assess which parameters were most likely to contribute to misprediction. Based on these analyses, simulations were repeated using where Enz ; 2 and Enz ; 2 are the amounts of active CYP3A4 at a given act H 3A4 act G 3A4 different assumptions regarding the Ind and IndC values entered into the time in the liver (eq. 3) and gut (eq. 4), respectively, Enz and Enz is the max 50 0,H-3A4 0,G-3A4 model as follows: basal amount of CYP3A in the liver and gut, respectively, and (Enzact(t) = E0 at t =0). Indmax is the maximal fold induction expressed as a fold over vehicle control. Indmax = • Use of a higher Indmax in the gut (16) than in the liver (8) but the same Emax +1.IndC50 is the concentration that supports half-maximal induction; KI is the IndC50 in both sites of interaction (0.32) (model B) concentration of mechanism-based inhibitor associated with half-maximal inactiva- • Use of a higher Indmax in both the gut and liver (16) but the same IndC50 tion rate of the enzyme (kinact(1/h)); It is the perpetrator concentration at time t in either (0.32) (model C) the liver or the gut. • Use of Indmax and IndC50 values derived from in vitro data without Derivation of Reference In Vivo Induction Parameters and their Role in calibration (mRNA) (model D)

Calibration. In vivo reference values describing the concentration–induction • Use of Indmax and IndC50 values derived from in vitro data without response of rifampicin (Indmax and IndC50) were derived using a study describing calibration (activity) (model E) Prediction of CYP3A4 Induction Using PBPK 825

• Use of a higher Indmax in both the gut and liver (12) but the same IndC50 value for mRNA versus activity with fold difference between the two (0.32) (model F) ranging from 0.6- to 1.3-fold (Fig. 2C). • Use of a higher Ind in both the gut and liver (20) but the same IndC max 50 Simulations Using the Rifampicin Base Model (Model A; Indmax (0.32) (model G) 8, IndC50 0.32 mM). The data in Table 3 show that both the magnitude of

After the best model was selected, the refined value of Indmax was used to interaction and the variability between studies were higher when MDZ was calibrate the in vitro data of the other inducers and the overall prediction accuracy administered orally compared with i.v. administration (median, 17.5-fold; for these inducers assessed. A schematic representation of this investigation is range, 8.0- to 64 vs. 2.0-fold (1.5- to 2.6-fold) reduction in MDZ AUC). shown in Fig. 1. Simulations of the clinical studies describing the changes in exposure Assessment of Prediction Accuracy. The ratio of the AUC of the substrate in of i.v. administered MDZ, before and after multiple dosing with the absence and the presence of an inhibitor of substrate metabolism (AUC(0–‘),inhibitor/ rifampicin, using the default settings in the rifampicin compound file AUC –‘ ) and the percent of change in the AUC are commonly used as a (0 ),control (model A) were in good agreement with the observed data (GMFE 1.21). basis for prediction of metabolic DDIs. In the presence of an enzyme inducer, this ratio gives values , 1; to aid interpretation, in this manuscript the reciprocal of Simulated studies describing the effect of multiple dosing of rifampicin on orally administered MDZ exposure predicted a higher fold this ratio has been used (AUC(0-‘),control /AUC(0-‘),induced) to yield ratios . 1 in the presence of an enzyme inducer. However, data were plotted both ways to show the change in exposure compared with i.v. administered MDZ (median fold comparison. The means of AUC ratios from the 10 simulated trials were compared change 6.5- versus 1.7-fold), in line with the observed situation (median against the mean AUC ratio from each in vivo study (fold error). In addition the fold change 18.1- versus 2.0-fold); however, the magnitude of in- acceptance criteria proposed by Guest et al. (2011) was also used. This is a more teraction was underpredicted for all clinical studies (GMFE 2.12), sensitive measure of concordance in reflecting absolute changes in AUC, despite the wide variability between the clinical studies (range of 1/AUC Downloaded from especially when these are small (Guest et al., 2011). Equation 7 and eq. 8 were ratios 8.0–64.3). used to calculate the geometric mean-fold error (GMFE) and the root-mean square Plotting the data as a percent change from control indicates excellent error, which were used to assess the precision of the predictions: prediction accuracy (Fig. 3, E and F), with all predictions for oral MDZ dosing falling between 0.8- and 1.25-fold of the observed value; mean log predicted DDI observed DDI however, comparison of these data as an interaction ratio or the GMFE ¼ 10 ð7Þ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi –

reciprocal of the ratio show that this is not the case (Fig. 3, A D). dmd.aspetjournals.org +ðpredicted DDI 2 observed ​DDIÞ2 Verification of Simulated Systemic Rifampicin Concentrations RMSE ¼ ð8Þ number of predictions and Victim Drug Properties. Although rifampicin concentrations were not reported for any of the clinical DDI studies (Table 3), independent studies describing the pharmacokinetics of rifampicin in healthy white volunteers were identified and simulated. The predicted plasma Results concentration-time profiles for rifampicin after multiple dose adminis- Induction Parameters Determined In Vitro. The in vitro param- tration were in reasonable agreement with the observed (Supplemental at ASPET Journals on October 1, 2021 eters (Indmax and IndC50) for the inducers investigated are shown in Fig. Fig. 1). Owing to a lack of information describing the metabolism of 2 and Table 2. Comparison of the data derived from assessment of rifampicin, the model used for rifampicin cannot account for auto- mRNA versus activity showed that efficacy was higher (1.3- to 2.0-fold induction, and hence the concentrations of the initial doses were under higher Indmax values; Fig. 2A), but potency (IndC50) was generally predicted. This was deemed acceptable as here the focus was on lower (1.0- to 3.3-fold; Fig. 2B) when measured by changes in mRNA predictions after multiple doses of rifampicin. Simulated key properties levels compared with changes in activity. When the ratio of Indmax to (fm and FG) were also in reasonable agreement with those that we IndC50 was compared, no systematic trend was seen for a higher or lower observed. (Supplemental Fig. 2)

Fig. 1. A schematic representation of the investigation that was split into two main stages: step 1 was the evaluation of different rifampicin models before the best model (model C) was evaluated for calibration of mRNA and activity data for the other inducers. This result was compared with no calibration and calibration with the original base model (model A). 826 Almond et al. Downloaded from dmd.aspetjournals.org

Fig. 2. A comparison of Indmax (A, diamonds), IndC50 (B, squares), and the ratio of Indmax:IndC50 (C, circles) derived from mRNA and activity data in four human hepatocyte donors (Hu1206, Hu1191, Hu1198, and Hu4193) after incubation with six in vitro inducers of CYP3A (rifampicin, CBZ, PHB, phenytoin, efavirenz, and at ASPET Journals on October 1, 2021 nifedipine). Data are plotted as mean 6 standard deviation. The lines of unity (unbroken line), 0.8- to 1.25-fold (dotted line) and 0.5 to 2-fold (dashed line) are shown.

Simulations Using the Modified Rifampicin Models (Models B– 8 gave the lowest prediction accuracy (GMFE 1.7 and 33.3% cases E). The accuracy of the rifampicin DDI simulations before and after within the acceptance limits). Although predictions with uncalibrated modifications to the base model are described in Table 5 and plotted in activity data and activity data calibrated against an Indmax of 16 were Fig. 4. All the alternative models performed better than the base model reasonably consistent, uncalibrated activity data gave the higher pre- but to varying degrees. Model B (where Indmax for the gut was increased diction accuracy (GMFE 1.39 vs. 1.49 and % within acceptance limits to 16, but Indmax in the liver was kept at 8) improved the predictions 83.3% vs. 66.7%). (GMFE 1.77 versus 2.12) but not as much as model C (where Indmax was changed to 16 in both the liver and the gut; GMFE 1.48 versus 2.12). The highest proportion of predictions to fall within the stringent criteria Discussion (Guest et al., 2011) was with models C and F (79.3% of cases). In this Changes to regulatory guidance from the FDA have promoted a study, the uncalibrated assessment of induction using mRNA and switch in emphasis from measuring activity to mRNA for assessment of activity yielded predictions that were also more accurate than the base induction in vitro (EMA, 2012; FDA, 2012). Although mRNA has model A (1.61 and 1.53 GMFE and 65.5% and 65.5% within acceptance utility as a sensitive marker, especially in cases where a compound is limits for model D activity and model E mRNA, respectively). both an inducer and a mechanism-based inhibitor (Fahmi et al., 2009), Additional tested Indmax values of 12 (model F) and 20 (model G) also the magnitude of mRNA changes can be several-fold greater than for improved the model compared with the base model (1.63 and 1.51 vs. activity for CYP3A4 (Luo et al., 2002; Martin et al., 2008; McGinnity 2.12, respectively). et al., 2009). In this investigation, full concentration-induction relation- Predicted DDIs with Inducers Other than Rifampicin. Simula- ships for mRNA and activity were derived in the same incubation for tions for inducers other than rifampicin (CBZ, PHY, and PHB) were run five clinical inducers (rifampicin, CBZ, PHY, PHB, and efavirenz) and using mRNA and activity data before and after calibration against one drug that induces in vitro but not in vivo (nifedipine). rifampicin. All calibration was performed using both the original (8) and When using in vitro data to quantitatively predict a clinical DDI, one refined (16)Indmax for rifampicin. Comparisons of predicted and question to consider is what defines a successful prediction. This may be observed fold changes in AUC (1/AUC ratio) are shown in Fig. 5. different early in a drug discovery project when a prediction accuracy of When mRNA data were used to predict the magnitude of induction, the 2- to 3-fold may be acceptable for ranking/compound selection, whereas prediction accuracy was similar for uncalibrated, calibrated with an in the later stages of clinical development, where the goals are to define Indmax of 8 and calibrated with an Indmax of 16, but GMFE was lowest DDI liability and support clinical trial design, a greater degree of (marginally) when the data were calibrated against an Indmax of 16 accuracy is required, perhaps within 1.25-fold. We have based our (Table 6). When activity data were used, calibration against an Indmax of assessments of prediction accuracy on calculated values of GMFE and Prediction of CYP3A4 Induction Using PBPK 827 Downloaded from dmd.aspetjournals.org at ASPET Journals on October 1, 2021

Fig. 3. A comparison of the observed and predicted (model A) magnitude of induction for the AUC (A, C, E) and Cmax (B, D, F) of midazolam (circles), nifedipine (squares), alfentanil (diamonds), triazolam (plus sign), alprazolam (cross), zolpidem (dash), and simvastatin (triangles) after their i.v. (open) and oral (closed) administration after multiple doses of rifampicin. Data are plotted as the interaction ratio (A, B), the reciprocal of the interaction ratio (C, D) and as percentage reduction in AUC (E) and Cmax (F). The lines of unity (unbroken line), 0.8- to 1.25-fold (dotted line), 0.5- to 2.0-fold (dashed line), and more cautious limits as suggested by Guest et al. (2011) (broken and dotted line) are shown. Solid vertical and horizontal lines mark 0.8- (A, B) and 1.25- (C, D) fold to show the clinical cutoffs for a DDI. 828 Almond et al.

TABLE 5 Summary of the accuracy of DDI predictions using different rifampicin models (A–G)

Model A Model B Model C Model D Model E Model F Model G

Observed a Indmax 8, Indmax 8 liver, 16 gut, Indmax 16 Indmax 22.7 Indmax 29.9 Indmax 12 Indmax 20 a IndC50 0.32 IndC50 0.32 IndC50 0.32 IndC50 0.30 IndC50 0.71 IndC50 0.32 IndC50 0.32 Rifampicin, i.v. MDZ Geometric mean fold induction 1.99 1.71 1.72 2.04 2.13 2.11 1.96 2.15 GMFE 1.21 1.21 1.16 1.18 1.18 1.16 1.20 RMSE 0.51 0.47 0.36 0.38 0.38 0.38 0.40 % Within acceptance limitsb 83.3 100 100 83.3 83.3 100 83.3 Rifampicin, oral MDZ Geometric mean fold induction 18.1 6.47 9.69 17.1 29.3 26.6 10.9 23.7 GMFE 3.26 2.21 1.70 1.96 1.85 1.99 1.75 RMSE 27.0 24.8 21.6 21.5 20.7 24.1 20.6 % Within acceptance limitsb 27.3 45.5 72.7 36.4 36.4 63.6 63.6 Rifampicin, all MDZ (i.v. and oral) Geometric mean fold induction n/a GMFE 2.30 1.79 1.48 1.64 1.58 1.65 1.53 RMSE 21.7 20.0 17.4 17.3 16.7 19.4 16.5 % Within acceptance limitsb 47.1 58.8 82.4 52.9 52.9 76.5 70.6 Rifampicin, all victims (i.v.) Downloaded from Geometric mean fold induction n/a GMFE 1.24 1.24 1.15 1.16 1.13 1.16 1.17 RMSE 0.53 0.56 0.35 0.36 0.35 0.42 0.37 % Within acceptance limitsb 90.0 100 100 90.0 90.0 100 90.0 Rifampicin, all victims (oral) Geometric mean fold induction n/a GMFE 2.81 2.12 1.69 1.91 1.80 1.96 1.72 RMSE 21.5 19.9 17.7 20.5 19.2 19.1 18.9 dmd.aspetjournals.org % Within acceptance limitsb 26.3 26.3 68.4 52.6 52.6 73.7 68.4 Rifampicin, all victim drugs Geometric mean fold induction n/a GMFE 2.12 1.77 1.48 1.61 1.53 1.63 1.51 RMSE 17.4 16.1 14.4 16.5 15.5 15.5 15.3 % Within acceptance limitsb 48.3 51.7 79.3 65.5 65.5 79.3 75.9

GMFE, geometric mean fold error; n/a, not applicable; RMSE, root mean square error. a

Default rifampicin induction parameters (V12). Geometric mean fold induction for observed data were calculated in a meta-analysis using published methodology (Einolf, 2007; Cubitt et al., 2011; at ASPET Journals on October 1, 2021 Ghobadi et al., 2011; Barter et al., 2013; Supplemental Table 3). bAcceptance limits proposed by Guest et al. (2011). root mean square error for consistency with the literature in this area and rifampicin were derived from two separate studies, one describing the have also used more conservative acceptance limits (Guest et al., 2011). change in metabolic ratio of an endogenous substrate (cortisol) during Although often overlooked, the variability observed in the clinic rifampicin dosing (Tran et al., 1999) and the other the kinetics of between studies with the same compounds can also impact the ability rifampicin (Acocella et al., 1971). Because of the variability in rifam- of an IVIVE approach to successfully predict the magnitude of DDI in all picin pharmacokinetics, it is possible that the plasma concentrations in individual studies. Because of variability in in vitro induction experi- the two studies were different. Second, monitoring the metabolic ratio of ments, the use of in vivo reference values for a calibrator compound have an endogenous compound may not provide information on changes in been recommended for the translation of in vitro induction effects to the gut metabolism as it is analogous to using a ratio calculated after i.v. in vivo situation (Almond et al., 2009). This approach assumes that the administration. The accuracy of DDI prediction was assessed using a efficacy and potency of an inducer relative to the calibrator is the same in range of models where Indmax was increased only in the gut or in both vitro as in vivo. Clearly, if a calibration approach is used, the values used gut and liver, respectively. Although all models improved predictions, for the in vivo calibration will also impact on whether the DDI model C gave the most accurate predictions when MDZ and other victim predictions are successful. In this study, the accuracy of these in vivo drugs (with ranging hepatic and gut extraction) were given orally. reference values was assessed initially by analyzing the accuracy of DDI Recent investigations have also reported a need for higher Indmax for prediction with rifampicin before assessment of their performance in rifampicin of 12.5- (Xia et al., 2014), 14.6-, (Baneyx et al., 2014) and calibration for other inducers. 11.5-fold (Wagner et al., 2015). These values are not dissimilar to the The original base model for rifampicin (model A) used Indmax values value of 16-fold used here and when used in our model gave comparable of 8 in the gut and liver and had a higher prediction accuracy for the DDI prediction accuracy. The current study is the only one to have used the between oral rifampicin and i.v. MDZ than when MDZ was also dosed refined rifampicin Indmax to calibrate in vitro induction data for other orally. This result could be explained by inaccuracy in the extent of inducers and demonstrate application of this strategy for these change in the first pass extraction in the liver (EH)and/orgut(EG)on compounds within a mechanistic dynamic PBPK model. In addition to dosing with rifampicin or may reflect that with a relatively high the in vivo reference Indmax and IndC50 values for rifampicin, other extraction compound, such as MDZ, there is a limit on the extent of factors that could potentially explain the underprediction of DDI when induction that can be observed when the compound is dosed i.v. as rifampicin was administered with oral victim drugs were investigated hepatic CL becomes limited by hepatic blood flow. but not shown to have a MDZ significant impact. These included Several factors were considered as explanations for the under consideration of: 1) induction of UGT1A4-mediated metabolism, 2) a prediction of the DDI between rifampicin and orally administered protein-binding displacement interaction leading to a transient increase victim drugs. First, the reference values used to predict in vivo effects of in the fu of the victim drug and increased first-pass clearance, 3) the Prediction of CYP3A4 Induction Using PBPK 829 Downloaded from dmd.aspetjournals.org at ASPET Journals on October 1, 2021

Fig. 4. A comparison of the observed and predicted magnitude of induction on the AUC (A, C, E, G, I) and Cmax (B, D, F, H, I) of midazolam (circles), nifedipine (squares), alfentanil (diamonds), triazolam (plus sign), alprazolam (cross) and simvastatin (triangles) after their i.v. (open) and oral (closed) administration after multiple doses of rifampicin (600 mg daily). Predictions were made with models A (A, B), model B (C, D), model C (E, F), model D (G, H), and model E (I, J). Data are plotted as the reciprocal of the interaction ratio. The lines of unity (unbroken line), 0.8- to 1.25-fold (dotted line), 0.5- to 2.0-fold (dashed line), and more cautious limits as suggested by Guest et al. (2011) (broken and dotted line) are shown. Solid vertical and horizontal lines mark 0.8-fold (A, B) and 1.25-fold (C, D) to show the clinical cutoffs for a DDI. 830 Almond et al. Downloaded from dmd.aspetjournals.org

Fig. 5. Comparison of the observed and predicted magnitude of change in 1/AUC ratio of orally administered CYP3A4 substrates after administration of multiple doses of CBZ (squares), phenytoin (circles), and PHB (triangles). Predictions are made using in vitro mRNA (A–C) and activity (D–F) data that are uncalibrated (A, D), calibrated at ASPET Journals on October 1, 2021 using Indmax 8, IndC50 0.32 (B, E), and calibrated using Indmax 16, IndC50 0.32. Data are plotted as the reciprocal of the interaction. The lines of unity (unbroken line), 0.8- to 1.25-fold (dotted line), 0.5- to 2.0-fold (dashed line), and more cautious limits as suggested by Guest et al. (2011) (broken and dotted line) are shown. Solid vertical and horizontal lines mark 0.8- (A, B) and 1.25- (C, D) fold to show the clinical cut offs for a DDI.

sensitivity to different values of first-order rate constants (kdegH and perpetrator and victim drugs across regions in the gut was assumed to be kdegG) that describe endogenous turnover of active enzyme in the liver uniform and not limited by solubility. Further research is required to and gut (Yang et al., 2008), 4) the impact of disparate regional fully elucidate the cause of under prediction before a mechanistic absorption between the victim and perpetrator along the gastrointestinal derivation of in vivo Indmax is possible. tract, and 5) sensitivity to different assumptions of the fraction unbound Despite the variability in in vitro assays of cytochrome induction, of drug within enterocytes (fugut) that is used to calculate both the FG direct entry of mRNA (model D) and activity (model E) data yielded (Yang et al., 2007) and the operational concentration of a perpetrator in DDI predictions that were in reasonable agreement with the observed the gut (Rowland Yeo et al., 2010), in line with recommendations (Zhao (GMFE 1.61 and 1.53 for models D and E, respectively, compared with et al., 2012). In the latter investigation, changing rifampicin fugut from 2.12 for the best model). The ratio of Indmax/IndC50 for mRNA and the 0.19 to 1 gave higher simulated unbound portal vein concentrations, but activity in this study were similar, with a tendency for the mRNA data to in both cases the free concentrations exceeded the IndC50 for rifampicin have both a higher Indmax and IndC50. Although this approach was (0.32 mM) across most of the dosing interval; hence, little effect on successful here, a drawback of this approach is that Indmax and IndC50 predictions was observed. In this investigation, absorption of both are influenced by interindividual variability across different donors. In a

TABLE 6 Summary of the predication accuracy of drug-drug interactions (1/AUC ratio) for the inducers

Six studies (carbamazepine, phenytoin, and phenobarbital) using mRNA and activity data, uncalibrated, calibrated against an Indmax =8 and calibrated against Indmax=16.

Activity mRNA Activity mRNA Activity mRNA

Uncalibrated Uncalibrated Calibrated (8) Calibrated (8) Calibrated (16) Calibrated (16) GMFE 1.39 1.44 1.68 1.46 1.49 1.35 RMSE 2.19 3.40 1.09 0.97 1.30 2.98 % Within acceptance limitsa 83.3 83.3 33.3 83.3 66.7 83.3

GMFE, geometric mean fold error; RMSE, root mean square error. aAcceptance limits proposed by Guest et al. (2011). Prediction of CYP3A4 Induction Using PBPK 831 previous study from this laboratory using different donors, the difference Backman JT, Olkkola KT, and Neuvonen PJ (1996) Rifampin drastically reduces plasma con- centrations and effects of oral midazolam. Clin Pharmacol Ther 59:7–13. in Indmax between the two experimental endpoints was approximately10- Baneyx G, Parrott N, Meille C, Iliadis A, and Lavé T (2014) Physiologically based pharmacoki- fold (Halladay et al., 2012), whereas other investigators have come to netic modeling of CYP3A4 induction by rifampicin in human: influence of time between sub- strate and inducer administration. Eur J Pharm Sci 56:1–15. similar conclusions (McGinnity et al., 2009). Considerable effort is Barter ZE, Tucker GT, and Rowland-Yeo K (2013) Differences in cytochrome p450-mediated required to fully characterize each hepatocyte lot by the generation of full pharmacokinetics between chinese and caucasian populations predicted by mechanistic physi- Ind and IndC data for a number of prototypical inducers to ensure ologically based pharmacokinetic modelling. Clin Pharmacokinet 52:1085–1100. max 50 Chung E, Nafziger AN, Kazierad DJ, and Bertino JS, Jr (2006) Comparison of midazolam and that an uncalibrated approach will be successful for a novel compound. simvastatin as cytochrome P450 3A probes. Clin Pharmacol Ther 79:350–361. Use of empirical scalars (d-factor) has been proposed for mechanistic Cubitt HE, Yeo KR, Howgate EM, Rostami-Hodjegan A, and Barter ZE (2011) Sources of in- terindividual variability in IVIVE of clearance: an investigation into the prediction of benzodi- static models (Fahmi et al., 2008; Fahmi et al., 2009) to account for any azepine clearance using a mechanistic population-based pharmacokinetic model. Xenobiotica 41: systematic deviation between in vitro and in vivo. In some ways, the 623–638. Data JL, Wilkinson GR, and Nies AS (1976) Interaction of quinidine with anticonvulsant drugs. N subsequent scrutiny and correction of in vitro data against a data set (from Engl J Med 294:699–702. the same characterized in vitro system) before entry into models is Dhuria S, Einolf H, Mangold J, Sen S, Gu H, Wang L, and Cameron S (2013) Time-dependent inhibition and induction of human cytochrome P4503A4/5 by an oral IAP antagonist, LCL161, analogous to the d-factor approach but is within a dynamic model. in vitro and in vivo in healthy subjects. J Clin Pharmacol 53:642–653. The advantages of a calibration-based approach are that it controls for Drusano GL, Townsend RJ, Walsh TJ, Forrest A, Antal EJ, and Standiford HC (1986) Steady-state serum pharmacokinetics of novobiocin and rifampin alone and in combination. Antimicrob the wide variability that is observed in vitro (such as that noted across Agents Chemother 30:42–45. independent laboratories) (Einolf et al., 2014); it allows the prospective Eap CB, Buclin T, Cucchia G, Zullino D, Hustert E, Bleiber G, Golay KP, Aubert AC, Baumann P, and Telenti A, et al. (2004) Oral administration of a low dose of midazolam (75 microg) as an in prediction of DDIs, with less emphasis for full characterization of the in vivo probe for CYP3A activity. Eur J Clin Pharmacol 60:237–246. vitro system; and provides flexibility in whether data from mRNA or Einolf HJ (2007) Comparison of different approaches to predict metabolic drug-drug interactions. Downloaded from activity are used. In this investigation, we evaluated the existing (Ind Xenobiotica 37:1257–1294. max Einolf HJ, Chen L, Fahmi OA, Gibson CR, Obach RS, Shebley M, Silva J, Sinz MW, Unadkat JD, 8) and refined (Indmax 16) the rifampicin model for the calibration of the and Zhang L, et al. (2014) Evaluation of various static and dynamic modeling methods to predict prototypical inducers CBZ, PHY, and PHB and showed calibration with clinical CYP3A induction using in vitro CYP3A4 mRNA induction data. Clin Pharmacol Ther 95:179–188. the refined model performed reasonably well. European Medicines Agency (EMA) (2012) Guideline on the Investigation of Drug Interactions. In summary, we have provided a systematic evaluation of the Committee for Human Medicinal Products, London. Fahmi OA, Hurst S, Plowchalk D, Cook J, Guo F, Youdim K, Dickins M, Phipps A, Darekar A, prediction of DDIs mediated by CYP3A4 induction using a mechanistic and Hyland R, et al. (2009) Comparison of different algorithms for predicting clinical drug-drug dmd.aspetjournals.org dynamic model. Use of a range of CYP3A substrates with i.v. and oral interactions, based on the use of CYP3A4 in vitro data: predictions of compounds as precipitants of interaction. Drug Metab Dispos 37:1658–1666. administration allowed correction of underprediction, which was then Fahmi OA, Maurer TS, Kish M, Cardenas E, Boldt S, and Nettleton D (2008) A combined model verified with independent predictions for inducers other than rifampicin. for predicting CYP3A4 clinical net drug-drug interaction based on CYP3A4 inhibition, in- Using a comprehensive data set generated using four hepatocyte donors, activation, and induction determined in vitro. Drug Metab Dispos 36:1698–1708. Floyd MD, Gervasini G, Masica AL, Mayo G, George AL, Jr, Bhat K, Kim RB, and Wilkinson GR we were able to compare the predictions made with mRNA and activity (2003) Genotype-phenotype associations for common CYP3A4 and CYP3A5 variants in the data, both calibrated and uncalibrated. Although we believe that basal and induced metabolism of midazolam in European- and African-American men and women. Pharmacogenetics 13:595–606. calibration with robust in vivo reference values is helpful to combat Food and Drug Administration (FDA) (2012) Guidance for Industry: Drug Interactions Studies— donor and laboratory variability, uncalibrated data also performed Study Design, Data Analysis, Implications for Dosing, and Labeling Recommendations. U.S. at ASPET Journals on October 1, 2021 Department of Health and Human Services, FDA, Silver Spring, MD. reasonably well with our data set based on prototypical inducers. Use Gandelman K, Zhu T, Fahmi OA, Glue P, Lian K, Obach RS, and Damle B (2011) Unexpected of an uncalibrated approach requires full characterization of the in vitro effect of rifampin on the pharmacokinetics of linezolid: in silico and in vitro approaches to explain its mechanism. JClinPharmacol51:229–236. induction seen within donors and laboratories with prototypical inducers Ghobadi C, Johnson TN, Aarabi M, Almond LM, Allabi AC, Rowland-Yeo K, Jamei M, to give an understanding of how that particular system extrapolates to the and Rostami-Hodjegan A (2011) Application of a systems approach to the bottom-up assessment of pharmacokinetics in obese patients: expected variations in clearance. Clin Pharmacokinet 50: in vivo situation. 809–822. Gorski JC, Vannaprasaht S, Hamman MA, Ambrosius WT, Bruce MA, Haehner-Daniels B, and Hall SD (2003) The effect of age, sex, and rifampin administration on intestinal and hepatic Acknowledgments cytochrome P450 3A activity. Clin Pharmacol Ther 74:275–287. The authors acknowledge Chenghong Zhang for technical assistance and Greupink R, Schreurs M, Benne MS, Huisman MT, and Russel FG (2013) Semi-mechanistic Jessica Waite and Eleanor Savill for assistance with manuscript preparation. physiologically-based pharmacokinetic modeling of clinical glibenclamide pharmacokinetics and drug-drug-interactions. Eur J Pharm Sci 49:819–828. Guest EJ, Aarons L, Houston JB, Rostami-Hodjegan A, and Galetin A (2011) Critique of the two- Authorship Contributions fold measure of prediction success for ratios: application for the assessment of drug-drug in- teractions. Drug Metab Dispos 39:170–173. Participated in research design: Almond, Gardner, Wong, Tay, Rowland- Gurley B, Hubbard MA, Williams DK, Thaden J, Tong Y, Gentry WB, Breen P, Carrier DJ, Yeo, Rostami-Hodjegan, Jamei, Kenny. and Cheboyina S (2006) Assessing the clinical significance of botanical supplementation on human cytochrome P450 3A activity: comparison of a milk thistle and black cohosh product to Conducted experiments: Almond, Mukadam, Wong, Tay. rifampin and clarithromycin. JClinPharmacol46:201–213. Contributed new reagents or analytic tools: Wong, Tay. Gurley BJ, Swain A, Hubbard MA, Hartsfield F, Thaden J, Williams DK, Gentry WB, and Tong Y Performed data analysis: Almond, Mukadam, Okialda, Wong, Hatley, Tay, (2008) Supplementation with goldenseal (Hydrastis canadensis), but not kava kava (Piper methysticum), inhibits human CYP3A activity in vivo. Clin Pharmacol Ther 831 :6 –69. Kenny. Halladay JS, Wong S, Khojasteh SC, and Grepper S (2012) An ‘all-inclusive’ 96-well cytochrome Wrote or contributed to the writing of the manuscript: Almond, Gardner, P450 induction method: measuring enzyme activity, mRNA levels, protein levels, and cyto- Hatley, Rowland-Yeo, Jamei, Rostami-Hodjegan, Kenny. toxicity from one well using cryopreserved human hepatocytes. J Pharmacol Toxicol Methods 66:270–275. Holtbecker N, Fromm MF, Kroemer HK, Ohnhaus EE, and Heidemann H (1996) The nifedipine- References rifampin interaction: evidence for induction of gut wall metabolism. Drug Metab Dispos 24: 1121–1123. Acocella G, Pagani V, Marchetti M, Baroni GC, and Nicolis FB (1971) Kinetic studies on ri- Huang S-M, Abernethy DR, Wang Y, Zhao P, and Zineh I (2013) The utility of modeling and fampicin. I. Serum concentration analysis in subjects treated with different oral doses over a simulation in drug development and regulatory review. JPharmSci102:2912–2923. period of two weeks. Chemotherapy 16:356–370. Jamei M, Dickinson GL, and Rostami-Hodjegan A (2009) A framework for assessing inter- Adams M, Pieniaszek HJ, Jr, Gammaitoni AR, and Ahdieh H (2005) Oxymorphone extended individual variability in pharmacokinetics using virtual human populations and integrating release does not affect CYP2C9 or CYP3A4 metabolic pathways. J Clin Pharmacol 45:337–345. general knowledge of physical chemistry, biology, anatomy, physiology and genetics: a tale Almond LM, Yang J, Jamei M, Tucker GT, and Rostami-Hodjegan A (2009) Towards a quanti- of ‘bottom-up’ vs ‘top-down’ recognition of covariates. Drug Metab Pharmacokinet 24: tative framework for the prediction of DDIs arising from cytochrome P450 induction. Curr Drug 53–75. Metab 10:420–432. Kharasch ED, Russell M, Mautz D, Thummel KE, Kunze KL, Bowdle A, and Cox K (1997) The Andreasen AH, Brøsen K, and Damkier P (2007) A comparative pharmacokinetic study in healthy role of cytochrome P450 3A4 in alfentanil clearance: implications for interindividual variability volunteers of the effect of carbamazepine and oxcarbazepine on cyp3a4. Epilepsia 48:490–496. in disposition and perioperative drug interactions. Anesthesiology 87:36–50. Backman JT, Kivistö KT, Olkkola KT, and Neuvonen PJ (1998) The area under the plasma Kharasch ED, Vangveravong S, Buck N, London A, Kim T, Blood J, and Mach RH (2011) concentration-time curve for oral midazolam is 400-fold larger during treatment with itracona- Concurrent assessment of hepatic and intestinal cytochrome P450 3A activities using deuterated zole than with rifampicin. Eur J Clin Pharmacol 54:53–58. alfentanil. Clin Pharmacol Ther 89:562–570. 832 Almond et al.

Kharasch ED, Walker A, Hoffer C, and Sheffels P (2004) Intravenous and oral alfentanil as in vivo Tran JQ, Kovacs SJ, McIntosh TS, Davis HM, and Martin DE (1999) Morning spot and 24-hour probes for hepatic and first-pass cytochrome P450 3A activity: noninvasive assessment by use of urinary 6 beta-hydroxycortisol to cortisol ratios: intraindividual variability and correlation under pupillary miosis. Clin Pharmacol Ther 76:452–466. basal conditions and conditions of CYP 3A4 induction. J Clin Pharmacol 39:487–494. Kyrklund C, Backman JT, Kivistö KT, Neuvonen M, Laitila J, and Neuvonen PJ (2000) Rifampin Ucar M, Neuvonen M, Luurila H, Dahlqvist R, Neuvonen PJ, and Mjörndal T (2004) Carbama- greatly reduces plasma simvastatin and simvastatin acid concentrations. Clin Pharmacol Ther zepine markedly reduces serum concentrations of simvastatin and simvastatin acid. Eur J Clin 68:592–597. Pharmacol 59:879–882. Link B, Haschke M, Grignaschi N, Bodmer M, Aschmann YZ, Wenk M, and Krähenbühl S (2008) Villikka K, Kivistö KT, Backman JT, Olkkola KT, and Neuvonen PJ (1997a) Triazolam is in- Pharmacokinetics of intravenous and oral midazolam in plasma and saliva in humans: usefulness effective in patients taking rifampin. Clin Pharmacol Ther 61:8–14. of saliva as matrix for CYP3A phenotyping. Br J Clin Pharmacol 66:473–484. Villikka K, Kivistö KT, Luurila H, and Neuvonen PJ (1997b) Rifampin reduces plasma concen- Luo G, Cunningham M, Kim S, Burn T, Lin J, Sinz M, Hamilton G, Rizzo C, Jolley S, and Gilbert trations and effects of zolpidem. Clin Pharmacol Ther 62:629–634. D, et al. (2002) CYP3A4 induction by drugs: correlation between a pregnane X receptor reporter Vlase L, Popa A, Neag M, Muntean D, Bâldea I, and Leucut¸a SE (2011) Pharmacokinetic in- gene assay and CYP3A4 expression in human hepatocytes. Drug Metab Dispos 30:795–804. teraction between zolpidem and carbamazepine in healthy volunteers. JClinPharmacol51: Martin P, Riley R, Back DJ, and Owen A (2008) Comparison of the induction profile for drug 1233–1236. disposition proteins by typical nuclear receptor activators in human hepatic and intestinal cells. Wagner C, Pan Y, Hsu V, Sinha V, and Zhao P (2016) Predicting the Effect of CYP3A Inducers on Br J Pharmacol 153:805–819. the Pharmacokinetics of Substrate Drugs Using Physiologically Based Pharmacokinetic (PBPK) McGinnity DF, Zhang G, Kenny JR, Hamilton GA, Otmani S, Stams KR, Haney S, Brassil P, Modeling: An Analysis of PBPK Submissions to the US FDA. Clin Pharmacokinet 2016 55: Stresser DM, and Riley RJ (2009) Evaluation of multiple in vitro systems for assessment of 475–483. DOI 10.10007/s40262-015-0330-y. CYP3A4 induction in drug discovery: human hepatocytes, pregnane X receptor reporter gene, Williamson BL, Purkayastha S, Hunter CL, Nuwaysir L, Hill J, Easterwood L, and Hill J (2011) and Fa2N-4 and HepaRG cells. Drug Metab Dispos 37:1259–1268. Quantitative protein determination for CYP induction via LC-MS/MS. Proteomics 11:33–41. Phimmasone S and Kharasch ED (2001) A pilot evaluation of alfentanil-induced miosis as a Xia B, Barve A, Heimbach T, Zhang T, Gu H, Wang L, Einolf H, Alexander N, Hanna I, and Ke J, noninvasive probe for hepatic cytochrome P450 3A4 (CYP3A4) activity in humans. Clin et al. (2014) Physiologically based pharmacokinetic modeling for assessing the clinical drug- Pharmacol Ther 70:505–517. drug interaction of alisporivir. Eur J Pharm Sci 63:103–112. Reitman ML, Chu X, Cai X, Yabut J, Venkatasubramanian R, Zajic S, Stone JA, Ding Y, Witter R, Xu Y, Zhou Y, Hayashi M, Shou M, and Skiles GL (2011) Simulation of clinical drug-drug and Gibson C, et al. (2011) Rifampin’s acute inhibitory and chronic inductive drug interactions: interactions from hepatocyte CYP3A4 induction data and its potential utility in trial designs. experimental and model-based approaches to drug-drug interaction trial design. Clin Pharmacol Drug Metab Dispos 39:1139–1148. Ther 89 – :234 242. Yang J, Jamei M, Yeo KR, Tucker GT, and Rostami-Hodjegan A (2007) Prediction of intestinal Downloaded from Rostami-Hodjegan A, Tamai I, and Pang KS (2012) Physiologically based pharmacokinetic first-pass drug metabolism. Current Drug Metabolism 8:676–684. (PBPK) modeling: it is here to stay! Biopharm Drug Dispos 33:47–50. Yang J, Liao M, Shou M, Jamei M, Yeo KR, Tucker GT, and Rostami-Hodjegan A (2008) Rowland Yeo K, Jamei M, Yang J, Tucker GT, and Rostami-Hodjegan A (2010) Physiologically Cytochrome p450 turnover: regulation of synthesis and degradation, methods for de- based mechanistic modelling to predict complex drug-drug interactions involving simultaneous termining rates, and implications for the prediction of drug interactions. Curr Drug Metab 9: competitive and time-dependent enzyme inhibition by parent compound and its metabolite in 384–394. both liver and gut—the effect of diltiazem on the time-course of exposure to triazolam. Eur J Zhao P, Rowland M, and Huang SM (2012) Best practice in the use of physiologically based Pharm Sci 39:298–309. pharmacokinetic modeling and simulation to address clinical pharmacology regulatory questions. Schellens JH, van der Wart JH, Brugman M, and Breimer DD (1989) Influence of enzyme Clin Pharmacol Ther 92:17–20. induction and inhibition on the oxidation of nifedipine, sparteine, mephenytoin and antipyrine in humans as assessed by a “cocktail” study design. JPharmacolExpTher249:638–645. dmd.aspetjournals.org Schmider J, Brockmöller J, Arold G, Bauer S, and Roots I (1999) Simultaneous assessment of CYP3A4 Address correspondence to: Lisa M. Almond, Simcyp Limited (a Certara andCYP1A2activityinvivowithalprazolamandcaffeine.Pharmacogenetics 9:725–734. Szalat A, Gershkovich P, Ben-Ari A, Shaish A, Liberman Y, Boutboul E, Gotkine M, Hoffman A, Company), Blades Enterprise Centre, John Street, Sheffield, S2 4SU, UK. E-mail: Harats D, and Leitersdorf E, et al. (2007) Rifampicin-induced CYP3A4 activation in CTX [email protected] patients cannot replace chenodeoxycholic acid treatment. Biochim Biophys Acta 1771 :839–844. at ASPET Journals on October 1, 2021 Prediction of DDIs arising from CYP3A induction using a physiologically-based dynamic model – SUPPLEMENTARY MATERIAL Lisa M Almond, Sophie Mukadam, Iain Gardner, Krystle Okialda, Susan Wong, Oliver Hatley, Suzanne Tay, Karen Rowland-Yeo, Masoud Jamei, Amin Rostami-Hodjegan and Jane R Kenny Drug Metabolism and Disposition

Table 1 – Input parameters of the victim drugs (substrates) used in simulations Parameter Value Reference Alfentanil MW 416.52 Meta-analysis (Meuldermans et al., 1982); (Bower and Hull, 1982); (Schuttler fu 0.12 and Stoeckel, 1982); (Meistelman et al., 1987); (Roure et al., 1987); (Beaumont et al., 2011); (Lemmens et al., 1992) (Meuldermans et al., 1982) B:P 0.66 (Beaumont et al., 2011) logP 2.16 (Mather, 1983) Compound type Monoprotic Base pKa(s) 6.5 (Meuldermans et al., 1982); (Mather, 1983) fa – predicted 0.99 Predicted from Caco-2 data (Gertz et al., 2010) ka (h-1) - predicted 4.8 Predicted from Caco-2 data (Gertz et al., 2010) Optimised to recover observed FG. (Kharasch and Stubbert, 2013); (Kharasch et al., 2011a); (Kharasch et al., 2004a); (Kharasch et al., 2005); (Kharasch et al., Qgut (L/h) 2.2 2007); (Kharasch et al., 2009); (Kharasch et al., 2008); (Kharasch et al., 2012); (Kharasch et al., 2011b) Caco-2 (Papp A-B) 293 (Gertz et al., 2010) fugut 1 Assumed Vss(L/kg) – predicted (Minimal PBPK) 0.397 Rodgers and Rowland method (Kp scalar 0.45 to recover the observed Vss)

Meta-analysis (Bower and Hull, 1982); (Bentley et al., 1983); (Bower and Sear, 1989); (Chauvin et al., 1987); (Ibrahim et al., 2003); (Kharasch et al., 2007); Vss (L/kg) - observed 0.371 (Kharasch et al., 1999); (Kharasch et al., 1997b); (Kharasch et al., 1997a); (Kharasch et al., 2011a); (Kharasch et al., 2004a); (Kharasch et al., 2011b); (Scott and Stanski, 1987); (Meistelman et al., 1987); (Meuldermans et al., 1988)

Meta-analysis (Bovill et al., 1982); (Bower and Hull, 1982); (Bower and Sear, 1989); (Camu et al., 1982); (Ibrahim et al., 2003); (Fragen et al., 1983); (Egan et CLiv (L/h) 19.9 al., 1996); (Helmers et al., 1984); (Mertens et al., 2001); (Scott and Stanski, 1987); (Kharasch and Stubbert, 2013); (Kharasch et al., 2011a); (Kharasch et al., 2004a); (Kharasch et al., 2005); (Kharasch et al., 2007); (Kharasch et al., 2009); (Kharasch et al., 1999); (Kharasch et al., 1997a); (Kharasch et al., 1997b); (Kharasch et al., 2008); (Kharasch et al., 2012); (Kharasch et al., 2011b); (McDonnell et al., 1982); (Meistelman et al., 1987); (Meuldermans et al., 1988); (Phimmasone and Kharasch, 2001); (Roure et al., 1987); (Schuttler and Stoeckel, 1982); (McDonnell et al., 2003); (Bentley et al., 1983) rCYP3A4 CLint (µL/min/pmol) 0.49 Calculated from CLiv using the Retrograde model and fmCYP3A4 = 0.93 Add Clint (µL/min/mg mic protein) 4.37 Calculated using the Retrograde model CLR (L/h) 0.06 Meta-analysis (Meuldermans et al., 1988); (Schuttler and Stoeckel, 1982) Alprazolam MW 308.8 (Scavone et al., 1988); (Ochs et al., 1986); (Moschitto and Greenblatt, 1983); fu 0.29 (Greenblatt et al., 1983) (Obach, 1999) & unpublished experimental data obtained by personal B:P 0.825 communication Compound type Monoprotic Base pKa(s) 2.4 fa 1 Assumed (Smith et al., 1984); (Lin et al., 1988); (Kirkwood et al., 1991); (Amchin et al., ka (h-1) 3.56 1998); (Greenblatt et al., 1983) Qgut (L/h) 15.6 Predicted from PSA fugut 1 Assumed PSA (A2) 43.07 Calculated with Marvin HBD 0 Calculated with Marvin (Lin et al., 1988); (Wong et al., 1998); (Kaplan et al., 1998); (Amchin et al., Vss(L/kg) (Minimal PBPK) 0.99 1998); (Stoehr et al., 1984) CL – Enzyme Kinetics

4-hydroxylation rCYP3A4 Vmax (µL/min/pmol) 17.5 Km (µM) 256.7 rCYP3A5 Vmax (µL/min/pmol) 5.99 Km (µM) 211.4 Meta-analysis (corrected for fumic & ISEF) oxylation (Hirota et al., 2001); (Galetin et al., 2004) rCYP3A4 Vmax (µL/min/pmol) 0.8 (Williams et al., 2002) Km (µM) 118.2 rCYP3A5 Vmax (µL/min/pmol) 2.29 Km (µM) 205.1

CLR (L/h) 0.678 (Fraser et al., 1991) Midazolam MW 325.8 Meta-analysis (Allonen et al., 1981); (Thummel et al., 1996); (Greenblatt et al., fu 0.032 1984); (Moschitto and Greenblatt, 1983) (Allonen et al., 1981); (Heizmann et al., 1983); (Bjorkman et al., 2001) & B:P 0.603 unpublished experimental data obtained by personal communication Compound type Ampholyte pKa(s) 10.95, 6.2 fa 1 Assumed ka (h-1) 3 (Allonen et al., 1981) Qgut (L/h) 14.0 Predicted from Caco-2 data (von Richter et al., 2009) fugut 1 Assumed Meta-analysis (Saari et al., 2006); (Ibrahim et al., 2002); (Wandel et al., 2000); Vss(L/kg) (Minimal PBPK) 1 (Schwagmeier et al., 1998); (Heizmann et al., 1983); (Allonen et al., 1981) Enzyme Kinetics

1-hydroxylation rCYP3A4 Vmax (µL/min/pmol) 5.23 Km (µM) 2.16 rCYP3A5 Vmax (µL/min/pmol) 19.7 Km (µM) 4.16 Meta-analysis, corrected for fumic & ISEF (Emoto et al., 2003); (Galetin et al., 4-hydroxylation 2004); (Huang et al., 2004); (Nakajima et al., 2002); (Soars et al., 2006); (Walsky rCYP3A4 V (µL/min/pmol) 5.2 max and Obach, 2004); (Weaver et al., 2003); (Williams et al., 2002); (Hyland et al., Km (µM) 31.8 2009) rCYP3A5 Vmax (µL/min/pmol) 4.03 Km (µM) 34.8 rUGT1A4 Vmax (µL/min/mg) 445 Km (µM) 40.3 CLR (L/h) 0.085 Fe 0.0033 from (Thummel et al., 1996) Nifedipine MW 346.3 fu 0.039 (Ahsan et al., 1993); (Krecic-Shepard et al., 2000) (Holtbecker et al., 1996) & unpublished experimental data obtained by personal B:P 0.685 communication logP 2.69 (Masumoto et al., 1995); (Volgyi et al., 2008); (Lombardo et al., 2000) Compound type Monoprotic Base pKa(s) 2.82 Marvin, ACD Labs fa 1 Assumed ka (h-1) 3.67 (Ahsan et al., 1993) Qgut (L/h) 15.8 Predicted from MDCK data (Polli et al., 2001) fugut 0.5 Modified to recover observed FG Vss(L/kg) (Minimal PBPK) 0.57 (Raemsch and Sommer, 1983); (Holtbecker et al., 1996); (Foster et al., 1983) CLiv (L/h) 33.6 (Holtbecker et al., 1996); (Raemsch and Sommer, 1983) Enzyme Kinetics Calculated using the retrograde model and fm3A4 reported in the literature (Ohno rCYP3A4 V (µL/min/pmol) 33.8 max et al., 2007; Foti et al., 2010) Meta-analysis from in vitro data (Carr et al., 2006); (Emoto and Iwasaki, 2007); Km (µM) 10.95 (Williams et al., 2002; Emoto and Iwasaki, 2006) Add HLM Clint (µL/min/mg) 133.32 Calculated using the retrograde model and fm3A4 reported in the literature CLR (L/h) negligible (Raemsch and Sommer, 1983) Quinidine MW 324.4 (Mihaly et al., 1987); (Kessler and Perez, 1981); (Hughes et al., 1975); (Kates et al., 1978); (Perez-Mateo and Erill, 1977); (Affrime and Reidenberg, 1975); fu 0.203 (Nilsen et al., 1978); (Ochs et al., 1978b); (Woo and Greenblatt, 1979); (Ochs et al., 1980); (Ochs et al., 1978a); (Kessler et al., 1978) (Hughes et al., 1975); (Rodgers and Rowland, 2007) & unpublished experimental B:P 0.88 data obtained by personal communication logP 2.88 (Hansch et al., 1995) Compound type Diprotic Base pKa(s) 4.2, 8.8 (Grube et al., 2009) fa 1 Assumed Optimised to recover observed tmax (Edwards et al., 1987); (Laganiere et al., ka (h-1) 3 1996) Qgut (L/h) 11.7 Predicted from Caco-2 data (von Richter et al., 2009) fugut 1 Assumed Vss(L/kg) – Predicted Corrected Poulin & Theil, corrected by Berezhkovskiy (Poulin and Theil, 2002); 2.03 Minimal PBPK – M1 (Berezhkovskiy, 2004) Enzyme Kinetics 3-hrdoxylation HLM CYP3A4 CLint (µL/min/mg) 20.7 N-oxidation Meta-analysis, corrected for fumic (Nielsen et al., 1999) HLM CYP3A4 CLint (µL/min/mg) 2.6 HLM CYP2C9 CLint (µL/min/mg) 0.40 HLM CYP2E1 CLint (µL/min/mg) 0.52 CLR (L/h) 1.95 (Darbar et al., 1997) CYP2D6 Ki (µM) 0.017 (Ching et al., 1995); (Broly et al., 1989); (Lalovic et al., 2004) (Nielsen et al., 1999); (Ngui et al., 2000); (Guengerich et al., 1986), fu 0.58 CYP3A4 Ki (µM) 40 mic (predicted)

3-OH Quinidine MW 340.4 fu 0.281 Simcyp Simulator B:P 1 Assumed logP 2 Calculated (ALogPS website & Marvin) Compound type Diprotic Base pKa(s) 4.2, 2.8 Calculated (Marvin & Moca) Vss(L/kg) – Predicted 1.28 (Rodgers and Rowland, 2007) Minimal PBPK – M2 CLpo (L/h) 52.6 (Lecocq et al., 1988); (Vozeh et al., 1985) CLR (L/h) 16.7 (Lecocq et al., 1988); (Vozeh et al., 1985) Simvastatin MW 418.57 fu 0.02 (Vickers et al., 1990) B:P 1 logP 4.68 (Hansch et al., 1995) Compound type Neutral pKa(s) - fa 1 Assumed ka (h-1) 1.43 Fitted to recover concentration-time profile Qgut (L/h) 8.07 Predicted from Caco-2 data ((Gertz et al., 2010)) fugut 1 Assumed Vss(L/kg) (Minimal PBPK) 2.26 Fitted to recover concentration time profiles Enzyme Kinetics HLM-CYP3A4 CLint (µL/min/mg) 1873 Optimised Add HLM-CYP CLint (µL/min/mg) 1873 Optimised CLR (L/h) 0 Zolpidem MW 307.39 fu 0.08 B:P 0.76 (Obach, 1999) Compound type Monoprotic Base pKa(s) 6.16 (Durand et al., 1992) fa 1 Assumed ka (h-1) 2.25 (Olubodun et al., 2003); (Patat et al., 1994) Qgut (L/h) 16.3 Predicted from MDCK II data (Mahar Doan et al., 2002) fugut 1 Assumed Vss(L/kg) (Minimal PBPK) 0.68 (Patat et al., 1994) Enzyme Kinetics

M3 rCYP1A2 Vmax (µL/min/pmol) 7.99 Km (µM) 38 rCYP2C9 Vmax (µL/min/pmol) 17.42 Km (µM) 103 rCYP2C19 Vmax (µL/min/pmol) 0.8 Km (µM) 133 rCYP2D6 Vmax (µL/min/pmol) 1.44 Km (µM) 190 rCYP3A4 Vmax (µL/min/pmol) 15.48 Km (µM) 123 rCYP2J2 Vmax (µL/min/pmol) 1.41 Km (µM) 340 M4 (Von Moltke et al., 1999), corrected for fumic & ISEF. rCYP1A2 Vmax (µL/min/pmol) 0.777 Km (µM) 40 rCYP2C9 Vmax (µL/min/pmol) 0.888 Km (µM) 81 rCYP2D6 Vmax (µL/min/pmol) 4.68 Km (µM) 214 rCYP3A4 Vmax (µL/min/pmol) 1.41 Km (µM) 340 rCYP2J2 Vmax (µL/min/pmol) 6.86 Km (µM) 399 M11 rCYP3A4 Vmax (µL/min/pmol) 6.89 Km (µM) 399 CLR (L/h) 0.18 fe = 0.01; (Salva and Costa, 1995).

Table 2 – Input parameters of the perpetrator drugs (inducers) used in simulations Parameter Value Reference Rifampicin MW 823 fu 0.15 (Burman et al., 2001) B:P 0.9 (Loos et al., 1985) logP 3.28 AlogPS website Compound type Ampholyte pKa(s) 1.7, 7.9 fa 1 Assumed ka (h-1) 0.51 (Drusano et al., 1986a) Vss(L/kg) (Minimal PBPK) 0.33 (Loos et al., 1985) CLiv (L/h) 7 Simcyp Library Value CLR (L/h) 1.2 Polk 2001 CYP3A4 Ki (µM) 10.5 (Kajosaari et al., 2005)– corrected for fumic CYP3A4 Ind (Fold, Emax +1) 8 max Base Model, Fitted from in vivo data. (Tran et al., 1999); (Acocella et al., 1971) CYP3A5 IndC50 (µM) 0.32 Carbamazepine MW 236.3 Meta-analysis ((Grimsley et al., 1991); (Di Salle et al., 1974); (Rawlins et al., 1975); (Ramsay et al., fu 0.25 1990); (Riva et al., 1984); (MacKichan and Zola, 1984); (Curran et al., 2011) B:P 1.07 Meta-analysis ((Bonneton et al., 1992); (de Groot et al., 1984) (Zhu et al., 2002); (Lombardo et al., 2000); (Henczi et al., 1995); (Wong et al., 2004); Dal Pozzo, A., et logP 2.22 al. 1989; (Wan et al., 2009); (Winiwarter et al., 1998) Compound type Neutral pKa(s) - fa 0.84 (Tchaparian et al., 2008); (Faigle and Feldmann, 1975); (Levy et al., 1975) ka (h-1) 0.5 (Geradin et al., 1976); (Grimsley et al., 1991); (Olling et al., 1999); (Wong et al., 1983) . Qgut (L/h) 12.6 Predicted from Peff,man (Winiwarter et al., 1998) Vss(L/kg) (Minimal PBPK) 0.78 (Ramsay et al., 1990) Enzyme Kinetics rCYP3A4 V (pmol/min/pmol) 0.72 max (Cazali et al., 2003); (Huang et al., 2004) Km (µM) 180 rCYP3A5 V (pmol/min/pmol) 1.44 max (Huang et al., 2004) Km (µM) 332 rCYP2C8 V (pmol/min/pmol) 0.03 max (Cazali et al., 2003) Km (µM) 741.7 rUGT2B7 Vmax (pmol/min/pmol) 4.18 (Staines et al., 2004) Km (µM) 25.4 Add rCYP CLint (µL /min/pmol) 0.010 Added to recover degree of auto-induction via enzymes other than CYP3A4 Add HLM CLint (µL/min/mg) 0.255 Other enzymes were combined from (Pearce et al., 2002) CLR (L/h) 0.0084 (Kim et al., 2005) Induction parameters mRNA/activity data Generated as part of this study (Table 2) Carbamazepine-10,11-epoxide MW 252.7 Meta-analysis (Ramsay et al., 1990); (Riva et al., 1984); (Riad and Sawchuk, 1998); (MacKichan and fu 0.48 Zola, 1984)

B:P 1.53 (Bonneton et al., 1992) logP 1.44 AlogPS website; PubChem website Compound type Neutral pKa(s) - Vss(L/kg) (Minimal PBPK) 0.78 Assumed equal to parent CLpo (L/h) 6.07 (Pisani et al., 1988); (Pisani et al., 1992); (Spina et al., 1988); (Tomson et al., 1983) CLR (L/h) 0.14 (Kim et al., 2005) Induction parameters Equal to parent Assumed based on data showing equipotency (Oscarson et al., 2006) Phenytoin MW 252.28 Meta-analysis (Odar-Cederlof and Borga, 1974); (Tassaneeyakul et al., 1992); (Ducharme et al., 1995); fu 0.10 (Kurata and Wilkinson, 1974); (Gugler et al., 1975); (Bochner et al., 1973) B:P 0.61 (Kurata and Wilkinson, 1974) Meta-analysis (Martinavarro-Dominguez et al., 2002); (Pade and Stavchansky, 1998); (Avdeef et al., logP 2.47 2000); (Taillardat-Bertschinger et al., 2002) Compound type Monoprotic Acid Meta-analysis (Pade and Stavchansky, 1998); (Avdeef et al., 2000); (Taillardat-Bertschinger et al., 2002); pKa(s) 8.15 (Loeuillet-Ritzler and Faller, 2004) fa 0.9 (Bochner et al., 1973) ka (h-1) 0.53 (Tassaneeyakul et al., 1992) Qgut (L/h) 13.2 Predicted from Caco-2 data (Irvine et al., 1999)

Vss(L/kg) (Minimal PBPK) 0.57 Meta-analysis (Gugler et al., 1976); (Odar-Cederlof and Borga, 1974); (Glazko et al., 1969); (Lund et al., 1974); (Perucca et al., 1978) CLiv (L/h) 1.88 Meta-analysis (Gugler et al., 1976); (Glazko et al., 1969); (Lund et al., 1974); (Perucca et al., 1978) Enzyme Kinetics rCYP2C9 Vmax (pmol/min/pmol) 0.24 Calculated from CLiv using retrograde model Km (µM) 4.1 (Rowland et al., 2008) rCYP2C19 Vmax (pmol/min/pmol) 1.53 Calculated from CLiv using retrograde model Km (µM) 36.8 (Giancarlo et al., 2001); (Bajpai et al., 1996) Add HLM CLint (µL/min/mg) 0.97 Calculated from CLiv using retrograde model CLR (L/h) 0.015 fe 0.008 (Kozelka and Hine, 1943); (Borga et al., 1979) CYP2C9 Ind (Fold, Emax +1) 10.7 max (Sahi et al., 2009) CYP2C9 IndC50 (µM) 9.8 CYP2B6 Ind (Fold, Emax +1) 1.9 max (Hariparsad et al., 2008) CYP2B6 IndC50 (µM) 15.3 CYP3A4/5 Indmax (Fold, Emax +1) mRNA/activity data Generated as part of this study (Table 2) CYP3A4/5 IndC50 (µM) mRNA/activity data Generated as part of this study (Table 2) Phenobarbital MW 232 (Wallin and Herngren, 1985); (Ehrnebo and Odar-Cederlof, 1975); (Ehrnebo et al., 1971);(Bender et al., fu 0.49 1975); (Shou et al., 2008) B:P 0.83 (Ehrnebo and Odar-Cederlof, 1975) logP 1.47 Drug Bank Database, (Xu et al., 2011) Compound type Monoprotic Acid pKa(s) 7.3 (Xu et al., 2011) fa 1 (Varma et al., 2012) ka (h-1) 2 Recovers the concentration-time profile of Wilensky et al., 1982 Vss(L/kg) (Minimal PBPK) 0.54 (Wilensky et al., 1982) CLiv (L/h) 0.3 (Xu et al., 2011) CLR (L/h) 0.07 fe 0.24 (Varma et al., 2012) CYP3A4/5 Indmax (Fold, Emax +1) mRNA/activity data Generated as part of this study (Table 2) CYP3A4/5 IndC50 (µM) mRNA/activity data Generated as part of this study (Table 2)

Table 3 Meta-analysis of studies where midazolam (victim drug) was administered orally

Study 1/AUC ratio Standard Deviation n

(Backman et al., 1996) 24.3 14.2 10

(Backman et al., 1998) 63.0 16.4 9

(Eap et al., 2004) 19.1 24.3* 4

(Gurley et al., 2006) 17.5 20.3* 19

(Gurley et al., 2008) 16.6 20.2* 16

(Reitman et al., 2011) 8.1 10.1* 11

(Kharasch et al., 2004b) 19.0 20.8* 8

(Gorski et al., 2003) 25.6 12.2* 52

Weighted Geometric Mean 18.1

CI of the observations 5.4-60.4

*Calculated using previously published methodology (Einolf, 2007; Cubitt et al., 2011; Ghobadi et al., 2011; Barter et al., 2013); 3 of the clinical studies used for assessment of DDI prediction (Floyd et al., 2003; Chung et al., 2006; Link et al., 2008) didn’t contain enough information to allow calculation of standard deviation and therefore are not included in the weighted geometric mean. Figure 1 Simulated (line) and observed (open circles) mean systemic concentration time profiles of inducers after multiple dosing of A) rifampicin 600 mg q.d. (Acocella et al., 1971), B) rifampicin 300 mg b.d. (Drusano et al., 1986b), C) carbamazepine 200 mg t.i.d with unequal dosing intervals (Eichelbaum et al., 1975), D) phenytoin 300 mg q.d. (Gugler et al., 1976) and E) phenobarbital 90 mg q.d. (Ferron et al., 2003).

Figure 2 Simulated (open) and reported (black) fmCYP3A4 (A) and FG (B) values for the victim drugs used in these analyses. Where multiple reports were available, an average was taken. In the cases of alfentanil and nifedipine a retrograde approach was used to recover the clearance and reported fm and FG incorporated. Clinical data gave a wide range of fm values (0.72-0.93) for alfentanil but either the inhibitor or trial design was not optimal and hence an fm was derived from in vitro data (0.93, Labroo et al) and verified using a clinical ketoconazole study. For other victim drugs fmCYP3A4 is propagated from in vitro metabolism data, incorporating other routes of elimination (for fm) and permeability data (for FG).

REFERENCES

Acocella G, Pagani V, Marchetti M, Baroni GC and Nicolis FB (1971) Kinetic studies on rifampicin.

I. Serum concentration analysis in subjects treated with different oral doses over a period of

two weeks. Chemotherapy 16:356-370.

Affrime M and Reidenberg MM (1975) The protein binding of some drugs in plasma from patients

with alcoholic liver disease. Eur J Clin Pharmacol 8:267-269.

Ahsan CH, Renwick AG, Waller DG, Challenor VF, George CF and Amanullah M (1993) The

influence of dose and ethnic origins on the pharmacokinetics of nifedipine. Clin Pharmacol

Ther 54:329-338.

Allonen H, Ziegler G and Klotz U (1981) Midazolam kinetics. Clin Pharmacol Ther 30:653-661.

Amchin J, Zarycranski W, Taylor KP, Albano D and Klockowski PM (1998) Effect of venlafaxine on

the pharmacokinetics of alprazolam. Psychopharmacol Bull 34:211-219.

Avdeef A, Berger CM and Brownell C (2000) pH-metric solubility. 2: correlation between the acid-

base titration and the saturation shake-flask solubility-pH methods. Pharm Res 17:85-89.

Backman JT, Kivisto KT, Olkkola KT and Neuvonen PJ (1998) The area under the plasma

concentration-time curve for oral midazolam is 400-fold larger during treatment with

itraconazole than with rifampicin. Eur J Clin Pharmacol 54:53-58.

Backman JT, Olkkola KT and Neuvonen PJ (1996) Rifampin drastically reduces plasma

concentrations and effects of oral midazolam. Clin Pharmacol Ther 59:7-13.

Bajpai M, Roskos LK, Shen DD and Levy RH (1996) Roles of cytochrome P4502C9 and cytochrome

P4502C19 in the stereoselective metabolism of phenytoin to its major metabolite. Drug

Metab Dispos 24:1401-1403.

Barter ZE, Tucker GT and Rowland-Yeo K (2013) Differences in cytochrome p450-mediated

pharmacokinetics between chinese and caucasian populations predicted by mechanistic

physiologically based pharmacokinetic modelling. Clin Pharmacokinet 52:1085-1100.

Beaumont K, Gardner I, Chapman K, Hall M and Rowland M (2011) Toward an integrated human

clearance prediction strategy that minimizes animal use. J Pharm Sci 100:4518-4535. Bender AD, Post A, Meier JP, Higson JE and Reichard G, Jr. (1975) Plasma protein binding of drugs

as a function of age in adult human subjects. J Pharm Sci 64:1711-1713.

Bentley JB, Finlay JH, Humphrey LR, Gandolfi AJ and Brown BR (1983) Obesity and Alfentanil

Pharmacokinetics Anesth Analg 62:251.

Berezhkovskiy LM (2004) Volume of distribution at steady state for a linear pharmacokinetic system

with peripheral elimination. J Pharm Sci 93:1628-1640.

Bjorkman S, Wada DR, Berling BM and Benoni G (2001) Prediction of the disposition of midazolam

in surgical patients by a physiologically based pharmacokinetic model. J Pharm Sci 90:1226-

1241.

Bochner F, Hooper WD, Sutherland JM, Eadie MJ and Tyrer JH (1973) The renal handling of

diphenylhydantoin and 5-(p-hydroxyphenyl)-5-phenylhydantoin. Clin Pharmacol Ther

14:791-796.

Bonneton J, Genton P and Mesdjian E (1992) Distribution of carbamazepine and its epoxide in blood

compartments in adolescent and adult epileptic patients. Biopharm Drug Dispos 13:411-416.

Borga O, Hoppel C, Odar-Cederlof I and Garle M (1979) Plasma levels and renal excretion of

phenytoin and its metabolites in patients with renal failure. Clin Pharmacol Ther 26:306-314.

Bovill JG, Sebel PS, Blackburn CL and Heykants J (1982) The pharmacokinetics of alfentanil

(R39209): a new opioid analgesic. Anesthesiology 57:439-443.

Bower S and Hull CJ (1982) Comparative pharmacokinetics of fentanyl and alfentanil. Br J Anaesth

54:871-877.

Bower S and Sear JW (1989) Disposition of alfentanil in patients receiving a renal transplant. J

Pharm Pharmacol 41:654-657.

Broly F, Libersa C, Lhermitte M, Bechtel P and Dupuis B (1989) Effect of quinidine on the

dextromethorphan O-demethylase activity of microsomal fractions from human liver. Br J

Clin Pharmacol 28:29-36.

Burman WJ, Gallicano K and Peloquin C (2001) Comparative pharmacokinetics and

pharmacodynamics of the rifamycin antibacterials. Clin Pharmacokinet 40:327-341. Camu F, Gepts E, Rucquoi M and Heykants J (1982) Pharmacokinetics of alfentanil in man. Anesth

Analg 61:657-661.

Carr B, Norcross R, Fang Y, Lu P, Rodrigues AD, Shou M, Rushmore T and Booth-Genthe C (2006)

Characterization of the rhesus monkey CYP3A64 enzyme: species comparisons of CYP3A

substrate specificity and kinetics using baculovirus-expressed recombinant enzymes. Drug

Metab Dispos 34:1703-1712.

Cazali N, Tran A, Treluyer JM, Rey E, d'Athis P, Vincent J and Pons G (2003) Inhibitory effect of

stiripentol on carbamazepine and saquinavir metabolism in human. Br J Clin Pharmacol

56:526-536.

Chauvin M, Lebrault C, Levron JC and Duvaldestin P (1987) Pharmacokinetics of alfentanil in

chronic renal failure. Anesth Analg 66:53-56.

Ching MS, Blake CL, Ghabrial H, Ellis SW, Lennard MS, Tucker GT and Smallwood RA (1995)

Potent inhibition of yeast-expressed CYP2D6 by dihydroquinidine, quinidine, and its

metabolites. Biochem Pharmacol 50:833-837.

Chung E, Nafziger AN, Kazierad DJ and Bertino JS, Jr. (2006) Comparison of midazolam and

simvastatin as cytochrome P450 3A probes. Clin Pharmacol Ther 79:350-361.

Cubitt HE, Yeo KR, Howgate EM, Rostami-Hodjegan A and Barter ZE (2011) Sources of

interindividual variability in IVIVE of clearance: an investigation into the prediction of

benzodiazepine clearance using a mechanistic population-based pharmacokinetic model.

Xenobiotica 41:623-638.

Curran RE, Claxton CR, Hutchison L, Harradine PJ, Martin IJ and Littlewood P (2011) Control and

measurement of plasma pH in equilibrium dialysis: influence on drug plasma protein binding.

Drug Metab Dispos 39:551-557.

Darbar D, Dell'Orto S, Morike K, Wilkinson GR and Roden DM (1997) Dietary salt increases first-

pass elimination of oral quinidine. Clin Pharmacol Ther 61:292-300. de Groot G, van Heijst AN and Maes RA (1984) Charcoal hemoperfusion in the treatment of two

cases of acute carbamazepine poisoning. J Toxicol Clin Toxicol 22:349-362. Di Salle E, Pacifici GM and Morselli PL (1974) Studies on plasma protein binding of carbamazepine.

Pharmacol Res Commun 6:193-202.

Drusano GL, Townsend RJ, Walsh TJ, Forrest A, Antal EJ and Standiford HC (1986a) Steady-state

serum pharmacokinetics of novobiocin and rifampin alone and in combination. Antimicrob

Agents Chemother 30:42-45.

Drusano GL, Townsend RJ, Walsh TJ, Forrest A, Antal EJ and Standiford HC (1986b) Steady-state

serum pharmacokinetics of novobiocin and rifampin alone and in combination. Antimicrob

Agents Chemother 30:42-45.

Ducharme MP, Slaughter RL, Warbasse LH, Chandrasekar PH, Van de Velde V, Mannens G and

Edwards DJ (1995) Itraconazole and hydroxyitraconazole serum concentrations are reduced

more than tenfold by phenytoin. Clin Pharmacol Ther 58:617-624.

Durand A, Thenot JP, Bianchetti G and Morselli PL (1992) Comparative pharmacokinetic profile of

two imidazopyridine drugs: zolpidem and alpidem. Drug Metab Rev 24:239-266.

Eap CB, Buclin T, Cucchia G, Zullino D, Hustert E, Bleiber G, Golay KP, Aubert AC, Baumann P,

Telenti A and Kerb R (2004) Oral administration of a low dose of midazolam (75 microg) as

an in vivo probe for CYP3A activity. Eur J Clin Pharmacol 60:237-246.

Edwards DJ, Lavoie R, Beckman H, Blevins R and Rubenfire M (1987) The effect of

coadministration of verapamil on the pharmacokinetics and metabolism of quinidine. Clin

Pharmacol Ther 41:68-73.

Egan TD, Minto CF, Hermann DJ, Barr J, Muir KT and Shafer SL (1996) Remifentanil versus

alfentanil: comparative pharmacokinetics and pharmacodynamics in healthy adult male

volunteers. Anesthesiology 84:821-833.

Ehrnebo M, Agurell S, Jalling B and Boreus LO (1971) Age differences in drug binding by plasma

proteins: studies on human foetuses, neonates and adults. Eur J Clin Pharmacol 3:189-193.

Ehrnebo M and Odar-Cederlof I (1975) Binding of amobarbital, pentobarbital and diphenylhydantoin

to blood cells and plasma proteins in healthy volunteers and uraemic patients. Eur J Clin

Pharmacol 8:445-453. Einolf HJ (2007) Comparison of different approaches to predict metabolic drug-drug interactions.

Xenobiotica 37:1257-1294.

Emoto C and Iwasaki K (2006) Enzymatic characteristics of CYP3A5 and CYP3A4: a comparison of

in vitro kinetic and drug-drug interaction patterns. Xenobiotica 36:219-233.

Emoto C and Iwasaki K (2007) Approach to predict the contribution of cytochrome P450 enzymes to

drug metabolism in the early drug-discovery stage: the effect of the expression of cytochrome

b(5) with recombinant P450 enzymes. Xenobiotica 37:986-999.

Emoto C, Murase S, Sawada Y, Jones BC and Iwasaki K (2003) In vitro inhibitory effect of 1-

aminobenzotriazole on drug oxidations catalyzed by human cytochrome P450 enzymes: a

comparison with SKF-525A and ketoconazole. Drug Metab Pharmacokinet 18:287-295.

Faigle JW and Feldmann KF (1975) Pharmacokinetic Data of Carbamazepine and Its Major

Metabolites in Man, in: Clinical Pharmacology of Anti-Epileptic Drugs (Schneider H, Janz D,

Gardner-Thorpe C, Meinardi H and Sherwin AL eds), pp 159-165, Springer Berlin

Heidelberg.

Floyd MD, Gervasini G, Masica AL, Mayo G, George AL, Jr., Bhat K, Kim RB and Wilkinson GR

(2003) Genotype-phenotype associations for common CYP3A4 and CYP3A5 variants in the

basal and induced metabolism of midazolam in European- and African-American men and

women. Pharmacogenetics 13:595-606.

Foster TS, Hamann SR, Richards VR, Bryant PJ, Graves DA and McAllister RG (1983) Nifedipine

kinetics and bioavailability after single intravenous and oral doses in normal subjects. J Clin

Pharmacol 23:161-170.

Foti RS, Rock DA, Wienkers LC and Wahlstrom JL (2010) Selection of alternative CYP3A4 probe

substrates for clinical drug interaction studies using in vitro data and in vivo simulation. Drug

Metab Dispos 38:981-987.

Fragen RJ, Booij LH, Braak GJ, Vree TB, Heykants J and Crul JF (1983) Pharmacokinetics of the

infusion of alfentanil in man. Br J Anaesth 55:1077-1081. Fraser AD, Bryan W and Isner AF (1991) Urinary screening for alprazolam and its major metabolites

by the Abbott ADx and TDx analyzers with confirmation by GC/MS. J Anal Toxicol 15:25-

29.

Galetin A, Brown C, Hallifax D, Ito K and Houston JB (2004) Utility of recombinant enzyme kinetics

in prediction of human clearance: impact of variability, CYP3A5, and CYP2C19 on CYP3A4

probe substrates. Drug Metab Dispos 32:1411-1420.

Geradin AP, Abadie FV, Campestrini JA and Theobald W (1976) Pharmacokinetics of carbamazepine

in normal humans after single and repeated oral doses. J Pharmacokinet Biopharm 4:521-

535.

Gertz M, Harrison A, Houston JB and Galetin A (2010) Prediction of human intestinal first-pass

metabolism of 25 CYP3A substrates from in vitro clearance and permeability data. Drug

Metab Dispos 38:1147-1158.

Ghobadi C, Johnson TN, Aarabi M, Almond LM, Allabi AC, Rowland-Yeo K, Jamei M and Rostami-

Hodjegan A (2011) Application of a systems approach to the bottom-up assessment of

pharmacokinetics in obese patients: expected variations in clearance. Clin Pharmacokinet

50:809-822.

Giancarlo GM, Venkatakrishnan K, Granda BW, von Moltke LL and Greenblatt DJ (2001) Relative

contributions of CYP2C9 and 2C19 to phenytoin 4-hydroxylation in vitro: inhibition by

sulfaphenazole, omeprazole, and ticlopidine. Eur J Clin Pharmacol 57:31-36.

Glazko AJ, Chang T, Baukema J, Dill WA, Goulet JR and Buchanan RA (1969) Metabolic

disposition of diphenylhydantoin in normal human subjects following intravenous

administration. Clin Pharmacol Ther 10:498-504.

Gorski JC, Vannaprasaht S, Hamman MA, Ambrosius WT, Bruce MA, Haehner-Daniels B and Hall

SD (2003) The effect of age, sex, and rifampin administration on intestinal and hepatic

cytochrome P450 3A activity. Clin Pharmacol Ther 74:275-287.

Greenblatt DJ, Abernethy DR, Locniskar A, Harmatz JS, Limjuco RA and Shader RI (1984) Effect of

age, gender, and obesity on midazolam kinetics. Anesthesiology 61:27-35. Greenblatt DJ, Divoll M, Abernethy DR, Moschitto LJ, Smith RB and Shader RI (1983) Alprazolam

kinetics in the elderly. Relation to antipyrine disposition. Arch Gen Psychiatry 40:287-290.

Grimsley SR, Jann MW, Carter JG, D'Mello AP and D'Souza MJ (1991) Increased carbamazepine

plasma concentrations after fluoxetine coadministration. Clin Pharmacol Ther 50:10-15.

Grube S, Langguth P, Junginger HE, Kopp S, Midha KK, Shah VP, Stavchansky S, Dressman JB and

Barends DM (2009) Biowaiver monographs for immediate release solid oral dosage forms:

quinidine sulfate. J Pharm Sci 98:2238-2251.

Guengerich FP, Muller-Enoch D and Blair IA (1986) Oxidation of quinidine by human liver

cytochrome P-450. Mol Pharmacol 30:287-295.

Gugler R, Azarnoff DL and Shoeman DW (1975) Diphenylhydantoin: correlation between protein

binding and albumin concentration. Klin Wochenschr 53:445-446.

Gugler R, Manion CV and Azarnoff DL (1976) Phenytoin: pharmacokinetics and bioavailability. Clin

Pharmacol Ther 19:135-142.

Gurley B, Hubbard MA, Williams DK, Thaden J, Tong Y, Gentry WB, Breen P, Carrier DJ and

Cheboyina S (2006) Assessing the clinical significance of botanical supplementation on

human cytochrome P450 3A activity: comparison of a milk thistle and black cohosh product

to rifampin and clarithromycin. J Clin Pharmacol 46:201-213.

Gurley BJ, Swain A, Hubbard MA, Hartsfield F, Thaden J, Williams DK, Gentry WB and Tong Y

(2008) Supplementation with goldenseal (Hydrastis canadensis), but not kava kava (Piper

methysticum), inhibits human CYP3A activity in vivo. Clin Pharmacol Ther 83:61-69.

Hansch C, Leo A and Hoekman D (1995) Exploring QSAR - Hydrophobic, Electronic, and Steric

Constants. American Chemical Society, Washington, DC.

Hariparsad N, Carr BA, Evers R and Chu X (2008) Comparison of immortalized Fa2N-4 cells and

human hepatocytes as in vitro models for cytochrome P450 induction. Drug Metab Dispos

36:1046-1055.

Heizmann P, Eckert M and Ziegler WH (1983) Pharmacokinetics and bioavailability of midazolam in

man. Br J Clin Pharmacol 16 Suppl 1:43S-49S. Helmers H, Van Peer A, Woestenborghs R, Noorduin H and Heykants J (1984) Alfentanil kinetics in

the elderly. Clin Pharmacol Ther 36:239-243.

Henczi M, Nagy J and Weaver DF (1995) Determination of octanol-water partition coefficients by an

HPLC method for anticonvulsant structure-activity studies. J Pharm Pharmacol 47:345-347.

Hirota N, Ito K, Iwatsubo T, Green CE, Tyson CA, Shimada N, Suzuki H and Sugiyama Y (2001) In

vitro/in vivo scaling of alprazolam metabolism by CYP3A4 and CYP3A5 in humans.

Biopharm Drug Dispos 22:53-71.

Holtbecker N, Fromm MF, Kroemer HK, Ohnhaus EE and Heidemann H (1996) The nifedipine-

rifampin interaction. Evidence for induction of gut wall metabolism. Drug Metab Dispos

24:1121-1123.

Huang W, Lin YS, McConn DJ, 2nd, Calamia JC, Totah RA, Isoherranen N, Glodowski M and

Thummel KE (2004) Evidence of significant contribution from CYP3A5 to hepatic drug

metabolism. Drug Metab Dispos 32:1434-1445.

Hughes IE, Ilett KF and Jellett LB (1975) The distribution of quinidine in human blood. Br J Clin

Pharmacol 2:521-525.

Hyland R, Osborne T, Payne A, Kempshall S, Logan YR, Ezzeddine K and Jones B (2009) In vitro

and in vivo glucuronidation of midazolam in humans. Br J Clin Pharmacol 67:445-454.

Ibrahim A, Karim A, Feldman J and Kharasch E (2002) The influence of parecoxib, a parenteral

cyclooxygenase-2 specific inhibitor, on the pharmacokinetics and clinical effects of

midazolam. Anesth Analg 95:667-673.

Ibrahim AE, Feldman J, Karim A and Kharasch ED (2003) Simultaneous assessment of drug

interactions with low- and high-extraction opioids: application to parecoxib effects on the

pharmacokinetics and pharmacodynamics of fentanyl and alfentanil. Anesthesiology 98:853-

861.

Irvine JD, Takahashi L, Lockhart K, Cheong J, Tolan JW, Selick HE and Grove JR (1999) MDCK

(Madin-Darby canine kidney) cells: A tool for membrane permeability screening. J Pharm Sci

88:28-33. Kajosaari LI, Laitila J, Neuvonen PJ and Backman JT (2005) Metabolism of repaglinide by CYP2C8

and CYP3A4 in vitro: effect of fibrates and rifampicin. Basic Clin Pharmacol Toxicol

97:249-256.

Kaplan GB, Greenblatt DJ, Ehrenberg BL, Goddard JE, Harmatz JS and Shader RI (1998) Single-

dose pharmacokinetics and pharmacodynamics of alprazolam in elderly and young subjects. J

Clin Pharmacol 38:14-21.

Kates RE, Sokoloski TD and Comstock TJ (1978) Binding of quinidine to plasma proteins in normal

subjects and in patients with hyperlipoproteinemias. Clin Pharmacol Ther 23:30-35.

Kessler KM, Humphries WC, Jr., Black M and Spann JF (1978) Quinidine pharmacokinetics in

patients with cirrhosis or receiving propranolol. Am Heart J 96:627-635.

Kessler KM and Perez GO (1981) Decreased quinidine plasma protein binding during hemodialysis.

Clin Pharmacol Ther 30:121-126.

Kharasch ED, Bedynek PS, Hoffer C, Walker A and Whittington D (2012) Lack of indinavir effects

on methadone disposition despite inhibition of hepatic and intestinal cytochrome P4503A

(CYP3A). Anesthesiology 116:432-447.

Kharasch ED, Bedynek PS, Walker A, Whittington D and Hoffer C (2008) Mechanism of ritonavir

changes in methadone pharmacokinetics and pharmacodynamics: II. Ritonavir effects on

CYP3A and P-glycoprotein activities. Clin Pharmacol Ther 84:506-512.

Kharasch ED, Francis A, London A, Frey K, Kim T and Blood J (2011a) Sensitivity of intravenous

and oral alfentanil and pupillary miosis as minimal and noninvasive probes for hepatic and

first-pass CYP3A induction. Clin Pharmacol Ther 90:100-108.

Kharasch ED, Hoffer C, Whittington D, Walker A and Bedynek PS (2009) Methadone

pharmacokinetics are independent of cytochrome P4503A (CYP3A) activity and

gastrointestinal drug transport: insights from methadone interactions with ritonavir/indinavir.

Anesthesiology 110:660-672.

Kharasch ED, Jubert C, Senn T, Bowdle TA and Thummel KE (1999) Intraindividual variability in

male hepatic CYP3A4 activity assessed by alfentanil and midazolam clearance. J Clin

Pharmacol 39:664-669. Kharasch ED, Russell M, Garton K, Lentz G, Bowdle TA and Cox K (1997a) Assessment of

cytochrome P450 3A4 activity during the menstrual cycle using alfentanil as a noninvasive

probe. Anesthesiology 87:26-35.

Kharasch ED, Russell M, Mautz D, Thummel KE, Kunze KL, Bowdle A and Cox K (1997b) The role

of cytochrome P450 3A4 in alfentanil clearance. Implications for interindividual variability in

disposition and perioperative drug interactions. Anesthesiology 87:36-50.

Kharasch ED and Stubbert K (2013) Cytochrome P4503A does not mediate the interaction between

methadone and ritonavir-lopinavir. Drug Metab Dispos 41:2166-2174.

Kharasch ED, Vangveravong S, Buck N, London A, Kim T, Blood J and Mach RH (2011b)

Concurrent assessment of hepatic and intestinal cytochrome P450 3A activities using

deuterated alfentanil. Clin Pharmacol Ther 89:562-570.

Kharasch ED, Walker A, Hoffer C and Sheffels P (2004a) 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.

Kharasch ED, Walker A, Hoffer C and Sheffels P (2004b) 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.

Kharasch ED, Walker A, Hoffer C and Sheffels P (2005) Sensitivity of intravenous and oral alfentanil

and pupillary miosis as minimally invasive and noninvasive probes for hepatic and first-pass

CYP3A activity. J Clin Pharmacol 45:1187-1197.

Kharasch ED, Walker A, Isoherranen N, Hoffer C, Sheffels P, Thummel K, Whittington D and Ensign

D (2007) Influence of CYP3A5 genotype on the pharmacokinetics and pharmacodynamics of

the cytochrome P4503A probes alfentanil and midazolam. Clin Pharmacol Ther 82:410-426.

Kim KA, Oh SO, Park PW and Park JY (2005) Effect of probenecid on the pharmacokinetics of

carbamazepine in healthy subjects. Eur J Clin Pharmacol 61:275-280.

Kirkwood C, Moore A, Hayes P, DeVane CL and Pelonero A (1991) Influence of menstrual cycle and

gender on alprazolam pharmacokinetics. Clin Pharmacol Ther 50:404-409. Kozelka FL and Hine CH (1943) Degradation products of Dilantin. J Pharmacol Exp Ther 77:175-

179.

Krecic-Shepard ME, Park K, Barnas C, Slimko J, Kerwin DR and Schwartz JB (2000) Race and sex

influence clearance of nifedipine: results of a population study. Clin Pharmacol Ther 68:130-

142.

Kurata D and Wilkinson GR (1974) Erythrocyte uptake and plasma binding of diphenylhydantoin.

Clin Pharmacol Ther 16:355-362.

Laganiere S, Davies RF, Carignan G, Foris K, Goernert L, Carrier K, Pereira C and McGilveray I

(1996) Pharmacokinetic and pharmacodynamic interactions between diltiazem and quinidine.

Clin Pharmacol Ther 60:255-264.

Lalovic B, Phillips B, Risler LL, Howald W and Shen DD (2004) Quantitative contribution of

CYP2D6 and CYP3A to oxycodone metabolism in human liver and intestinal microsomes.

Drug Metab Dispos 32:447-454.

Lecocq B, Jaillon P, Lecocq V, Ferry A, Gardin ME, Jarreau C, Leroyer R, Pays M and Jarreau FX

(1988) Clinical pharmacology of hydroxy-3(S)-dihydroquinidine in healthy volunteers

following oral administration. J Cardiovasc Pharmacol 12:445-450.

Levy RH, Pitlick WH, Troupin AS, Green JR and Neal JM (1975) Pharmacokinetics of

carbamazepine in normal man. Clin Pharmacol Ther 17:657-668.

Lin KM, Lau JK, Smith R, Phillips P, Antal E and Poland RE (1988) Comparison of alprazolam

plasma levels in normal Asian and Caucasian male volunteers. Psychopharmacology (Berl)

96:365-369.

Link B, Haschke M, Grignaschi N, Bodmer M, Aschmann YZ, Wenk M and Krahenbuhl S (2008)

Pharmacokinetics of intravenous and oral midazolam in plasma and saliva in humans:

usefulness of saliva as matrix for CYP3A phenotyping. Br J Clin Pharmacol 66:473-484.

Loeuillet-Ritzler F and Faller B (2004) High-Throughput pKa with the Sirius Profiler SGA using a

co-solvent approach, in: 3rd Lipophilicity Symposium LogP2004, Zürich, Switzerland.

Lombardo F, Shalaeva MY, Tupper KA, Gao F and Abraham MH (2000) ElogPoct: a tool for

lipophilicity determination in drug discovery. J Med Chem 43:2922-2928. Loos U, Musch E, Jensen JC, Mikus G, Schwabe HK and Eichelbaum M (1985) Pharmacokinetics of

oral and intravenous rifampicin during chronic administration. Klin Wochenschr 63:1205-

1211.

Lund L, Alvan G, Berlin A and Alexanderson B (1974) Pharmacokinetics of single and multiple doses

of phenytoin in man. Eur J Clin Pharmacol 7:81-86.

MacKichan JJ and Zola EM (1984) Determinants of carbamazepine and carbamazepine 10,11-

epoxide binding to serum protein, albumin and alpha 1-acid glycoprotein. Br J Clin

Pharmacol 18:487-493.

Mahar Doan KM, Humphreys JE, Webster LO, Wring SA, Shampine LJ, Serabjit-Singh CJ, Adkison

KK and Polli JW (2002) Passive permeability and P-glycoprotein-mediated efflux

differentiate central nervous system (CNS) and non-CNS marketed drugs. J Pharmacol Exp

Ther 303:1029-1037.

Martinavarro-Dominguez A, Capella-Peiro ME, Gil-Agusti M, Marcos-Tomas JV and Esteve-Romero

J (2002) Therapeutic drug monitoring of anticonvulsant drugs by micellar HPLC with direct

injection of serum samples. Clin Chem 48:1696-1702.

Masumoto K, Takeyasu A, Oizumi K and Kobayashi T (1995) [Studies of novel 1,4-dihydropyridine

Ca antagonist CS-905. I. Measurement of partition coefficient (log P) by high performance

liquid chromatography (HPLC)]. Yakugaku Zasshi 115:213-220.

Mather LE (1983) Clinical pharmacokinetics of fentanyl and its newer derivatives. Clin

Pharmacokinet 8:422-446.

McDonnell CG, Malkan D, Van Pelt FD and Shorten GD (2003) Elimination of alfentanil delivered

by infusion is not altered by the chronic administration of atorvastatin. Eur J Anaesthesiol

20:662-667.

McDonnell TE, Bartkowski RR, Bonilla BS, Henthorn TK and Williams JJ (1982) Nonuniformity of

Alfentanil pharmacokinetics in healthy adults. Anesthesiology 57:A236.

Meistelman C, Saint-Maurice C, Lepaul M, Levron JC, Loose JP and Mac Gee K (1987) A

comparison of alfentanil pharmacokinetics in children and adults. Anesthesiology 66:13-16. Mertens MJ, Vuyk J, Olofsen E, Bovill JG and Burm AG (2001) Propofol alters the pharmacokinetics

of alfentanil in healthy male volunteers. Anesthesiology 94:949-957.

Meuldermans W, Van Peer A, Hendrickx J, Woestenborghs R, Lauwers W, Heykants J, Vanden

Bussche G, Van Craeyvelt H and Van der Aa P (1988) Alfentanil pharmacokinetics and

metabolism in humans. Anesthesiology 69:527-534.

Meuldermans WE, Hurkmans RM and Heykants JJ (1982) Plasma protein binding and distribution of

fentanyl, sufentanil, alfentanil and lofentanil in blood. Arch Int Pharmacodyn Ther 257:4-19.

Mihaly GW, Ching MS, Klejn MB, Paull J and Smallwood RA (1987) Differences in the binding of

quinine and quinidine to plasma proteins. Br J Clin Pharmacol 24:769-774.

Moschitto LJ and Greenblatt DJ (1983) Concentration-independent plasma protein binding of

benzodiazepines. J Pharm Pharmacol 35:179-180.

Nakajima M, Tane K, Nakamura S, Shimada N, Yamazaki H and Yokoi T (2002) Evaluation of

approach to predict the contribution of multiple cytochrome P450s in drug metabolism using

relative activity factor: effects of the differences in expression levels of NADPH-cytochrome

P450 reductase and cytochrome b(5) in the expression system and the differences in the

marker activities. J Pharm Sci 91:952-963.

Ngui JS, Tang W, Stearns RA, Shou M, Miller RR, Zhang Y, Lin JH and Baillie TA (2000)

Cytochrome P450 3A4-mediated interaction of diclofenac and quinidine. Drug Metab Dispos

28:1043-1050.

Nielsen TL, Rasmussen BB, Flinois JP, Beaune P and Brosen K (1999) In vitro metabolism of

quinidine: the (3S)-3-hydroxylation of quinidine is a specific marker reaction for cytochrome

P-4503A4 activity in human liver microsomes. J Pharmacol Exp Ther 289:31-37.

Nilsen OG, Leren P, Aakesson I and Jacobsen S (1978) Binding of quinidine in sera with different

levels of triglycerides, cholesterol, and orosomucoid protein. Biochem Pharmacol 27:871-

876.

Obach RS (1999) Prediction of human clearance of twenty-nine drugs from hepatic microsomal

intrinsic clearance data: An examination of in vitro half-life approach and nonspecific binding

to microsomes. Drug Metab Dispos 27:1350-1359. Ochs HR, Greenblatt DJ, Labedzki L and Smith RB (1986) Alprazolam kinetics in patients with renal

insufficiency. J Clin Psychopharmacol 6:292-294.

Ochs HR, Greenblatt DJ, Woo E, Franke K and Smith TW (1978a) Effect of propranolol on

pharmacokinetics and acute electrocardiographic changes following intravenous quinidine in

humans. Pharmacology 17:301-306.

Ochs HR, Greenblatt DJ, Woo E and Smith TW (1978b) Reduced quinidine clearance in elderly

persons. Am J Cardiol 42:481-485.

Ochs HR, Grube E, Greenblatt DJ, Woo E and Bodem G (1980) Intravenous quinidine:

pharmacokinetic properties and effects on left ventricular performance in humans. Am Heart J

99:468-475.

Odar-Cederlof I and Borga O (1974) Kinetics of diphenylhydantoin in uraemic patients: consequences

of decreased plasma protein binding. Eur J Clin Pharmacol 7:31-37.

Ohno Y, Hisaka A and Suzuki H (2007) General framework for the quantitative prediction of

CYP3A4-mediated oral drug interactions based on the AUC increase by coadministration of

standard drugs. Clin Pharmacokinet 46:681-696.

Olling M, Mensinga TT, Barends DM, Groen C, Lake OA and Meulenbelt J (1999) Bioavailability of

carbamazepine from four different products and the occurrence of side effects. Biopharm

Drug Dispos 20:19-28.

Olubodun JO, Ochs HR, von Moltke LL, Roubenoff R, Hesse LM, Harmatz JS, Shader RI and

Greenblatt DJ (2003) Pharmacokinetic properties of zolpidem in elderly and young adults:

possible modulation by testosterone in men. Br J Clin Pharmacol 56:297-304.

Oscarson M, Zanger UM, Rifki OF, Klein K, Eichelbaum M and Meyer UA (2006) Transcriptional

profiling of genes induced in the livers of patients treated with carbamazepine. Clin

Pharmacol Ther 80:440-456.

Pade V and Stavchansky S (1998) Link between drug absorption solubility and permeability

measurements in Caco-2 cells. J Pharm Sci 87:1604-1607. Patat A, Trocherie S, Thebault JJ, Rosenzweig P, Dubruc C, Bianchetti G, Court LA and Morselli PL

(1994) EEG profile of intravenous zolpidem in healthy volunteers. Psychopharmacology

(Berl) 114:138-146.

Pearce RE, Vakkalagadda GR and Leeder JS (2002) Pathways of carbamazepine bioactivation in vitro

I. Characterization of human cytochromes P450 responsible for the formation of 2- and 3-

hydroxylated metabolites. Drug Metab Dispos 30:1170-1179.

Perez-Mateo M and Erill S (1977) Protein binding of salicylate and quinidine in plasma from patients

with renal failure, chronic liver disease and chronic respiratory insufficiency. Eur J Clin

Pharmacol 11:225-231.

Perucca E, Makki K and Richens A (1978) Is phenytoin metabolism dose-dependent by enzyme

saturation or by feedback inhibition? Clin Pharmacol Ther 24:46-51.

Phimmasone S and Kharasch ED (2001) A pilot evaluation of alfentanil-induced miosis as a

noninvasive probe for hepatic cytochrome P450 3A4 (CYP3A4) activity in humans. Clin

Pharmacol Ther 70:505-517.

Pisani F, Fazio A, Artesi C, Oteri G, Spina E, Tomson T and Perucca E (1992) Impairment of

carbamazepine-10, 11-epoxide elimination by valnoctamide, a valpromide isomer, in healthy

subjects. Br J Clin Pharmacol 34:85-87.

Pisani F, Fazio A, Oteri G, Spina E, Perucca E and Bertilsson L (1988) Effect of valpromide on the

pharmacokinetics of carbamazepine-10, 11-epoxide. Br J Clin Pharmacol 25:611-613.

Polli JW, Wring SA, Humphreys JE, Huang L, Morgan JB, Webster LO and Serabjit-Singh CS (2001)

Rational use of in vitro P-glycoprotein assays in drug discovery. J Pharmacol Exp Ther

299:620-628.

Poulin P and Theil FP (2002) Prediction of pharmacokinetics prior to in vivo studies. 1. Mechanism-

based prediction of volume of distribution. J Pharm Sci 91:129-156.

Raemsch KD and Sommer J (1983) Pharmacokinetics and metabolism of nifedipine. Hypertension

5:II18-24. Ramsay RE, McManus DQ, Guterman A, Briggle TV, Vazquez D, Perchalski R, Yost RA and Wong

P (1990) Carbamazepine metabolism in humans: effect of concurrent anticonvulsant therapy.

Ther Drug Monit 12:235-241.

Rawlins MD, Collste P, Bertilsson L and Palmer L (1975) Distribution and elimination kinetics of

carbamazepine in man. Eur J Clin Pharmacol 8:91-96.

Reitman ML, Chu X, Cai X, Yabut J, Venkatasubramanian R, Zajic S, Stone JA, Ding Y, Witter R,

Gibson C, Roupe K, Evers R, Wagner JA and Stoch A (2011) Rifampin's acute inhibitory and

chronic inductive drug interactions: experimental and model-based approaches to drug-drug

interaction trial design. Clin Pharmacol Ther 89:234-242.

Riad LE and Sawchuk RJ (1998) A partial area difference analysis for estimating elimination rate

constants and distribution volumes of metabolites. J Pharm Sci 87:769-773.

Riva R, Contin M, Albani F, Perucca E, Ambrosetto G, Gobbi G, Cortelli P, Procaccianti G and

Baruzzi A (1984) Free and total plasma concentrations of carbamazepine and carbamazepine-

10,11-epoxide in epileptic patients: diurnal fluctuations and relationship with side effects.

Ther Drug Monit 6:408-413.

Rodgers T and Rowland M (2007) Mechanistic approaches to volume of distribution predictions:

understanding the processes. Pharm Res 24:918-933.

Roure P, Jean N, Leclerc AC, Cabanel N, Levron JC and Duvaldestin P (1987) Pharmacokinetics of

alfentanil in children undergoing surgery. Br J Anaesth 59:1437-1440.

Rowland A, Elliot DJ, Knights KM, Mackenzie PI and Miners JO (2008) The "albumin effect" and in

vitro-in vivo extrapolation: sequestration of long-chain unsaturated fatty acids enhances

phenytoin hydroxylation by human liver microsomal and recombinant cytochrome P450 2C9.

Drug Metab Dispos 36:870-877.

Saari TI, Laine K, Leino K, Valtonen M, Neuvonen PJ and Olkkola KT (2006) Effect of voriconazole

on the pharmacokinetics and pharmacodynamics of intravenous and oral midazolam. Clin

Pharmacol Ther 79:362-370.

Sahi J, Shord SS, Lindley C, Ferguson S and LeCluyse EL (2009) Regulation of cytochrome P450

2C9 expression in primary cultures of human hepatocytes. J Biochem Mol Toxicol 23:43-58. Salva P and Costa J (1995) Clinical pharmacokinetics and pharmacodynamics of zolpidem.

Therapeutic implications. Clin Pharmacokinet 29:142-153.

Scavone JM, Greenblatt DJ, Locniskar A and Shader RI (1988) Alprazolam pharmacokinetics in

women on low-dose oral contraceptives. J Clin Pharmacol 28:454-457.

Schuttler J and Stoeckel H (1982) [Clinical pharmacokinetics of alfentanyl (author's transl)].

Anaesthesist 31:10-14.

Schwagmeier R, Alincic S and Striebel HW (1998) Midazolam pharmacokinetics following

intravenous and buccal administration. Br J Clin Pharmacol 46:203-206.

Scott JC and Stanski DR (1987) Decreased fentanyl and alfentanil dose requirements with age. A

simultaneous pharmacokinetic and pharmacodynamic evaluation. J Pharmacol Exp Ther

240:159-166.

Shou M, Hayashi M, Pan Y, Xu Y, Morrissey K, Xu L and Skiles GL (2008) Modeling, prediction,

and in vitro in vivo correlation of CYP3A4 induction. Drug Metab Dispos 36:2355-2370.

Smith RB, Kroboth PD, Vanderlugt JT, Phillips JP and Juhl RP (1984) Pharmacokinetics and

pharmacodynamics of alprazolam after oral and IV administration. Psychopharmacology

(Berl) 84:452-456.

Soars MG, Grime K and Riley RJ (2006) Comparative analysis of substrate and inhibitor interactions

with CYP3A4 and CYP3A5. Xenobiotica 36:287-299.

Spina E, Tomson T, Svensson JO, Faigle JW and Bertilsson L (1988) Single-dose kinetics of an

enteric-coated formulation of carbamazepine-10,11-epoxide, an active metabolite of

carbamazepine. Ther Drug Monit 10:382-385.

Staines AG, Coughtrie MW and Burchell B (2004) N-glucuronidation of carbamazepine in human

tissues is mediated by UGT2B7. J Pharmacol Exp Ther 311:1131-1137.

Stoehr GP, Kroboth PD, Juhl RP, Wender DB, Phillips JP and Smith RB (1984) Effect of oral

contraceptives on triazolam, temazepam, alprazolam, and lorazepam kinetics. Clin Pharmacol

Ther 36:683-690. Taillardat-Bertschinger A, Martinet CA, Carrupt PA, Reist M, Caron G, Fruttero R and Testa B

(2002) Molecular factors influencing retention on immobilized artifical membranes (IAM)

compared to partitioning in liposomes and n-octanol. Pharm Res 19:729-737.

Tassaneeyakul W, Veronese ME, Birkett DJ, Doecke CJ, McManus ME, Sansom LN and Miners JO

(1992) Co-regulation of phenytoin and tolbutamide metabolism in humans. Br J Clin

Pharmacol 34:494-498.

Tchaparian E, Tang L, Xu G, Huang T and Jin L (2008) Cell Based Experimental Models as Tools for

the Prediction of Human Intestinal Absorption, in: 15th North American ISSX Meeting San

Diego.

Thummel KE, O'Shea D, Paine MF, Shen DD, Kunze KL, Perkins JD and Wilkinson GR (1996) Oral

first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-

mediated metabolism. Clin Pharmacol Ther 59:491-502.

Tomson T, Tybring G and Bertilsson L (1983) Single-dose kinetics and metabolism of

carbamazepine-10,11-epoxide. Clin Pharmacol Ther 33:58-65.

Tran JQ, Kovacs SJ, McIntosh TS, Davis HM and Martin DE (1999) Morning spot and 24-hour

urinary 6 beta-hydroxycortisol to cortisol ratios: intraindividual variability and correlation

under basal conditions and conditions of CYP 3A4 induction. J Clin Pharmacol 39:487-494.

Varma MV, Gardner I, Steyn SJ, Nkansah P, Rotter CJ, Whitney-Pickett C, Zhang H, Di L, Cram M,

Fenner KS and El-Kattan AF (2012) pH-Dependent solubility and permeability criteria for

provisional biopharmaceutics classification (BCS and BDDCS) in early drug discovery. Mol

Pharm 9:1199-1212.

Vickers S, Duncan CA, Chen IW, Rosegay A and Duggan DE (1990) Metabolic disposition studies

on simvastatin, a cholesterol-lowering prodrug. Drug Metab Dispos 18:138-145.

Volgyi G, Deak K, Vamos J, Valko K and Takacs-Novak K (2008) RPTLC determination of log P of

structurally diverse neutral compounds. JPlanar Chromatogr 21:143-149.

Von Moltke LL, Greenblatt DJ, Granda BW, Duan SX, Grassi JM, Venkatakrishnan K, Harmatz JS

and Shader RI (1999) Zolpidem metabolism in vitro: responsible cytochromes, chemical

inhibitors, and in vivo correlations. Br J Clin Pharmacol 48:89-97. von Richter O, Glavinas H, Krajcsi P, Liehner S, Siewert B and Zech K (2009) A novel screening

strategy to identify ABCB1 substrates and inhibitors. Naunyn Schmiedebergs Arch

Pharmacol 379:11-26.

Vozeh S, Uematsu T, Guentert TW, Ha HR and Follath F (1985) Kinetics and electrocardiographic

changes after oral 3-OH-quinidine in healthy subjects. Clin Pharmacol Ther 37:575-581.

Wallin A and Herngren L (1985) Distribution of phenobarbital in whole blood during pregnancy and

perinatally--an in vitro study. Eur J Clin Pharmacol 29:187-191.

Walsky RL and Obach RS (2004) Validated assays for human cytochrome P450 activities. Drug

Metab Dispos 32:647-660.

Wan H, Ahman M and Holmen AG (2009) Relationship between brain tissue partitioning and

microemulsion retention factors of CNS drugs. J Med Chem 52:1693-1700.

Wandel C, Witte JS, Hall JM, Stein CM, Wood AJ and Wilkinson GR (2000) CYP3A activity in

African American and European American men: population differences and functional effect

of the CYP3A4*1B5'-promoter region polymorphism. Clin Pharmacol Ther 68:82-91.

Weaver R, Graham KS, Beattie IG and Riley RJ (2003) Cytochrome P450 inhibition using

recombinant proteins and mass spectrometry/multiple reaction monitoring technology in a

cassette incubation. Drug Metab Dispos 31:955-966.

Wilensky AJ, Friel PN, Levy RH, Comfort CP and Kaluzny SP (1982) Kinetics of phenobarbital in

normal subjects and epileptic patients. Eur J Clin Pharmacol 23:87-92.

Williams JA, Ring BJ, Cantrell VE, Jones DR, Eckstein J, Ruterbories K, Hamman MA, Hall SD and

Wrighton SA (2002) Comparative metabolic capabilities of CYP3A4, CYP3A5, and

CYP3A7. Drug Metab Dispos 30:883-891.

Winiwarter S, Bonham NM, Ax F, Hallberg A, Lennernas H and Karlen A (1998) Correlation of

human jejunal permeability (in vivo) of drugs with experimentally and theoretically derived

parameters. A multivariate data analysis approach. J Med Chem 41:4939-4949.

Wong KS, Kenseth J and Strasburg R (2004) Validation and long-term assessment of an approach for

the high throughput determination of lipophilicity (log POW) values using multiplexed,

absorbance-based capillary electrophoresis. J Pharm Sci 93:916-931. Wong SL, Locke C, Staser J and Granneman GR (1998) Lack of multiple dosing effect of sertindole

on the pharmacokinetics of alprazolam in healthy volunteers. Psychopharmacology (Berl)

135:236-241.

Wong YY, Ludden TM and Bell RD (1983) Effect of erythromycin on carbamazepine kinetics. Clin

Pharmacol Ther 33:460-464.

Woo E and Greenblatt DJ (1979) Pharmacokinetic and clinical implications of quinidine protein

binding. J Pharm Sci 68:466-470.

Xu Y, Zhou Y, Hayashi M, Shou M and Skiles GL (2011) Simulation of clinical drug-drug

interactions from hepatocyte CYP3A4 induction data and its potential utility in trial designs.

Drug Metab Dispos 39:1139-1148.

Zhu C, Jiang L, Chen TM and Hwang KK (2002) A comparative study of artificial membrane

permeability assay for high throughput profiling of drug absorption potential. Eur J Med

Chem 37:399-407.