Acta Pharmaceutica Sinica B 2016;6(5):430–440

Chinese Pharmaceutical Association

Institute of Materia Medica, Chinese Academy of Medical Sciences

Acta Pharmaceutica Sinica B

www.elsevier.com/locate/apsb
www.sciencedirect.com

REVIEW

PBPK modeling and simulation in drug research

and development
Xiaomei Zhuanga, Chuang Lub,n

a
State Key Laboratory of Toxicology and Medical Countermeasures, Beijing Institute of Pharmacology and Toxicology,
Beijing 100850, China
b
Department of DMPK, Biogen, Inc., Cambridge, MA 02142, USA

Received 18 March 2016; revised 25 April 2016; accepted 26 April 2016

KEY WORDS Abstract Physiologically based pharmacokinetic (PBPK) modeling and simulation can be used to
PBPK;
predict the pharmacokinetic behavior of drugs in humans using preclinical data. It can also explore the
PK prediction; effects of various physiologic parameters such as age, ethnicity, or disease status on human
Absorption; pharmacokinetics, as well as guide dose and dose regiment selection and aid drug–drug interaction risk
Metabolism; assessment. PBPK modeling has developed rapidly in the last decade within both the field of academia
Drug–drug interaction; and the pharmaceutical industry, and has become an integral tool in drug discovery and development. In
Special population this mini-review, the concept and methodology of PBPK modeling are briefly introduced. Several case
studies were discussed on how PBPK modeling and simulation can be utilized through various stages of
drug discovery and development. These case studies are from our own work and the literature for better
understanding of the absorption, distribution, metabolism and excretion (ADME) of a drug candidate, and
the applications to increase efficiency, reduce the need for animal studies, and perhaps to replace clinical
trials. The regulatory acceptance and industrial practices around PBPK modeling and simulation is also
discussed.

& 2016 Chinese Pharmaceutical Association and Institute of Materia Medica, Chinese Academy of Medical
Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

n
Corresponding author. Tel.: þ1 6176793365.
E-mail address: chuang.lu@biogen.com (Chuang Lu).
Peer review under responsibility of Institute of Materia Medica, Chinese Academy of Medical Sciences and Chinese Pharmaceutical Association.

http://dx.doi.org/10.1016/j.apsb.2016.04.004
2211-3835 & 2016 Chinese Pharmaceutical Association and Institute of Materia Medica, Chinese Academy of Medical Sciences. Production and hosting by
Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
PBPK modeling and simulation in drug research 431

1. Introduction Non-Eliminating tissues:

VT  dCT=dt ¼ QT  CA QT  CV T ð1Þ
The concept of physiologically based pharmacokinetic (PBPK)
models was first introduced by Teorell in 19371. For several where Q is blood flow (L/h), C is concentration (mg/L), V is
decades, growing efforts have been made to refine PBPK models volume (L); and T represents tissues, A represents arterial, V
so that they can be applied in drug development2. Thanks to the represents venous.
advancement in computing power and increasing access to

CV T ¼ CT = K p =B:P ð2Þ
preclinical data, especially in vitro data, on absorption, distribu-
tion, metabolism and excretion (ADME). PBPK modeling and where Kp is tissue to plasma partition coefficient of the compound
simulation currently receives extensive attention during drug and B:P is the ratio of blood to plasma.
discovery and development3,4, and in submissions for regulatory Eliminating tissues:
filing and reviews5,6. As a unique tool, PBPK models can be used
VT  dCT=dt ¼ QT  CA QT  CV T  CLint  CV u T ð3Þ
to estimate the pharmacokinetic (PK) profile of a compound based
on its preclinical ADME data and can be used to assess the where CLint is the intrinsic clearance of the compound (L/h), and u
exposure in a target organ after the administration of a drug by is unbound.
taking into account the rate of absorption and disposition in that Different from the conventional PK models, PBPK model is
organ, as well as metabolism within that organ if it is applicable. composed of two main parts—an anatomical “backbone” which
Based on the PK data generated from one dose schedule, the contains species specific physiological parameters that are inde-
PBPK model can be used to evaluate the PK profile of different pendent of the drug and hence can be applied to any compounds,
dose schedules and/or dose routes. Based on the PK data from one and a drug-specific part which consists of the individual drug's
ethnic population, the PBPK model can be used to predict the PK ADME properties applied to the relevant processes within each
profile in different ethnic populations as well as populations of tissue compartment. Parameters for incorporating into PBPK
various age and disease stages. This mini-review describes the models are either drug-dependent (e.g., binding to blood, fub;
PBPK methodology used in drug discovery and development and tissue-to-plasma distribution coefficient, KPT; tissue permeability–
specific examples of its application together with the regulatory surface area product, PST; enzymatic activity, Vmax/Km) or drug-
acceptance and industrial common applications. independent (e.g. blood flows, QT; tissue volumes, VT; tissue
composition). The accuracy of the PBPK prediction of ADME
parameters by the model not only depends on the present knowl-
2. PBPK methodology edge of animal or human physiology, but also on the physio-
chemical and biochemical properties of the test compounds.
PBPK models are made up of compartments corresponding to the
different physiological organs of the body, linked by the circulat-
ing blood system. Each compartment is exactly described by a 3. The key points in PBPK model construction
tissue volume and blood flow rate that is specific to the species of
interest. Each tissue is defined with assumptions of either 3.1. Acquisition of drug dependent parameters
perfusion-rate-limited or permeability-rate-limited. Perfusion-rate-
limited kinetics tends to exist for small lipophilic molecules where PBPK modeling is a bottom-up approach that integrates a large
the blood flow to the tissue proved to be the limiting process of the number of drug specific data, parameters on species physiology
absorption. Permeability-rate-limited kinetics occurs for more (system data), and a good understanding of all active processes
hydrophilic and larger molecules where the permeability across affecting the pharmacokinetic properties of a drug. System-
the cell membrane becomes the limiting process of absorption7. dependent parameters (e.g., tissue volume, blood flow, glomerular
Drug is disposed via the exile blood flow after being metabolized filtration rate, amount of microsomal protein/hepatocytes per gram
in the organ, if applicable. of liver, plasma protein, enzyme, and transporter abundance) for
A schematic of a PBPK model is shown in Fig. 1. The mass human and preclinical species are available in the literature and
balance differential equations used in these models have been have been compiled in the commercial PBPK platforms, including
described previously8 and follow the principles shown below. GastroPlus (www.simulations-plus.com), PKSIM (www.system
sbiology.com/products/pk-sim.html), Simcyp (www.simcyp.com),
ADMEWORKS DDI Simulator (http://www.fqs.pl/chemistry_
materials_life_science/products/ddi_simulator), CLOEPK (http://
www.cyprotex.com/insilico/), and many other modeling software.
For example, PBPK models in humans specify ethnic population
(specific system parameters) account for variability (standard
deviation or range) and the covariation between these parameters
in that ethnic population. Drug dependent parameters include
physicochemical properties (molecular weight, pKa, basic or acidic
nature of the drug), solubility (logD) and permeability, blood cell
and plasma protein binding (e.g. fraction unbound in plasma (fu,p),
blood plasma partitioning [B:P]), transporter contribution to drug
disposition, and in vitro data on the metabolism by hepatic or ex-
hepatic enzymes (e.g., intrinsic clearance (CLint)). A lack of
sufficient in vitro and in vivo data may hamper the use of this
Figure 1 Schematic of a PBPK model. approach. These compound-specific parameters are often deter
 432 Xiaomei Zhuang, Chuang Lu

Table 1 Data requirement for building a PBPK model in Simcyps.

Parameter Unit convert to In vitro test system

Molecular weight g/mol Physicochemistry property measurement, less

prefer an in silico prediction
logP Octanol:water partition coefficient
pKa (s) Physicochemistry property measurement, less
prefer to use an in silico prediction
Compound type Base, acid, neutral Based on the chemical structure or pH-
dependent solubility test
pH-dependent solubility mg/mL Measured in buffer with different pH
Plasma protein binding fu In vitro in human plasma (pay attention to
whether compound binds to AGP)
Blood–plasma partitioning B:P In vitro in human blood
Apparent permeability 10  6 cm/s Caco-2 , MDCK
Intrinsic clearance in microsomes, or S9, or mL/min/mg for microsomes and S9, In vitro assay, or use in vivo clearance if
hepatocytes, or rhCYP uL/min/million cells for hepatocytes, available
uL/min/pmol for rhCYP
Protein concentration in in vitro test mg/mL In vitro assay for intrinsic clearance
In vitro test matrix binding fu Measure the free fraction using the same
protein concentration in the in vitro test
system
Vmax and Km (if study for saturable PK; study pmol/min/mg, mmol/L The same in vitro system where intrinsic
metabolic-mediated DDI as a victim) clearance was determined
Percent of enzyme (e.g. CYP) contribution to the fm In vitro reaction phenotyping
metabolism (study DDI as a victim; study
metabolic-mediated DDI as a perpetrator)
Reversible inhibition, IC50 mmol/L Human liver microsomes or suitable in vitro
system
Mechanism-based CYP inhibition, kinact, KI h–1, mmol/L
CYP Induction, Jmax, EC50 fold induction, mmol/L Human hepatocytes with positive controls in
3 donors

Some transporter data can be incorporated; when clinical data become available, CL, Vss, fa, Ka, etc., can be incorporated to refine the initial model.

mined using in vitro assays or sometimes in silico models. Table 1 prospectively to simulate unstudied scenarios and, when
lists the input data required for building a basic PBPK model, in appropriate, these predictions can be incorporated into regula-
this case, for Simcyp simulation. tory submissions, product labels, additional post-approval
studies, and next generation follow-on drugs10.

3.2. Combination of “bottom-up” and “middle-out” methods to During this stage, mismatches between simulation and
create and refine a PBPK model observation may frequently occur and parameter sensitivity
analysis is critical to identify the inputs that have the most
PBPK modeling typically uses a “bottom-up” approach and is influence on a simulated profile. The selection of which
initially constructed based on preclinical data during the early drug parameters to focus upon for the parameter sensitivity analysis
discovery stage. Compound-specific parameters generated using requires a good understanding of the nature of each input data,
in vitro models are used to predict in vivo PK profile in preclinical as well as how they may impact the simulated profile. It is also
species and humans. For example, a “bottom-up” methodology for important to have an understanding of how the input data are
the clinical PK profile prediction proposed by Jones et al.9 is generated and the associated errors, and also an awareness of
described as following: the reasonable range of input values.

(1) Verification of intravenous disposition prediction in preclinical

species, for example, assessment of most appropriate Kp
prediction methodology taking into account method assump- 4. Applications of PBPK modeling during drug research and
tions, assessment of the prediction accuracy and the physico- development
chemical properties of the particular compound;
(2) Verification of oral absorption prediction in preclinical species PBPK models are routinely applied from the early discovery stage,
over a range of doses to further assess prediction accuracy; where there is limited data captured for any compound of interest,
(3) Simulation of disposition and absorption in humans—using to late drug development, where large amounts of data are
appropriate CL and Kp prediction methods selected based on the available9. PBPK modeling can be categorized into three major
preclinical verification step. Once preclinical or clinical in vivo roles, that can be used to inform regulatory communications, that
data are available, the mechanistic PBPK models can be further have impacted clinical development decisions and that promote the
refined and updated (“middle-out” approach) and applied mechanistic understanding of clinical observations11.
PBPK modeling and simulation in drug research 433

5. Lead optimization or candidate evaluation, a case study rat pharmacology study. A bioavailability of 16.9% was predicted in
humans. However, after decreasing the oral dose from 287 mg to
Unlike late development stages where PK data from animal can be 57.4 mg (the low end of the projected human efficacious dose), the
used to refine the PBPK model built on in vitro data, in drug predicted bioavailability increased from 16.9% to 35.1%, whereas
discovery, these processes mainly rely on the use of physico- no change in elimination parameters such as t1/2 was observed,
chemical properties, in vitro data, and increasingly in silico data. suggesting that solubility did play a role in the absorption of YQA-
This example illustrates the use of a PBPK absorption model 14 in humans. These acceptable PK properties make YQA-14 an
(GastroPlus v. 8.5) in the prediction of human oral bioavailability improved candidate for further development as a potential dopamine
from preclinical studies for a candidate compound. YQA-14 is a D3R antagonism for the treatment of drug addiction in clinic.
novel and selective dopamine D3 receptor antagonist, with the
potential to treat drug addiction. Earlier compounds in its structural
class tend to have poor oral bioavailability in humans due to the 6. Drug–drug interaction (DDI) potential prediction, a case
pronounced metabolism from aldehyde oxidase (AO). The aim of study
this study was to simulate the clinical pharmacokinetic behavior of
YQA-14 using a PBPK model to assess the likelihood of This example demonstrates the potential of using PBPK modeling in
developing YQA-14 as a clinical candidate12. YQA-14 is a the prediction of DDI risk13. Naturally occurring furanocoumarin
lipophilic and basic compound with three pKa values (6.91, compounds psoralen (PRN) and isopsoralen (IPRN) are bioactive
9.30, and 10.91) and a logD7.4 value of 2.15. At pH 6.5, the constituents in herbaceous plants. They are widely used as active
solubility of YQA-14 was 0.004 mg/mL. It was stable in human ingredients in many Chinese herbal medicines. Both PRN and IPRN
liver cytosolic fractions (less AO metabolism liability compared to showed potent reversible inhibition of CYP1A2 in human liver
the previous candidates), and the liver microsomal clearances and microsomes (HLMs). In addition, time-dependent inhibition of
in vivo clearances were moderate in rats, dogs (in vitro and in vivo) CYP1A2 was observed with IPRN but not PRN. In an attempt to
and humans (in vitro only). It also had moderate bioavailability in assess the potential DDI risk, Simcyp simulations were conducted to
preclinical species. For human PK prediction, a “bottom-up” full predict phenacetin (a CYP1A2 substrate) AUC changes under the
PBPK model was first built by inputting the main parameters co-administration of PRN or IPRN by allowing perpetrator and
obtained from in vitro studies (Table 2). This model was then victim dosed at the same time once a day for 10 days. A reduced
validated and modified by in vivo PK profiles of rats and dogs PBPK model was built using the basic physicochemical data listed in
(Fig. 2). After oral administration, YQA-14 was rapidly absorbed Table 314. Simulations were performed in healthy subjects (n¼ 100,
in preclinical species with a Tmax around 0.5–1 h; this is consistent 50% men, aged 40–65 years) by using a Simcyp population-based
with the high permeability obtained from the Caco-2 assessment. simulator (version 11, Simcyp Ltd., Sheffield, UK). A population of
Oral bioavailability in rats and dogs were 15.6% and 45.9%, smokers was constructed by modifying the CYP1A2 abundance in
respectively. Because rats have a higher hepatic clearance, both healthy subjects from 52 to 94 pmol/mg microsomal protein to
in vitro and in vivo, then might have a higher pre-system mimic individuals who smoke 20 cigarettes per day14. The Simcyp
metabolism. Poor solubility could be another reason for the lower default phenacetin profile was used without further modification. The
bioavailability in rats because a higher dose was given to rats maximum allowed daily doses of PRN and IPRN (60 mg, Chinese
compared to dogs. After the preclinical model was validated, Pharmacopoeia Commission, 2010) were used to predict the worst-
physicochemical properties, models/modules used to predict tissue case scenario of DDI. Fig. 3 presents the 10-day simulations of
distribution, compound dissolution and precipitation information, plasma concentration–time profiles of a 1500 mg daily dose of
combined with respective in vitro human data (clearance, plasma phenacetin with a 60 mg daily dose of PRN or IPRN. The results
protein and microsomal binding, and RBC partitioning) were showed that PRN increased the AUC of phenacetin by 1.71-fold and
utilized to simulate human plasma concentration vs. time profiles 2.12-fold in healthy volunteers and smokers, respectively, whereas
of YQA-14 at 287 mg QD, a therapeutic dose extrapolated from the IPRN increased the AUC of phenacetin by 3.24-fold and 5.01-fold in
healthy volunteers and smokers, respectively. It is worth noting that
Table 2 Input data used in the GastroPlus™ PBPK model. in this simulation, the smoker population has lower basal AUC
because of their high CYP1A2 activity. Co-administration of the
Parameter Value moderate reversible inhibitor PRN was not able to bring the AUC
back to the level of a healthy volunteer, suggesting an incomplete
Molecular weight (g/mol) 442.95
balance of the higher CYP1A2 activity induced by the smoke.
pKa 6.91, 9.30, 10.91
logD at pH 7.4 2.15
However, when a more potent inhibition IPRN was applied (both
Caco-2 permeability (10–6 cm/s) 19.90, 22.13 reversible and time-dependent CYP1A2 inhibitor), the AUC in the
(propranolol, control) smoke population was comparable to the level of healthy volunteer
Aqueous solubility at pH 6.5 (mg/mL) 0.004 suggesting that the inhibition and inactivation effects by IPRN
Rbp in rat, dog, human 0.70, 0.67, 0.65 balanced off the CYP1A2 activity induced by smoke. On the other
%Fu in rat, dog, human plasma and 1.29, 1.05, 0.96, 64 hand, the change of clearance and AUC were more profound when
human liver microsomes the co-administration IPRN or PRN in smoke population.
CLint in rat, dog and human liver 57.60, 6.42, 13.46
microsomes (mL/min/kg)
In vitro predicted hepatic clearance in rat, 31.60, 5.49, 8.05
7. Human PK and DDI prediction to avoid clinical DDI
dog and human (mL/min/kg)
In vivo clearance, rat, dog (mL/min/kg) 29.7, 8.3
trials, a case study
Dose (rat, dog, and human, mg/kg, QD) 25, 5, 4.1
Orteronel (TAK-700) is an oral, nonsteroidal, reversible, selective
This table is adapted from Ref. 12 with permission. 17,20-lyase inhibitor that was in development for the treatment of
434 Xiaomei Zhuang, Chuang Lu

Figure 2 Observed (□) and PBPK model–simulated (-) plasma concentration–time profiles of YQA-14 in rats (A and B) and dogs (C and D) after
a single i.v. (A and C) or (p.o.) (B and D) administration. Observed plasma concentration–time profiles (OBS) were obtained for rats and dogs after
single i.v. and p.o. administration of YQA-14 at 25 and 5 mg/kg, respectively (n¼ 3 rats/group; n¼ 4 dogs/group). This figure is adapted from Ref.
12 with permission.

patients with metastatic castration-resistant prostate cancer. In vitro with AUC changes all less than 1.25-fold, with the criteria
CYP inhibition study in human liver microsomes showed that considered as no DDI by the FDA16.
orteronel is a moderate inhibitor in CYP1A2, 2C8, 2C9, and 2C19,
with IC50 values of 17.8, 27.7, 30.8 and 38.8 mmol/L, respectively.
However, it showed no inhibition in CYP2B6, 2D6 or 3A4/5 8. Dose guidance for renal impairment patients, a case study
(IC504100 mmol/L, Table 4)15. The Cmax of orteronel in patients
who had consumed a high-fat meal was at average of 9.18 mmol/L This case illustrates how PBPK modeling can inform appropriate
and thus, the [I]/IC50 ratio calculated using a basic static model dosing of renal impairment (RI) patients in phase I/III studies and
showed that orteronel could cause as high as 1.84-fold of DDI. thereby enable characterization of safety and efficacy in the RI
Following the FDA DDI guidance16, if basic static models show patients during the late stage of drug development17. Data from
that a perpetrator has the potential of causing DDI (i.e. human ADME study revealed that orteronel (see last example) is a
[I]/IC5040.1), following up DDI assessment using a PBPK model drug that is primarily cleared by kidney excretion. The extent of
under the dynamic conditions with both substrate and inhibitor is orteronel biotransformation is minimal, with cytochrome P450
recommended. A PBPK model was then built with physicochem- isozymes having only a minor role. Thus, patients with renal
ical and preclinical data and oral clearance from a human phase I impairment may have increased exposure to orteronel because of
trial because in vitro metabolic clearance does not reflect the total their impaired urinary excretion capability. A PBPK model was
body clearance (Table 5). The resulting model well described the built as described in the last case study. The predicted PK profile
observed clinical PK (Fig. 4). This model was then used to was then validated using clinical PK data before applying the
simulate DDI potential with a set of sensitive CYP probe model to simulate PK profile of orteronel in moderate (glomerular
substrates, theophylline, repaglinide, (S)-warfarin, and omeprazole filtration rate (GFR), 30–60 mL/min) or severe (GFR, o30 mL/min)
for CYP1A2, 2C8, 2C9, and 2C19, respectively (built in com- RI patients. By comparing the PBPK model outputs with the
pound profiles within the Simcyp software, no further modification population PK analysis results from phase 2 trials, it was
was made). The DDI potential of orteronel toward these 4 CYPs at demonstrated that PBPK modeling can accurately predict the
the dynamic concentration scenario was simulated. As shown in effect of moderate and severe RI on the PK profile of orteronel
Table 6, orteronel would not cause DDI with any of the 4 CYPs (i.e. fold increase in AUC). In this model, exposure to orteronel
PBPK modeling and simulation in drug research 435

Table 3 Input data of PRN and IPRN for Simcyps simulation.

Parameter PRN IPRN

Molecular weight 186.17 186.17

logD7.4 1.63 1.32
Blood–plasma partition co-efficient (B/P) 0.82 0.65
Plasma protein binding (fu) 0.283 0.126
Microsomal protein binding at 0.5 mg/mL (fu) 0.745 0.906
Apparent permeability value: Papp (10–6 cm/s) Caco-2 51.6 44.6
(calibration compound atenolol Papp ¼ 1.40  10  6 cm/s)
Microsomal clearance (μL/min/mg) 14.5 8.0
CYP1A2 IC50 (μmol/L) 0.26 0.22
CYP1A2 KI (μmol/L) 0.40
CYP1A2 kinact (min-1) 0.05

For reversible inhibition, Ki were estimated using IC50/2;

Both compounds are in neutral condition under physiological pH, thus pKa was not available;
1400 mg phenacetin QD  10 and 60 mg PRN or IPRN QD  10 were applied;
This table is adapted from Ref. 14 with permission.

Figure 3 Simcyp simulation results of phenacetin AUC0–24 at 1400 mg daily  10 days in the presence of IPRN (60 mg daily  10 days) and
absence of IPRN in healthy subjects (A) and smokers (B), or the presence of PRN (60 mg daily  10 days) and absence of PRN in healthy subjects
(C) and smokers (D). The outer curves represent phenaceitn concentration in the presence of PRN or IPRN. This figure is adapted from Ref. 14
with permission. IPRN, isopsoralen; PRN, psoralen.

increased as a reversed function of the estimated proportion of may not require dose adjustments as they were predicted to have
orteronel cleared by the kidney, aligning with the degree of renal only a 20% higher exposure compared to the healthy subjects (a
impairment. The AUC for orteronel, when given at the clinical scenario not included in the PBPK modeling). Patients with severe
dose of 400 mg BID, was predicted to increase by 52% in patients RI given orteronel 200 mg BID were predicted to have similar
with moderate RI and 83% in patients with severe RI compared orteronel plasma concentrations as control subjects given 400 mg
with the healthy population group. Furthermore, the PBPK BID. In summary, this case demonstrates that for a drug being
simulation also predicted that a reduced dose of orteronel of eliminated primarily via renal route, the PBPK modeling approach
220 mg BID (or a rounded dose of 200 mg BID) would achieve can play a key role for guiding dose selection. This analysis helped
exposures in severely impaired subjects comparable to those seen the inclusion of patients with RI in phase III trials with appropriate
in subjects with normal renal function treated at 400 mg BID dose adjustment. That could serve as an alternative to a dedicated
(Fig. 5). Then a PopPK model was built to determine if dose RI study, or suggests that a reduced-size study in severe RI
adjustments might be required for renal RI patients in the clinical patients may be sufficient to assess the exposure risk in other RI
setting. Results of the PopPK suggested that patients with mild RI patients.
436 Xiaomei Zhuang, Chuang Lu

Table 4 Orteronel [I]/Ki values and predicted AUC ratio using static model.

Parameter CYP1A2 CYP2C8 CYP2C9 CYP2C19

Orteronel IC50 (μmol/L) 17.8 27.7 30.8 38.8

[I]/Ki 1.03 0.66 0.60 0.47
Substrate (fm) Theophylline (0.90) Repaglinide (0.64) (S)-warfarin (1.00) Omeprazole (0.87)
AUC ratio 1.84 1.34 1.60 1.39

Abbreviations: CYP, cytochrome P450; [I], inhibitor concentration that is the total plasma maximum concentration (Cmax); IC50, 50% inhibitory
concentration; Ki, inhibition dissociation constant.
Note: The mean Cmax in the subjects with the high-fat meal was 9.18 μmol/L. Ki ¼ IC50/2, assuming competitive inhibition. The fm was adapted from
Simcyps v 11. AUC ratio was calculated using the basic static equation: AUCR¼ 1/(fm/((1þ[I]/Ki)þ(1–fm))).
This table is adapted from Ref. 15 with permission.

Table 5 Orteronel input data for PBPK M&S.

Parameter Value

Compound type Monoprotic base

Molecular weight 307.35
logD7.4 1.322
pKa 6.600
Blood–plasma partition coefficient (B/P) 1.39
Plasma protein binding (fu) 0.403
Main binding protein HSA
Microsomal protein binding at 0.5 mg/mL (fu,mic) 0.961
fu (gut) 1
fa 0.86
Ka (L/h) 0.79
Qgut (L/h) 8.394
Apparent intrinsic permeability value: Papp (10–6 cm/s) Caco-2 9.05
Calibration compound (propranolol) value: Papp (10–6 cm/s) Caco-2 25.1
Clinical oral clearance (CL/F), (L/h) 16.9
Human ADME clearance routes (renal, hepatic, other) 53%, 28%, 19%
Clinical oral clearance, %CV 15.7
Clinical volume of distribution (Vd/F), (L/kg) 1.4
Clinical volume of distribution, %CV 30.2
CYP1A2 Ki (mmol/L)a 8.9
CYP2C8 Ki (mmol/L)a 13.8
CYP2C9 Ki (mmol/L)a 15.4
CYP2C19 Ki (mmol/L)a 19.4

Abbreviations: %CV, percent coefficient of variation; ADME, absorption, distribution, metabolism, excretion; fa,
fraction absorbed; fu, fraction unbound; fu (gut), apparent unbound fraction in enterocytes; HSA, human serum
albumin; IC50, 50% inhibitory concentration; Ka, first-order absorption rate constant; Ki, reversible inhibition constant;
logD7.4, logarithm of the octanol–water partition coefficient at pH 7.4; Papp, apparent passive permeability; pKa,
logarithmic acid dissociation constant; Qgut, hypothetical blood flow term that is used to indicate complex interplay
among passive intestinal permeability, active transport, enterocyte drug binding, blood flows to enterocytes, and gut
metabolism.
This table is adapted from Ref. 15 with permission.
a
All inhibition was assumed conservatively to be reversible; Ki values were calculated: IC50/2.

9. Bridge healthy adults to special populations simulation of pharmacokinetics in adult animal species was first
conducted. After a reasonable simulation of pharmacokinetics in
PBPK models can be utilized to extrapolate the drug pharmaco- the adult animal is achieved with a refinement, prediction of
kinetic behavior in healthy volunteer to patient populations that are human pharmacokinetics was performed using information cap-
a challenge to obtain PK profiles for, such as predicting doses and tured during the refinement of the animal model. For prediction of
drug exposures in children and infants18 and patients suffered from juvenile PK profile in humans, the same methodology is followed.
impaired renal or liver function11,15. The work of Parrott and First, a juvenile animal model is generated, which accounts for
colleagues19 exemplified the usage of a mechanistic PBPK model age-dependent differences that are known to impact the PK
in predicting the pharmacokinetics of a neuraminidase inhibitor behavior of the drug, then the model was verified by comparison
oseltamivir and its active metabolite oseltamivir carboxylate (OC) with the data obtained from a juvenile animal study. Finally, the
which are for the treatment and prophylaxis of influenza A and B prediction of juvenile humans was done using a PBPK model that
infections in infants and neonates. In their strategy (Fig. 6), the accounted for age dependency in humans and information gathered
PBPK modeling and simulation in drug research 437

Figure 4 Simulated and actual mean orteronel concentration-versus-

time curves. The line represents the simulated mean area under the
concentration-versus-time curve after a single dose of orteronel at
400 mg; the circles represent the actual data points from the high-fat
diet group (n ¼42) treated with a single dose of orteronel 400 mg. This
figure was adapted from Ref. 15 with permission.

Table 6 DDI analysis: simulated area under the concentra-

tion–time curve ratios for orteronel.

CYP/substrate Dose Orteronel IC50

(mmol/L)

CYP1A2/theophylline (SV) 125 mg TID 17.8

CYP2C8/repaglinide (SV) 0.25 mg BID 27.7
CYP2C9/(S)-warfarin (Sim) 10 mg QD 30.8 Figure 5 Physiologically based pharmacokinetic (PBPK) simulation
CYP2C19/omeprazole, 20 mg BID 38.8
of orteronel in (A) healthy subjects (observed and simulated values),
enteric-coated (SV)
subjects with moderate renal impairment (simulated values), and
Abbreviations: BID, twice daily; CYP, cytochrome P450; DDI, subjects with severe renal impairment (simulated values), and
drug–drug interaction; IC50, 50% inhibitory concentration; QD, (B) regression of orteronel clearance vs. glomerular filtration rate
once daily; Sim, profile based on in vitro data; SV, profile based on (GFR) based on PBPK simulations in healthy subjects, subjects with
in vivo data; TID, 3 times daily. moderate renal impairment, and subjects with severe renal impairment.
This table was adapted from Ref. 15 with permission. Observed data for healthy subjects (high-fat diet group, n ¼42) were
obtained from clinical study C21007. The clinical scenario assumed
100% bioavailability with all uncharacterized metabolism treated as
from adult human and juvenile animal studies for refinement. This
hepatic clearance (orteronel dose: 400 mg BID for 10 days). CL, total
provided the first insight of drug exposure in juveniles since a PK
clearance; RI, renal impairment. This figure was adapted from Ref. 17
study in this population is hard to come by.
with permission.

10. Regulatory submission primarily metabolized by CYP3A and 2D621,22. Among com-
pounds in the BCS classification, the PK profile of type I
PBPK modeling has been gaining acceptance at various regulatory compounds with high solubility and high permeability usually
bodies as part of submission package. Discussion of modeling and can be predicted quite well from their preclinical data. On the other
simulation approaches can be found in the updated DDI guidance hand, prediction of exposure changes due to CYP induction has
from both the European Medicines Agency (EMA)20 and the U.S. not been well validated, neither had the compounds of mix of CYP
Food and Drug Administration (FDA)16. Recently, The FDA inducer and time-dependent inhibitor. PK prediction involving
hosted a workshop at their White Oak Campus in Silver Spring, transporters is still not reliable due to poor understanding of the
MD, USA. At the workshop, the director of the Center of Drug scaling factors used to extrapolate in vitro data to in vivo
Evaluation and Research (CDER), Dr. Janet Woodcock, concluded disposition. The same is true for metabolism and disposition in
that “the modeling work performed thus far at CDER has the gut. Due to the complexity of the metabolism, absorption, and
contributed tremendously to overall drug development, in terms transporter activity involved at the different segments of the
of safety and efficacy, which ultimately result in patient benefits”6. gastrointestinal tract, and the unique nature of the of individual
Both FDA scientists and industrial and academic representatives compound, PBPK modeling for gut absorption (Fg) has yet to be
agreed that the current advance in PBPK modeling enable us to optimized22,23. Drug metabolizing enzyme-transporter interplay,
predict investigational drugs as a substrate of drug metabolizing PK prediction in organ impairment population, and allometry
enzyme with high confidence, especially when the drug is scaling down to children younger than 2 years of age (ontogeny
438 Xiaomei Zhuang, Chuang Lu

Figure 6 PBPK modeling strategy employed to predict exposure in neonates and infants. A stepwise approach is followed with verification
against in vivo data at each step. Simulations in juveniles are based on a model incorporating age dependencies in physiology and incorporating
data from relevant in vitro systems. Verification in juvenile animals allows for model refinement before prediction in children. This figure was
adapted from Ref. 19 with permission.

and maturation), are among the areas that still need more research. discourage further investment in that drug candidate if it is not a
There are much experiences lacked in the area of pregnancy, first-in-class drug candidate. For a similar reason, if a drug
obesity, and the geriatric population, as well as food effect, needs to be administrated multiple times a day, it may face
formulation, and pH effects. The prediction of intracellular challenge in marketing if it is neither a first-in-class nor a best-
concentration is also often a challenge24,25. in-class drug candidate.
As PBPK modeling advances, the FDA has seen an increase in 2) At the candidate selection stage of drug discovery, should a
modeling work in submission packages. Much of the modeling drug candidate be partially metabolized by polymorphic
work is cited in drug product labels to illustrate the degree or lack enzymes, such as CYP2D6 or 2C19, a PBPK model can be
of DDI risk with co-administration of market drugs5,6,22,26–30. applied to simulate the exposure in population including poor
From July 1, 2008 to December 31, 2013, there were 112 PBPK metabolizer to determine whether poor metabolizers need to be
packages submitted to the FDA including 5 run by the agency. excluded in the first-in-human (FIH) trials. The model can be
Among these packages, most of the studies (76/112) were DDI- further refined with the data from FIH healthy volunteer trials to
related. Most of the DDI simulation (45/76) had no clinical data help to design a DDI study (e.g., dose adjustment) in the poor
available for comparison, and of these eight studies were for metabolizer population.
perpetrators. Most of the DDI submissions were for CYP inhibi- 3) At the drug development stage, DDI risk simulation is the most
tion risk evaluation, only 8 cases were for CYP induction and popular application for PBPK modeling. Whether a drug
1 was for transporter inhibition. A few cases (3) had clinical data candidate is a substrate of drug metabolizing enzymes or a
available for building and optimizing the final models22. In most perpetrator, a DDI risk simulation with standard care medica-
situations, the PBPK models were included in the submission of tions can assess the risk of co-medication with these standard
IND or NDA. care market drugs. Oftentimes, the most potent perpetrator or
the most sensitive substrate is used in the initial simulations to
assess the worst case scenario. If enough safety margins are
11. Common industrial application for PBPK modeling presented at the worst case scenario, then clinical trials with
moderately sensitive substrates or perpetrator may get waivered.
In the pharmaceutical industry, PBPK modeling is used for 4) Many drugs are cleared via hepatic metabolism. Some may be
purposes, such as mechanistic studies, aiding internal drug excreted via renal excretion. Thus, exposure simulation in organ
discovery or clinical development decisions, and informing impairment patients is important to know. There are successful
regulatory communication including filing at various stages (e.g., examples of such a simulation described above and in the
IND and NDA). It is mostly applied at the development stage. literature11,15.
Below are a few outlines of its routine applications: 5) In DDI trial, dependent on the half-life of the substrate, the dose
frequency and dose duration of the inhibitor need to be
1) At the lead optimization stage of drug discovery, PBPK optimized to cover the duration of the exposure of the substrate
modeling can provide human PK prediction at clinical dose as much as possible. The more exposure overlap between the
and dose schedule. A high projected dose (e.g., 41 g/day) may substrate and the inhibitor, the better we can capture the DDI
PBPK modeling and simulation in drug research 439

perpetrators in patients is also not ethical and practical, the PBPK

modeling, in this case, can provide information about “what if” all
of those drugs are co-administered together. On the other hand, as
discussed earlier, PBPK modeling is a bottom-up approach, its
results dependent on the quality of the input data. Although
software are available for the prediction of physicochemical
properties of compounds, such as logP and pKa, in authors
experience, it is critical to use measured values to get a reliable
PBPK prediction, especially when predicting human PK profile,
rather than the AUC ratio for DDI purpose. For example, for a set
of clinical candidates (about 40 compounds), the number of
compounds for which the predicted PK profile within two fold
of observed clinical values dropped from about 70% to half of that
when in silico predicted logP and pKa were used (unpublished
Figure 7 Application of physiologically based pharmacokinetic data). Transporter is another emerging area of PBPK modeling,
modeling and simulation in various stages of drug discovery and however, most of the data generated are qualitative to answer the
development. Models were initially built with preclinical data, and question of yes or no of whether a compound is a substrate of a
later refined with available clinical information. This figure was transporter. PBPK modeling relies on kinetic data, such as the
adapted from Ref. 15 with permission. clearance of the compound via that transporter. Thus, additional
data of transporter clearance are needed for PBPK modeling.
potential between these two drugs31. In a crossover study
design, the washout period is changed upon the application of
an inhibitor and therefore needs to be simulated prior the study. References
A lengthy washout period is costly and also not convenient to
the patients, whereas a washout period too short would affect 1. Teorell T. Kinetics of distribution of substances administered to the
the data quality. body. I. The extravascular modes of administration. Arch Int Pharma-
6) For many high clearance compounds and substrate of CYP3A4/5, codyn Ther 1937;57:205–25.
the CYP3A4/5 inhibitor ritonavir is often used to enhance the 2. Reddy M, Yang RSH, Clewell III HJ, Andersen ME. In: Physiologi-
exposure. PBPK modeling can be used to simulate whether, in cally based pharmacokinetic modeling: science and applications.
Hoboken, NJ: Wiley; 2005.
the presence of ritonavir, the exposure of the drug candidate can
3. Rowland M, Peck C, Tucker G. Physiologically-based pharmacoki-
reach the level of efficacious exposure. When DDI liability
netics in drug development and regulatory science. Annu Rev
cannot be avoided, for example part of the chemical structure is Pharmacol Toxicol 2011;51:45–73.
responsible for the pharmacological activity but also carries 4. Jones HM, Chen Y, Gibson C, Heimbach T, Parrott N, Peters SA,
DDI liability, modeling and simulation can be applied to et al. Physiologically based pharmacokinetic modeling in drug
evaluate whether alternative routes of administration (e.g., discovery and development: a pharmaceutical industry perspective.
intravenous) would reduce the DDI liability. Drugs target for Clin Pharmacol Ther 2015;97:247–62.
inhalation or other non-conventional routes are better evaluated 5. Sinha V, Zhao P, Huang SM, Zineh I. Physiologically based
using PBPK modeling to get the feasibility test. pharmacokinetic modeling: from regulatory science to regulatory
policy. Clin Pharmacol Ther 2014;95:478–80.
6. Wagner C, Zhao P, Pan Y, Hsu V, Grillo J, Huang SM, et al.
Application of physiologically based pharmacokinetic (PBPK) model-
12. Conclusions ing to support dose selection: report of an FDA public workshop on
PBPK. CPT Pharmacomet Syst Pharmacol 2015;4:226–30.
PBPK modeling is a useful tool for the prediction of human PK 7. Brown RP, Delp MD, Lindstedt SL, Rhomberg LR, Beliles RP.
profile from preclinical data. Once FIH PK data or human ADME Physiological parameter values for physiologically based pharmacoki-
data becomes available, the model can be further fine-tuned as netic models. Toxicol Ind Health 1997;13:407–84.
8. Jones HM, Parrott N, Jorga K, Lavé T. A novel strategy for
illustrated in Fig. 715. It is a good tool for evaluating and
physiologically based predictions of human pharmacokinetics. Clin
optimizing clinical trial design, for example, to select the dose
Pharmacokinet 2006;45:511–42.
and dose schedule. It helps to understand the individual variability 9. Jones HM, Dickins M, Youdim M, Gosset JR, Attkins NJ, Hay TL,
and parameters that have the most impact on human PK profile et al. Application of PBPK modelling in drug discovery and develop-
through sensitivity analysis. Hence, PBPK modeling provides a ment at Pfizer. Xenobiotica 2012;42:94–106.
practical solution for extrapolating PK profile from healthy 10. Peters SA, Ungell AL, Dolgos H. Physiologically based pharmacoki-
population to some ethnical, special age, or disease populations netic (PBPK) modeling and simulation: applications in lead optimiza-
where clinical PK study is the hardest to conduct. In the DDI tion. Curr Opin Drug Discov Dev 2009;12:509–18.
prediction area, PBPK modeling can help to determine the 11. Shardlow CE, Generaux GT, Patel AH, Tai GY, Tran T, Bloomer JC.
washout period in a crossover study design to set the minimal Impact of physiologically based pharmacokinetic modeling
but sufficient clinical trial duration. It can also be applied as an and simulation in drug development. Drug Metab Dispos
2013;41:1994–2003.
alternative to DDI trials in some special populations, such as
12. Liu F, Zhuang XM, Yang CP, Li Z, Xiong S, Zhang ZW, et al.
pediatrics and organ-impairment patients where the actual DDI Characterization of preclinical in vitro and in vivo ADME properties
trial is hard to conduct due to logistical or ethical issues. Thus, it and prediction of human PK using a physiologically based pharma-
can sometimes provide waiver for conducting unnecessary clinical cokinetic model for YQA-14, a new dopamine D3 receptor antagonist
DDI trials which then speeds up the drug development process and candidate for treatment of drug addiction. Biopharm Drug Dispos
put fewer burdens on patients. Conducting DDI trials with multiple 2014;35:296–307.
440 Xiaomei Zhuang, Chuang Lu

13. Plowchalk DR, Rowland Yeo K. Prediction of drug clearance in a analysis of physiologically based pharmacokinetic modeling submis-
smoking population: modeling the impact of variable cigarette con- sions to the US Food and Drug Administration. Clin Pharmacokinet
sumption on the induction of CYP1A2. Eur J Clin Pharmacol 2015;54:117–27.
2012;68:951–60. 23. Peters SA, Jones CR, Ungell AL, Hatley OJD. Predicting drug
14. Zhuang XM, Zhong YH, Xiao WB, Li H, Lu C. Identification and extraction in the human gut wall: assessing contributions from drug
characterization of psoralen and isopsoralen as potent CYP1A2 metabolizing enzymes and transporter proteins using preclinical
reversible and time-dependent inhibitors in human and rat preclinical models. Clin Pharmacokinet 2016. Available from: 〈http://dx.doi.
studies. Drug Metab Dispos 2013;41:1914–22. org/10.1007/s40262-015-0351-6〉.
15. Lu C, Suri A, Shyu WC, Prakash S. Assessment of cytochrome P450- 24. Chu X, Korzekwa K, Elsby R, Fenner K, Galetin A, Lai Y, Matsson P,
mediated drug–drug interaction potential of orteronel and exposure et al. Intracellular drug concentrations and transporters: measurement,
changes in patients with renal impairment using physiologically based modeling, and implications for the liver. Clin Pharmacol Ther
pharmacokinetic modeling and simulation. Biopharm Drug Dispos 2013;94:126–41.
2014;35:543–52. 25. Lu C, Li P, Gallegos R, Uttamsingh V, Xia CQ, Miwa GT, et al.
16. FDA, Center for Drug Evaluation and Research (CDER). Guidance for Comparison of intrinsic clearance in liver microsomes and hepatocytes
industry: drug interaction studies-study design, data analysis, implications for from rats and humans: evaluation of free fraction and uptake in
dosing, and labeling recommendations [Draft Guidance]. U.S. Department of
hepatocytes. Drug Metab Dispos 2006;34:1600–5.
Health and Human Services, Food and Drug Administration: Rockville,
26. Huang SM. PBPK as a tool in regulatory review. Biopharm Drug
MD; 2012. Available from: 〈http://www.fda.gov/downloads/Drugs/Guidance
Dispos 2012;33:51–2.
ComplianceRegulatoryInformation/Guidances/ucm292362.pdf〉.
27. Leong R, Vieira MLT, Zhao P, Mulugeta Y, Lee CS, Huang SM, et al.
17. Suri A, Chapel S, Lu C, Venkatakrishnan K. Physiologically based and
Regulatory experience with physiologically based pharmacokinetic
population PK modeling in optimizing drug development: a predict-
modeling for pediatric drug trials. Clin Pharmacol Ther 2012;91:926–31.
learn-confirm analysis. Clin Pharmacol Ther 2015;98:336–44.
28. Wagner C, Pan Y, Hsu V, Sinha V, Zhao P. Predicting the effect of
18. Manolis E, Pons G. Proposals for model-based paediatric medicinal
CYP3A inducers on the pharmacokinetics of substrate drugs using
development within the current European Union regulatory frame-
physiologically based pharmacokinetic (PBPK) modeling: an analysis
work. Br J Clin Pharmacol 2009;68:493–501.
19. Parrott N, Davies B, Hoffmann G, Koerner A, Lave T, Prinssen E, of PBPK submissions to the US FDA. Clin Pharmacokinet
et al. Development of a physiologically based model for oseltamivir 2016;55:475–83.
and simulation of pharmacokinetics in neonates and infants. Clin 29. Zhao P, Rowland M, Huang SM. Best practice in the use of
Pharmacokinet 2011;50:613–23. physiologically based pharmacokinetic modeling and simulation to
20. EMA, Committee for Human Medicinal Products (CHMP). Guideline address clinical pharmacology regulatory questions. Clin Pharmacol
on the investigation of drug interactions [Final]. European Medicines Ther 2012;92:17–20.
Agency: London; 2012. Available from: 〈http://www.ema.europa.eu/ 30. Zhao P, Vieira MLT, Grillo JA, Song PF, Wu TC, Zheng JH, et al.
docs/en_GB/document_library/Scientific_guideline/2012/07/ Evaluation of exposure change of nonrenally eliminated drugs in
WC500129606.pdf〉. patients with chronic kidney disease using physiologically based
21. Vieira MD, Kim MJ, Apparaju S, Sinha V, Zineh I, Huang SM, et al. pharmacokinetic modeling and simulation. J Clin Pharmacol
PBPK model describes the effects of comedication and genetic 2012;52 Suppl 1:91S–108S.
polymorphism on systemic exposure of drugs that undergo multiple 31. Zhao P, Ragueneau-Majlessi I, Zhang L, Strong JM, Reynolds KS,
clearance pathways. Clin Pharmacol Ther 2014;95:550–7. Levy RH, et al. Quantitative evaluation of pharmacokinetic inhibition
22. Wagner C, Pan Y, Hsu V, Grillo JA, Zhang L, Reynolds KS, et al. of CYP3A substrates by ketoconazole: a simulation study. J Clin
Predicting the effect of cytochrome P450 inhibitors on substrate drugs: Pharmacol 2009;49:351–9.