Article

Comparative assessment of In Vitro-In Vivo extrapolation methods used for predicting hepatic metabolic clearance of drugs

Consultant, 4009 Sylvia Daoust, Québec City, Québec G1X 0A6, Canada. .
Journal of Pharmaceutical Sciences (Impact Factor: 2.59). 11/2012; 101(11):4308-26. DOI: 10.1002/jps.23288
Source: PubMed

ABSTRACT

The purpose of this study was to perform a comparative analysis of various in vitro--in vivo extrapolation (IVIVE) methods used for predicting hepatic metabolic clearance (CL) of drugs on the basis of intrinsic CL data determined in microsomes. Five IVIVE methods were evaluated: the conventional and conventional bias-corrected methods using the unbound fraction in plasma (fup), the Berezhkovskiy method in which the fup is adjusted for drug ionization, the Poulin et al. method using the unbound fraction in liver (fuliver), and the direct scaling method, which does not consider any binding corrections. We investigated the effects of the following scenarios on the prediction of CL: the use of preclinical or human datasets, the extent of plasma protein binding, the magnitude of CL in vivo, and the extent of drug disposition based on biopharmaceutics drug disposition classification system (BDDCS) categorization. A large and diverse dataset of 139 compounds was collected, including those from the literature and in house from Genentech. The results of this study confirm that the Poulin et al. method is robust and showed the greatest accuracy as compared with the other IVIVE methods in the majority of prediction scenarios studied here. The difference across the prediction methods is most pronounced for (a) albumin-bound drugs, (b) highly bound drugs, and (c) low CL drugs. Predictions of CL showed relevant interspecies differences for BDDCS class 2 compounds; the direct scaling method showed the greatest predictivity for these compounds, particularly for a reduced dataset in rat that have unexpectedly high CL in vivo. This result is a reflection of the direct scaling method's natural tendency to overpredict the true metabolic CL. Overall, this study should facilitate the use of IVIVE correlation methods in physiologically based pharmacokinetics (PBPK) model. (C) 2012 Wiley Periodicals, Inc. and the American Pharmacists Association J Pharm Sci 101:43084326, 2012

Full-text

Available from: Sami Haddad, May 06, 2014
Comparative Assessment of
In Vitro
In Vivo
Extrapolation
Methods used for Predicting Hepatic Metabolic Clearance of
Drugs
PATRICK POULIN,
1
CORNELIS E. C. A. HOP,
2
QUYNH HO,
2
JASON S. HALLADAY,
2
SAMI HADDAD,
3
JANE R. KENNY
2
1
Consultant, 4009 Sylvia Daoust, Qu
´
ebec City, Qu
´
ebec G1X 0A6, Canada
2
DMPK, Genentech Inc., South San Francisco, California 94080
3
D
´
epartement de Sant
´
e Environnementale et Sant
´
e au Travail, IRSPUM, Facult
´
edeM
´
edecine, Universit
´
edeMontr
´
eal, Montr
´
eal,
Qu
´
ebec H3T 1J4, Canada
Received 23 May 2012; revised 26 June 2012; accepted 17 July 2012
Published online 13 August 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23288
ABSTRACT: The purpose of this study was to perform a comparative analysis of various
in vitro--in vivo extrapolation (IVIVE) methods used for predicting hepatic metabolic clearance
(CL) of drugs on the basis of intrinsic CL data determined in microsomes. Five IVIVE methods
were evaluated: the “conventional and conventional bias-corrected methods” using the unbound
fraction in plasma (fu
p
), the “Berezhkovskiy method” in which the fu
p
is adjusted for drug
ionization, the “Poulin et al. method” using the unbound fraction in liver (fu
liver
), and the
“direct scaling method,” which does not consider any binding corrections. We investigated the
effects of the following scenarios on the prediction of CL: the use of preclinical or human
datasets, the extent of plasma protein binding, the magnitude of CL in vivo, and the extent
of drug disposition based on biopharmaceutics drug disposition classification system (BDDCS)
categorization. A large and diverse dataset of 139 compounds was collected, including those
from the literature and in house from Genentech. The results of this study confirm that the
Poulin et al. method is robust and showed the greatest accuracy as compared with the other
IVIVE methods in the majority of prediction scenarios studied here. The difference across
the prediction methods is most pronounced for (a) albumin-bound drugs, (b) highly bound
drugs, and (c) low CL drugs. Predictions of CL showed relevant interspecies differences for
BDDCS class 2 compounds; the direct scaling method showed the greatest predictivity for these
compounds, particularly for a reduced dataset in rat that have unexpectedly high CL in vivo.
This result is a reflection of the direct scaling method’s natural tendency to overpredict the
true metabolic CL. Overall, this study should facilitate the use of IVIVE correlation methods in
physiologically based pharmacokinetics (PBPK) model. © 2012 Wiley Periodicals, Inc. and the
American Pharmacists Association J Pharm Sci 101:4308–4326, 2012
Keywords: disposition; microsomes; clearance; unbound fraction; computational ADME;
in vitro–in vivo extrapolation; in vitro–in vivo correlation; IVIVE; pharmacokinetics; PBPK
modeling
Abbreviations used: AAG, alpha1-acid glycoprotein; AFE,
average-fold error; AL, albumin; BCS, bioclassification system;
BDDCS, biopharmaceutics drug disposition classification system;
CCC, concordance correlation coefficient; CL, clearance; CL
int
,in-
trinsic CL; fu
inc
, unbound fraction in incubation; fu
liver
, unbound
fraction in liver; fu
p
, unbound fraction in plasma; fu
p-app
, appar-
ent unbound fraction in plasma; IVIVE, in vitro–in vivo extrapola-
tion; K
m
, Michaelis–Menten constant; PhRMA, Pharmaceutical Re-
search and Manufacturers of America; PLR, plasma-to-whole liver
concentration ratio of albumin; Q
liver
, blood flow rate to liver; R
BP
,
blood-to-plasma concentration ratio; RMSE, root-mean-squared er-
ror.
Correspondence to: Dr. Patrick Poulin (Telephone: +418-802-
3985; E-mail: patrick-poulin@videotron.ca)
INTRODUCTION
Various methods are available to predict human phar-
macokinetics with some based on preclinical in vivo
data and others utilizing human in vitro data. in vitro
methods are most convenient because they require
minimal amount of compound and do not necessitate
animal studies. However, the predictivity of in vitro
methods depends on the model used and the input
Journal of Pharmaceutical Sciences, Vol. 101, 4308–4326 (2012)
© 2012 Wiley Periodicals, Inc. and the American Pharmacists Association
4308 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 1
COMPARATIVE ANALYSIS OF IVIVE METHODS 4309
parameters. Factors influencing the predictive per-
formance of an in vitroin vivo extrapolation (IVIVE)
method for hepatic metabolic clearance (CL) are re-
lated to several input parameters; namely, binding
terms such as the unbound fraction in plasma (fu
p
)
and in incubation medium (fu
inc
)aswellasthein-
trinsic CL (CL
int
) and liver blood flow rate (Q
liver
).
1–11
Wan et al.
1
studied the impact of these input param-
eters on the CL estimate in rat and human datasets.
The authors concluded that the simplified IVIVE
method, disregarding binding data (i.e., direct scal-
ing), might be sufficiently good for IVIVE evalua-
tions. Obach
2
also suggested disregarding all binding
data to predict human CL for basic and neutral com-
pounds, whereas for acidic compounds, he suggested
including all binding terms (i.e., fu
p
/fu
inc
). Recently,
Berezhkovskiy et al.
12,13
and Poulin et al.,
14
both of
whom also studied the impact of the binding terms
on CL estimates, presented two novel IVIVE meth-
ods. Berezhkovskiy’s method consists of replacing fu
p
with an apparent fu
p
(fu
p-app
) that considers drug ion-
ization differences between the plasma and liver cells.
Poulin et al.
14
further analyzed the concept of binding
terms and suggested converting the value of fu
p-app
to
an unbound fraction in the liver (fu
liver
) to take also
into account the role of extracellular binding proteins
on the passive uptake of drugs in hepatocytes. Us-
ing a dataset of 25 drugs, the Poulin et al.
14
method
showed the greatest accuracy as compared with other
IVIVE methods on the basis of several statistical
parameters.
14
Recently, Halifax and Houston
15
used
a larger dataset and demonstrated superior precision
and lower bias in the majority of cases for the novel
method of Poulin et al.; however, these authors are not
in total agreement on the mechanistic justification
of the method advocated by Poulin et al.
14
Instead,
Halifax and Houston
15
proposed an empirical scaling
method involving a conventional model, but corrected
for the average-fold error (AFE) (i.e., the conventional
bias-corrected method). Therefore, a consensus on the
use of IVIVE methods could not be agreed upon, and
hence, further testing is needed.
The purpose of this study was to further investigate
the published IVIVE methods by using large and di-
verse datasets from human, monkey, dog, and rat.
This study might help to identify potential outlier
drugs and apply further refined IVIVE methods to
identify the strengths and limitations of these meth-
ods.
METHODS
The overall strategy consisted of evaluating the effect
of the following scenarios on predictive performance
of various IVIVE methods for CL based on microso-
mal data: (a) the preclinical and human datasets, (b)
the extent of plasma protein binding [i.e., drugs bound
to albumin (AL), drugs bound to alpha1-acid glycopro-
tein (AAG), and drugs highly bound in plasma], (c) the
magnitude of CL under in vivo conditions (i.e., very
low, low, medium, and high CL), and (d) the extent
of drug disposition based on the biopharmaceutics
drug disposition classification system (BDDCS) and/
or bioclassification system (BCS) categorizations.
16
Furthermore, we explored the effect of hepatic up-
take on CL estimations by using the current IVIVE
methods for a reduced dataset of drugs in rats. For
this dataset, the Poulin et al.
14
method was compared
with the direct scaling method. We theorized that the
direct scaling method may be advantageous when CL
in vivo is unexpectedly high because this method nat-
urally overpredicts the true metabolic CL, as is re-
ported in the literature.
1,2,7,13
Finally, we present a
sensitivity analysis to demonstrate how the different
IVIVE methods vary with input parameters related to
drug ionization, plasma protein binding, and/or CL
int
.
Comparative Analysis of IVIVE Methods
Five IVIVE methods that have undergone previ-
ous comparative assessments were the focus of fur-
ther evaluation in this study.
1,2,14,15
These IVIVE
methods are (a) the “conventional” and “conven-
tional bias-corrected” methods using the fu
p
,(b)the
“Berezhkovskiy method” in which the fu
p
is adjusted
for drug ionization on either side of the plasma mem-
brane on the basis on pH differences, (c) the “Poulin
et al.
14
method” using the fu
liver
to adjust in addi-
tion for protein-facilitated uptake because of the po-
tential ionic interactions between the plasma-protein-
bound-drug complex and the cell surface of the hepa-
tocytes, and (d) the “direct scaling method” that does
not consider any binding corrections. Table 1 summa-
rizes all equations related to these IVIVE methods.
Recently, Halifax and Houston
15
reported an empiri-
cal method, the “conventional bias-corrected method”,
which involves multiplying the predicted CL values
from the conventional method with the corresponding
average bias of underprediction to reduce the under-
prediction. The average bias of underprediction was
obtained from the AFE observed for each dataset (i.e.,
each prediction scenario) of this study. This empirical
method was also evaluated in this study. The well-
stirred model was considered for the purpose of this
study. Furthermore, the parallel tube model was also
used for high CL compounds for all IVIVE methods
tested because it is expected that the prediction ac-
curacy for these drugs will increase with the parallel
tube model.
2
Estimation of the Input Parameters
The five IVIVE methods scale CL
int
determined in
microsomes from in vitro-to-in vivo conditions by us-
ing a physiologically based scaling factor based on
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 2
4310 POULIN ET AL.
Table 1. IVIVE Methods Tested in this Study as Reported from the Literature
1,2,12–14
Methods Model Equations of Plasma CL
Conventional method CL =
Q
liver
R
BP
CL
int,in vivo
fu
p
Q
liver
R
BP
+ CL
int,in vivo
fu
p
(1)
Adapted from Berezhkovskiy
a
CL =
Q
liver
R
BP
CL
int,in vivo
fu
papp
/fu
inc
Q
liver
R
BP
+ CL
int,in vivo
fu
papp
/fu
inc
(2)
where
fu
papp
=
F
1
fu
p
1+
(
F
1
1
)
fu
p
(3)
and
F
1
=
f
unionizedplasma
f
unionizedliver,cells
(4)
and
f
unionized neutral
= 1(5)
f
unionizedmonoproticacid
= 1/
1 +
10
pHpK
a

(6)
f
unionizedmonoproticbase
= 1/
1 +
10
pK
a
pH

(7)
f
unionizeddiproticbases
=
1/
1 +
10
pK
a1
pH
+ 10
pK
a1
+pK
a2
2pH

(8)
Poulin et al.
14
method CL =
Q
liver
R
BP
CL
int,in vivo
fu
liver
/fu
inc
Q
liver
R
BP
+ CL
int,in vivo
fu
liver
/fu
inc
(9)
where
fu
liver
=
PLRfu
papp
1+
(
PLR1
)
fu
papp
(10)
Direct scaling method CL =
Q
liver
R
BP
CL
int,in vivo
Q
liver
R
BP
+ CL
int,in vivo
(11)
AL, albumin; AAG, alpha1-acid glycoprotein; Q
liver
, liver blood flow rate; R
BP
, blood-to-plasma ratio; fu
p
, unbound fraction in plasma;
fu
p-app
, apparent unbound fraction considering the pH gradient; fu
liver
, unbound fraction in liver considering the protein-facilitated
uptake and pH gradient; fu
inc
, unbound fraction in incubation medium (microsomes); F
I
, ionization factor; f
unionized
, fraction unionized;
CL
int
, intrinsic clearance; PLR, plasma-to-whole liver ratio of AL or AAG. For the conventional-bias corrected method, the AFE value is
used as described in the Method.
a
The calculation of fu
p-app
has been made from the binding isotherm to cover both the highly and the less bound drugs, and to avoid
fu
p-app
> 1 for basic drugs.
hepatic microsomal recovery from whole liver to con-
vert CL
int
in vitro in :L/(min mg proteins) to mL/(min
kg body weight) for CL
int
in vivo (i.e., 900, 800, 1908,
and 1720 mg protein per kg body weight in humans,
monkeys, dogs, and rats, respectively).
1,2,6,14
In ad-
dition, Q
liver
was taken to be 20.7, 43.6, 30.3, and
65 mL/(min kg) in humans, monkeys, dogs, and rats,
respectively.
6,14
In in vitro experiments it is doubtful whether the
pH gradient between the extracellular and intra-
cellular spaces of liver is maintained. Because mi-
crosomes were incubated with a buffer at pH 7.4,
the in vitro measurements of CL
int
do not account
for pH differences in extracellular and intracellular
spaces of liver. Thus, application of the methods of
Berezhkovskiy
12,13
and Poulin et al.
14
may represent
a significant improvement because the pH gradient
for the ionized drugs is taken into account (Table 1).
The pH values used were 7.4 for plasma and 7.0 for
liver cells. The pH value of liver cells was chosen to
represent the mean of most of the values reported in
the literature, which ranged from 6.89 to 7.12.
17–19
As mentioned, the Poulin et al.
14
method consists
of adjusting the well-stirred model for the protein-
facilitated uptake in addition to the effect of the pH
gradient. These effects are considered in the estima-
tion of fu
liver
(Table 1). The traditional assumption is
that equilibrium between the free and protein-bound
drug is instantaneous, such that the metabolism pro-
cess is driven by a constant supply of unbound drug
concentration in plasma. However, in this study, we
also have assumed that unbound drug concentration
in liver and plasma differs because of the protein-
facilitated uptake of drugs into hepatocytes, and
hence, the current fu
liver
represents a correction of
the observed fu
p
to take into account the extracellu-
lar protein binding in plasma relative to liver in addi-
tion to the effect of the pH gradient. Thus, the current
fu
liver
should not be totally comparable to the overall
unbound fraction determined in liver homogenates,
for example, which would consider binding to diverse
components (e.g., lipids).
It was determined that AL plays a major role in
the protein-facilitated uptake of drugs in hepatocytes
as compared with the AAG.
14
Consequently, Poulin
et al.
14
taken into account the extracellular protein
binding in plasma relative to liver [or the plasma-to-
whole liver concentration ratio (PLR) of AL or AAG]
in the calculation of fu
liver
. The experimentally deter-
mined value of PLR for AL in rat and human is about
13.3.
14,20,21
For monkey, the same PLR value of 13.3
was used in this study. Briefly, the PLR value of AL
was obtained by converting the measured plasma-to-
extracellular fluid concentration ratio into plasma–w-
hole liver ratio to be in agreement with the well-
stirred model, which assumes an homogeneous drug
distribution in the liver.
14
The volume of interstitial
fluid in liver is used as the conversion factor because
the level of AL in liver cells (0.4 mg/g) is negligible
as compared with the extracellular fluid, and hence,
of serum (40 mg/g).
14,22
However, for dog, the vol-
ume of interstitial fluid in the liver is far greater
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 3
COMPARATIVE ANALYSIS OF IVIVE METHODS 4311
than in other species (0.29 vs. 0.18).
23,24
Conse-
quently, when species difference in the volume of in-
terstitial space is considered, the PLR value in dog is
estimated to be 8.5 (i.e., assuming that the plasma-to-
extracellular fluid concentration ratio of AL for liver
is species invariant). For each drug bound to AAG,
the PLR value for this protein was set to unity, and
hence, fu
liver
is equal to fu
p-app
14
(Table 1). The values
of PLR can be debated and were questioned by Hali-
fax and Houston.
15
However, it should be emphasized
that the values of PLR across species were chosen in
a truly prospective fashion considering the measured
PLR data presented in the literature without knowing
the predictivity of this IVIVE method.
Datasets
The drug datasets and the corresponding experi-
mentally determined input parameters are presented
in the Appendix (see Supplementary information).
The datasets consist of drugs obtained from the
literature
1–7,14,25
and drug candidates synthesized at
Genentech (South San Francisco, California). A Phar-
maceutical Research and Manufacturers of America
(PhRMA) initiative
6,7
also published concise and com-
plete datasets with experimental data for several
drugs in rat, dog, monkey, and/or human. For the
studies of Wan et al.
1
and Hosea et al.,
5
only the
neutral drugs were considered because the essential
pK
a
values of the proprietary compounds were not
presented in these studies. In addition, the in vitro
and in vivo CL as well as the essential input pa-
rameters was experimentally determined in rat for
21 compounds at Genentech as described in the Ap-
pendix (see Supplementary information). As men-
tioned, a reduced dataset for rat containing five drugs
for which CL in vivo is governed by an active up-
take process (digoxin, ketanserin, PhRMA #37, #40,
and #43) is also presented in the Appendix (see Sup-
plementary information).
6,26–28
The total number of
drugs studied here is 139, covering a large range of
drug properties. The in vitro CLdatausedinthis
study were obtained from in vitro metabolic assays by
using plasma-free microsomal incubations. The main
binding protein (AL or AAG) was identified for each
literature drug on the basis of mechanistic studies
published in the literature.
14
When this information
was not available (28 proprietary compounds), bind-
ing to AAG was assumed to be preferred for strongly
basic drugs, whereas binding to AL was assumed to
be preferred for acidic and neutral drugs. However,
for each Genentech compound, the main binding pro-
tein was determined experimentally as described in
the Appendix (see Supplementary information) (i.e.,
the binding ratio of human AAG to human AL was
determined by using a dialysis system).
Modeling Assumptions
The drugs used in this study are thought to be elimi-
nated by hepatic oxidative metabolic CL under in vivo
conditions in each species. The liver metabolism
for several of the marketed compounds is governed
mainly by cytochrome P450 (CYP) enzymes.
1–6,14
Al-
though this information was unknown for the propri-
ety compounds, we assumed that their metabolism
was governed similarly. Transporter-mediated pro-
cesses that could possibly be responsible for drug up-
take or drug efflux from hepatocytes were neglected,
except for the five drugs in the reduced rat dataset.
Because the microsomal data were determined in
plasma-free incubations, it was assumed that drug
distribution from plasma to hepatocytes was not im-
peded by limited diffusion processes under in vivo
conditions. In the metabolic stability assays, the sub-
strate concentration was expected to be well below
the apparent Michaelis–Menten constant (K
m
). It was
also assumed that the in vivo CL of drugs follows
the free drug hypothesis according to the well-stirred
model. In this case, binding to plasma proteins was
assumed to be reversible and not saturated at the
conditions studied.
Evaluation of Predictive Performance
The prediction accuracy was assessed by comparing
predicted versus observed values of CL by using sev-
eral statistical parameters.
6,15,24
The AFE and root-
mean-squared error (RMSE) were calculated and are
presented for each prediction method. Furthermore,
the concordance coefficient of correlation (CCC) global
is presented, which evaluates the global degree to
which pairs of predicted and observed data fall on
the line of unity passing through the origin. Specific-
fold errors of deviation between the predicted and ob-
served values (percentage of fold error 2and5)
were also calculated. Finally, plots of predicted ver-
sus observed CL values are also presented for each
IVIVE method.
Sensitivity Analysis
Theoretical simulations of plasma CL in human over a
large range of CL
int
were investigated with the Poulin
et al.,
14
Berezhkovskiy, and direct scaling IVIVE
methods. A span of predicted CL values is presented
for two drug examples, namely, a strong acid (pK
a
=
2) and a base (pK
a
= 10) either bound (fu
p
= 0.01)
or less bound (fu
p
= 0.9) in plasma. For the Poulin
et al.
14
method the acidic and basic drugs were de-
fined to bind to AL and AAG, respectively.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 4
4312 POULIN ET AL.
RESULTS
Comparative Assessment for Various IVIVE Methods for
Predicting CL
Several IVIVE calculation methods of CL were com-
pared by using the same drug datasets, and the com-
parative assessment was made on the basis of several
statistical parameters. The overall statistical sum-
mary in terms of accuracy, precision, and correlation
is listed in Tables 2–4 for the different scenarios of
prediction. The plots of predicted versus observed CL
values for each method are shown in Figures 1–5,
whereas Figure 6 compares the precision bias across
the IVIVE methods. The sensitivity analysis is pre-
sented in Figure 7.
All Datasets
The results of this study confirm that the Poulin
et al.
14
method is robust and showed the greatest ac-
curacy among the IVIVE methods tested in the major-
ity of the prediction scenarios. The statistical analysis
is generally in favor of the Poulin et al.
14
method be-
cause superior, or at least comparable, statistics were
obtained by this method as compared with those ob-
tained by other methods (Tables 2–4); this is also cor-
roborated graphically in Figures 1–5. The empirical,
conventional bias-corrected method suggested by Hal-
ifax and Houston
15
showed lower predictivity as com-
pared with the other methods studied here. This em-
pirical method was, therefore, not fully investigated in
this study. Moreover, the conventional bias-corrected
method required analysis of several datasets to first
determinate the AFE correction factor, which depends
on the composition of the datasets (i.e., number and
disposition of compounds) as shown by the variabil-
ity in the AFE values across the various prediction
scenarios of this study. Therefore, it is not surprising
to observe that the resulting AFE values of the con-
ventional bias-corrected method are close to unity in
most cases, however, other statistical parameters are
not much improved over other methods (Tables 2–4).
An important observation is that the Poulin et al.
14
method generally provides AFE values close to unity,
whereas the Berezhkovskiy and conventional meth-
ods provide lower AFE value and the direct scal-
ing method gives the highest AFE value. Therefore,
a systematic underprediction of CL in vivo was ob-
served for the Berezhkovskiy and conventional meth-
ods, whereas the direct scaling method overpredicted
the CL as compared with the Poulin et al.
14
method.
In particular, the Poulin et al.
14
method presented
the lowest bias in precision especially for the follow-
ing prediction scenarios: AL-bound drugs and those
drugs with low fu
p
and low CL values (Fig. 6). Ac-
cordingly, the global CCC value is closest to unity
for the Poulin et al.
14
method, and hence, the RMSE
Table 2. Comparative Assessment of IVIVE Methods Used to Predict Plasma CL According to Preclinical and
Human Datasets
Percentage of Twofold
or Less Error
Percentage of Fivefold
or Less Error AFE RMSE CCC
All datasets (
n
= 134)
Poulin et al.
14
77 96 0.99 0.27 0.91
Berezhkovskiy 59 83 0.52 0.50 0.79
Conventional 32 58 0.22 0.85 0.55
Conventional bias corrected 38 83 1.02 0.56 0.75
Direct scaling 70 92 1.49 0.41 0.74
Human (
n
= 48)
Poulin et al.
14
88 100 0.96 0.19 0.96
Berezhkovskiy 69 83 0.52 0.48 0.83
Conventional 25 56 0.21 0.83 0.57
Direct scaling 67 90 1.66 0.50 0.59
Dog and monkey (
n
= 22)
Poulin et al.
14
82 100 1.08 0.21 0.79
Berezhkovskiy 50 73 0.46 0.56 0.42
Conventional 41 50 0.24 0.91 0.21
direct scaling 77 95 1.49 0.27 0.60
Rat (this study) (
n
= 21)
Poulin et al.
14
71 95 0.99 0.30 0.82
Berezhkovskiy 43 91 0.56 0.41 0.70
Conventional 19 52 0.23 0.73 0.46
Direct scaling 62 81 1.40 0.50 0.14
Rat (literature) (
n
= 43)
Poulin et al.
14
65 91 0.99 0.35 0.59
Berezhkovskiy 60 84 0.53 0.52 0.53
Conventional 42 67 0.24 0.91 0.33
Direct scaling 74 98 1.38 0.28 0.59
AFE, average-fold error; RMSE, root-mean-square error; CCC, concordance correlation coefficient.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 5
COMPARATIVE ANALYSIS OF IVIVE METHODS 4313
Table 3. Comparative Assessment of IVIVE Methods Used to Predict Plasma CL According to Drugs Bound to
AL and AAG and Drugs Highly Bound in Plasma (fu
p
0.01)
Percentage of Twofold
or Less Error
Percentage of Fivefold
or Less Error AFE RMSE CCC
Highly bound (
n
= 10)
Poulin et al.
14
60 100 1.06 0.30 0.82
Berezhkovskiy 40 80 0.36 0.65 0.44
Conventional 0 0 0.05 1.42 0.06
Conventional bias corrected 40 80 1.04 0.60 0.27
Direct scaling 50 70 2.56 0.56 0.25
AL bound (
n
= 72)
Poulin et al.
14
74 96 1.21 0.28 0.92
Berezhkovskiy 40 71 0.36 0.64 0.75
Conventional 28 50 0.19 0.95 0.54
Conventional bias corrected 38 78 0.97 0.61 0.74
Direct scaling 64 90 1.72 0.46 0.74
AAG bound (
n
= 62)
Poulin et al.
14
81 97 0.79 0.26 0.89
Berezhkovskiy 81 97 0.79 0.26 0.89
Conventional 37 68 0.28 0.72 0.55
Conventional bias corrected 39 90 1.0 0.47 0.75
Direct scaling 77 94 1.26 0.33 0.74
AFE, average-fold error; RMSE, root-mean-square error; CCC, concordance correlation coefficient.
Table 4. Comparative Assessment of IVIVE Methods Used to Predict Plasma CL According to the Magnitude of
CL In Vivo
Percentage of Twofold
or Less Error
Percentage of Fivefold
or Less Error AFE RMSE CCC
Very low CL (5% of
Q
liver
)(
n
= 13)
Poulin et al.
14
62 92 1.15 0.33 0.86
Berezhkovskiy 15 23 0.16 0.89 0.63
Conventional 8 31 0.11 1.04 0.46
Conventional bias corrected 62 92 1.63 0.36 0.86
Direct scaling 31 54 5.67 0.94 0.42
Low CL (25% of
Q
liver
)(
n
= 44)
Poulin et al.
14
59 91 1.19 0.36 0.87
Berezhkovskiy 46 73 0.41 0.66 0.71
Conventional 27 55 0.20 0.91 0.51
Conventional bias corrected 43 73 0.99 0.58 0.71
Direct scaling 34 75 2.99 0.66 0.53
Medium CL (
n
= 70)
Poulin et al.
14
83 100 0.96 0.22 0.75
Berezhkovskiy 64 87 0.59 0.42 0.44
Conventional 31 54 0.21 0.90 0.21
Conventional bias corrected 31 83 0.98 0.58 0.38
Direct scaling 84 100 1.14 0.21 0.75
High CL (75% of
Q
liver
)(
n
= 20)
Well-stirred model
Poulin et al.
14
95 100 0.74 0.18 0.79
Berezhkovskiy 70 90 0.57 0.32 0.59
Conventional 45 80 0.40 0.47 0.40
Conventional bias corrected 70 100 1.01 0.25 0.70
Direct scaling 100 100 0.82 0.12 0.90
Parallel tube model
Poulin et al.
14
95 100 0.91 0.13 0.88
Berezhkovskiy 85 95 0.70 0.27 0.66
Conventional 55 85 0.48 0.42 0.46
Conventional bias corrected 60 100 1.01 0.28 0.66
Direct scaling 100 100 0.98 0.08 0.95
AFE, average-fold error; RMSE, root-mean-square error; CCC, concordance correlation coefficient.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 6
4314 POULIN ET AL.
Poulin et al.
Predicted CL
Observed CL
Figure 1. Predicted CL versus observed CL for the IVIVE
method proposed by Poulin et al.
14
according to all datasets
(CCC = 0.91 and n = 134). The solid line indicates the
best fit (unity). Short and long dashed lines on either side
of unity represent twofold and threefold error, respectively.
CL is in mL/(min kg).
Berezhkovskiy
Predicted CL
Observed CL
Figure 2. Predicted CL versus observed CL for the IVIVE
method proposed by Berezhkovskiy according to all datasets
(CCC = 0.79 and n = 134). The solid line indicates the best
fit (unity). Short and long dashed lines on either side of
unity represent twofold and threefold error, respectively.
CL is in mL/(min kg).
value is the lowest. Furthermore, this method often
shows the greatest accuracy based on the fold errors
of deviation between the predicted and observed val-
ues (percentage of fold error 2and5). It is use-
ful to take a closer look at the prediction scenarios in
more detail because more information can be obtained
on the predictivity of one method compared with
another.
Predictivity of Preclinical and Human Datasets
On the basis of all statistical parameters, the Poulin
et al.
14
method was the best performing prediction
method when considering the current human dataset
(n = 48; Table 2). This method resulted in 88% of
predicted CL within twofold error as compared with
the observed values in human, whereas this number
was lower for the other methods tested (25%–69%).
Again, no systematic underestimation or overestima-
tion of CL in human was observed with the Poulin
Conventional
Predicted CL
Observed CL
Figure 3. Predicted CL versus observed CL for the con-
ventional IVIVE method according to all datasets (CCC
= 0.55 and n = 134). The solid line indicates the best fit
(unity). Short and long dashed lines on either side of unity
represent twofold and threefold error, respectively. CL is in
mL/(min kg).
Conventional bias corrected
Predicted CL
Observed CL
Figure 4. Predicted CL versus observed CL for the con-
ventional bias-corrected IVIVE method according to all
datasets (CCC = 0.75 and n = 134). The solid line indi-
cates the best fit (unity). Short and long dashed lines on
either side of unity represent twofold and threefold error,
respectively. CL is in mL/(min kg).
Direct scaling
Predicted CL
Observed CL
Figure 5. Predicted CL versus observed CL for the direct
scaling IVIVE method according to all datasets (CCC = 0.74
and n = 134). The solid line indicates the best fit (unity).
Short and long dashed lines on either side of unity represent
twofold and threefold error, respectively. CL is in mL/(min
kg).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 7
COMPARATIVE ANALYSIS OF IVIVE METHODS 4315
AL-bound
Low CL
Prediction scenarious
Low f
up
Figure 6. AFE values obtained from the different IVIVE
methods for different prediction scenarios. Black squares,
direct scaling method; red circles, Poulin et al.
14
method;
green triangles, Berezhkovskiy method; and pink dia-
monds, conventional method.
et al.
14
method (AFE = 0.96), whereas the other meth-
ods underpredicted or overpredicted human CL (AFE
= 0.21, 0.52, and 1.66). Similar findings are obtained
for the dog and monkey datasets. For the rat datasets,
the predictivity of the Poulin et al.
14
method is supe-
rior (n = 21; Genentech dataset) or comparable (n
= 43; literature dataset) to the predictivity tested by
other IVIVE methods. In addition, the AFE values re-
sulting from the predictivity for the two rat datasets
is still closest to unity for the Poulin et al.
14
method
(Table 2). Furthermore, it is observed in Table 2 that
the Poulin et al.
14
method is the most accurate for the
rat dataset of Genentech compounds for which all in-
put parameters have been experimentally determined
(including the main binding protein).
Table 2 indicates that the predictive performance
decreases in the rat datasets as compared with the
human dataset for the conventional, Berezhkovskiy,
and Poulin et al.
14
IVIVE methods. The average CL
of the rat datasets represents 62% of the liver blood
flow rate, whereas this number decreases to 32% for
the human dataset (Appendix; see Supplementary in-
formation). This means that the compounds for which
we have human data may have less complex pharma-
cokinetics in humans as compared with the rats (i.e.,
mainly CYP-mediated oxidative metabolism). This
aspect is considered further in the Discussion section.
Conversely, the predictive performance of the direct
scaling method increased in the rat datasets as com-
pared with the human dataset. Therefore, the direct
scaling method was further evaluated with a reduced
rat dataset of five drugs for which their CL in vivo
is unexpectedly high (Table A2; see Supplementary
information) as well as according to the effect of the
BDDCS and/or BCS categorization of the literature
drugs
16
(Table A1; see Supplementary information)
as presented below.
a Acidic drug;f = 0.01
30
20
10
0
30
20
10
0
0 1000 2000 3000 4000
0 1000 2000 3000 4000
Calculated plasma CL [mL/(min/ kg)]
up
c Basic drug;f = 0.01
Intrinsic CL [mL/(min/ kg)]
up
d Basic drug;f = 0.9
up
b Acidic drug;f = 0.9
up
30
20
10
0
0
1000
2000 3000
4000
30
20
10
0
0
1000
2000
3000
4000
Figure 7. Theoretical simulations of plasma CL in human over a large range of CL
int
.(a)An
acidic drug highly bound to AL (pK
a
= 2 and fu
p
= 0.01). (b) An acidic drug not significantly
bound to AL (pK
a
= 2 but fu
p
= 0.9). (c) A basic drug highly bound to AAG (pK
a
= 10 and fu
p
=
0.01). (d) A basic drug not significantly bound to AAG (pK
a
= 10 but fu
p
= 0.9). The other input
parameters in the IVIVE methods were set equal to 1 (R
BP
) and 0.5 (fu
inc
). Some simulations
are similar and, therefore, overlap (b and d). Poulin et al.’s
14
method is shown by a black solid
line, Berezhkovskiy’s method is shown by a dashed gray line (long dashed lines), and the direct
scaling is shown by a dashed blue line (short dashed lines).
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 8
4316 POULIN ET AL.
Assessing Prediction of CL in Rat for a Reduced Dataset
The Poulin et al.
14
method correctly predicted rat CL
within threefold error for only one of the five drugs
in which CL in vivo in rat was unexpectedly high
because of active uptake, whereas the direct scal-
ing method correctly predicted three drugs. The di-
rect scaling method provided improved prediction per-
formance in this circumstance because it naturally
shows trends of overprediction of CL as compared
with the Poulin et al.
14
method. The reason being
the resulting AFE value is much greater for the di-
rect scaling method (0.43) than for the Poulin et al.
14
method (0.08) for these five drugs.
Predictivity According to the BDDCS and/or BCS Class
Relevant interspecies differences were observed in
predictivity between class 1 and 2 compounds when
either the BDDCS or BCS information was used. For
example, for the rat dataset, the Poulin et al.
14
method
generated 88% (15/17) of the CL estimates within
twofold error for class 1 compounds, but this num-
ber decreased significantly to 30% (3/10) for class 2
compounds. Conversely, the predictivity for the hu-
man dataset was more comparable between the two
drug classes, with 89% (25/28) and 79% (11/14) of the
CL estimates within twofold error for class 1 and 2
compounds, respectively. Another example is that the
prediction of CL in vivo of ketanserin and ritanserin
(BDDCS class 2) in human is within twofold error,
whereas in the rat, the predicted values are greatly
underpredicted (up to a factor of sixfold) (not shown).
The disparity between predictions of class 2 com-
pounds in rat is particularly evident with the Poulin
et al.
14
method as compared with the direct scaling
method. The reason is that the direct scaling method
provides higher predicted CL values, which is benif-
ical when CL in vivo is higher than expected in rat
especially for those class 2 compounds.
Predictivity According to Plasma Protein Binding
As expected, the Poulin et al.
14
method showed su-
perior predictive performance particularly for drugs
bound mainly to AL and those highly bound in
plasma (fu
p
values 0.01) (Table 3). For example,
the Poulin et al.
14
method provided AFE values rang-
ing from 1.06 to 1.21, whereas the conventional and
Berezhkovskiy methods obtained much lower AFE
values (0.05–0.36). The direct scaling method did not
perform better as AFE values ranged from 1.72 to
2.56 (Fig. 6). Note that the Berezhkovskiy and Poulin
et al.
14
methods behave the same for (basic) com-
pounds bound to AAG because PLR is equal to unity
for AAG as explained in the Methods section (i.e.,
fu
liver
is equal to fu
p-app
) (Tables 1 and 3).
Predictivity According to the Magnitude of Drug CL
An important difference in predictivity was observed
among the IVIVE methods tested when the magni-
tude of CL in vivo of the drug was considered (Table
4). For drugs with a low CL in vivo (25% of Q
liver
), the
Poulin et al.
14
method is by far the most accurate pre-
diction method. This is even more evident for drugs
with a very low CL in vivo (5% of Q
liver
). Indeed, 62%
of CL predictions for drugs with low and very low CL
in vivo are within twofold error with the Poulin et al.
14
method, whereas the estimates range from 8% to 31%
for the other IVIVE methods. The AFE value for the
Poulin et al.
14
method is 1.15, whereas this value is
much greater (5.67) for the direct scaling method and
is lower for the conventional (0.11) and Berezhkovskiy
(0.16) methods (Fig. 6). Conversely, for the medi-
um–high CL range (>25% of Q
liver
), the Poulin et al.
14
and direct scaling methods are comparable, followed
by the conventional and Berezhkovskiy methods. The
reasonable predictivity of the direct scaling method
particularly for the high CL compounds is in line
with previous findings on trends of over-prediction
of CL compared to the other IVIVE methods. The re-
sults for high CL compounds slightly improved for
all IVIVE methods when the parallel-tube model is
used as compared with the well-stirred model. This
is reflected in the values of AFE and CCC, which are
closer to unity, whereas the values of RMSE decreased
(Table 4).
Sensitivity Analysis
The sensitivity analysis demonstrates the importance
of compound properties and binding parameters that
are reflective of specific mechanistic determinants rel-
evant to prediction of the CL values. Figure 7 illus-
trates differences among the methods tested, partic-
ularly at lower CL
int
values; at higher CL
int
values,
the IVIVE methods are more similar. Furthermore,
the class of drug affected CL predictions. This is more
noticeable when an acidic drug is highly bound to AL
than when a basic drug is highly bound to AAG. The
differences observed in the sensitivity analysis are in
accordance with the superior predictivity of the Poulin
et al.
14
method especially in the problematic areas of
high protein binding and low CL, as reported previ-
ously (Tables 3 and 4). Thus, the importance of fu
p
as
an input parameter is obvious, but it should be high-
lighted that it is experimentally hard to determine
fu
p
accurately for highly bound drugs.
DISCUSSION
We conducted a comparative analysis of five promis-
ing IVIVE methods by using a dataset of 139 drugs ob-
tained in preclinical species and human. The findings
of this assessment suggest that the method proposed
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 9
COMPARATIVE ANALYSIS OF IVIVE METHODS 4317
by Poulin et al.
14
offers a significant improvement in
the prediction of CL as compared with other IVIVE
methods. In other words, the results confirm the valid-
ity of the novel method published by Poulin et al.
14
for
calculation of hepatic metabolic CL, which accounts
for pH differences in extracellular and intracellular
water of liver as well as the protein-facilitated uptake.
Berezhkovskiy et al.
13
previously studied the effect of
a pH gradient on the estimation of CL
int
under in vitro
conditions, but this study and two published compara-
tive analyses
14,15
demonstrate that considering solely
a pH gradient in the IVIVE of CL is not sufficient, and
hence, the protein-facilitated uptake should also be
included for more accurate extrapolations. Notably,
the general success of the novel IVIVE method of
Poulin et al.
14
is probably because of the adjustment
of the well-stirred model with fu
liver
. Because no con-
sideration of mechanisms related to the main bind-
ing proteins in liver was made with the other IVIVE
methods tested, this may explain why they generally
provided a lower prediction performance as compared
with the Poulin et al.
14
method. In summary, the dif-
ference across these methods is most pronounced for
(a) AL-bound drugs, (b) low CL drugs, and (c) highly
bound drugs (Figs. 6 and 7).
Halifax and Houston
15
suggested that the predic-
tion bias was CL dependent and protein binding de-
pendent for all methods, indicating important sources
of bias from in vitro methodology. The extent of CL
and protein binding are acknowledged to be poten-
tial sources of error that could cause discrepancy be-
tween in vitro and in vivo values. In this study, rel-
atively lower predictivities were obtained for low CL
compounds as compared with high CL compounds as
well as for highly bound drugs as compared with less
bound drugs (Tables 3 and 4). Halifax and Houston
15
suggested that prediction from microsomes, and par-
ticularly from hepatocytes, might be improved beyond
any of the methods assessed in this study through
the use of an empirical correction factor to eliminate
both the average bias and the CL dependency bias.
Sohlenius-Sternbeck et al.
29
also presented a method
of removing the systematic bias through application
of empirical correction factors derived from regres-
sion analyses applied to the in vitro and in vivo data
for a defined set of reference compounds. However,
Sohlenius-Sternbeck et al.
29
optimized their regres-
sion equations from hepatocyte data only, and hence,
they cannot be applied in the present study where
microsomal data were used. Indeed, these two em-
pirical methods are fully dependent on the analy-
sis of the in vivo datasets. In contrast, the Poulin
et al.
14
method only requires compound-specific input
and no prior analysis of a large dataset to provide
CL predictions. The bias for CL and protein bind-
ing was reduced in this study with the mechanistic
method of Poulin et al.
14
as compared with the em-
pirical technique from Halifax and Houston
15
on the
basis of several statistical parameters for problem-
atic areas (i.e., compounds with low fu
p
and low CL
values). Both this study and the literature
14
demon-
strate that considering fu
liver
in IVIVE represents a
significant statistical improvement in these problem-
atic areas as compared with other methods (Fig. 6);
therefore, justification for the mechanistic modelling
proposed here, including that for highly bound drugs,
is well supported. The Poulin et al.
14
method is sen-
sitive to the value of PLR, but the value was de-
termined on the basis of the measured data avail-
able in the literature, and the analysis presented
here provides justification for the value and the novel
approach.
Mechanistic plasma protein binding assays are re-
quired to identify the major binding protein to ap-
ply the IVIVE calculation method of Poulin et al.
14
prospectively. In this context, the present study pro-
posed a new experimental setting to define whether
AL or AGG is the main binding protein in plasma for
21 Genentech compounds (i.e., when AAG/AL is dial-
ysed against each other, a ratio greater than 0.6 sug-
gested AAG is preferred; Appendix; see Supplemen-
tary information). Many highly binding drugs may
bind to both proteins (AL or AAG). We assumed ei-
ther one or the other protein, but in reality the bind-
ing ratio of 0.6 is an arbitrary cutoff. For compounds
that bind equally to both proteins, we may potentially
consider a binding to AL invoking PLR correction in
the estimation of fu
liver
and we will explore that fur-
ther in a subsequent analysis. Any significant errors
in experimental assessment of fu
p
would confound
the predictability of IVIVE, particularly for highly
bound drugs.
10,14
For several propriety compounds
(i.e., Pfizer, J&J, and PhRMA compounds), the main
binding protein was unknown, and so it was estimated
on the basis of the class of drug, which may have af-
fected the comparative assessment. The paucity and
variability of literature data relating to IVIVE demon-
strate the importance of generating experimentally
consistent data for the purpose of method validation
because the data quality is crucial for the prediction
accuracy, particularly for compounds with low fu
p
and
CL values.
1,10,14,30
In light of the fact that discrepan-
cies are observed in CL
int
,fu
inc
,andfu
p
values as
well as in in vivo plasma concentrations between lab-
oratories, Beaumont et al.
10
strongly recommend us-
ing the same concentration (preferably a low one) of
compounds for both the in vitro metabolic stability
and binding assays to avoid possible concentration-
dependent effects. In fact, the substrate concentra-
tion should be well below the apparent K
m
.Another
source of error is the variability in the lipid content
and composition of liver microsomes,
31
both of which
may influence the measure of K
m
particularly in a
human population.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 10
4318 POULIN ET AL.
Despite potential experimental errors, protein-
facilitated uptake has been reported as an important
factor affecting the CL of highly bound drugs un-
der in vitro conditions. Recently, Wattanachai et al.
32
showed that supplementation of human liver micro-
some incubations with bovine serum AL resulted in a
3.6-fold increase in the microsomal CL
int
for paclitaxel
6"-hydroxylation, due mainly to a reduction in the
K
m
.AlowerK
m
value is in accordance with a protein-
facilitated mechanism because of the presence of ad-
ditional ionic interactions between the protein–drug
complex and microsomes compared with when the
drug is used alone in the incubation system. This
observation from Wattanachai et al.,
32
on the basis
of microsomal data, corroborates the findings of sev-
eral other authors that used hepatocytes data.
14,33–37
The reason is that liver microsomes and membranes
of hepatocytes have a similar composition of lipids to
which the protein–drug complex may potentially ag-
gregate under in vitro and in vivo conditions.
14,31,38
Overall, this is in accordance with the pharmacoki-
netics in liver of AL-bound drugs
14,33,39
and charged
macromolecules (e.g., antibodies),
40,41
which is gov-
erned by the binding to the protein and the presence
of positive charge on the protein, and, hence, is in
line with the use of the current fu
liver
in IVIVE. Con-
versely, Halifax and Houston
15
seem not in total accor-
dance with the implication of the protein-facilitated
mechanism but these authors did not provide any ro-
bust explanation to support their disagreement. Nev-
ertheless, Halifax et al.
42
studied the impact of drug
permeability in hepatocytes to understand the poor
in vitro-to-in vivo correlation of CL of highly bound
drugs. These authors demonstrated that the predic-
tion accuracy was not dependent on the relative per-
meability of drugs in hepatocytes indicating the ab-
sence of a general rate limitation by passive hepato-
cyte uptake on metabolic CL.
Protein-facilitated uptake has been reported to
be more important for AL-bound drugs than AAG-
bound drugs,
14,33–37,39
which may explain why Poulin
et al.’s
14
method shows improved predictive perfor-
mance for drugs that bind mainly to AL as compared
with the other methods studied. Inversely, drugs that
bind mainly to AAG seem to require no additional
correction, which may explain why the predictive per-
formance for AAG-bound drugs is more comparable
across the IVIVE methods tested (particularly the
Berezhkovskiy and Poulin et al.
14
methods)
14
(Table
3). These observations are reflected in the sensitivity
analysis depicted in Figure 7.
Prediction accuracy has been shown to be gener-
ally better in human than in rat (Table 2), probably
because it is expected that pharmacokinetics in hu-
mans is less complex (e.g., hepatic CL mediated by
CYP-oxidative metabolism) as a consequence of the
screening processes in drug discovery and develop-
ment. Here, we point out that the preclinical datasets
(particularly the rat datasets) might contain a signifi-
cant amount of drugs for which CL in vivo is unexpect-
edly high; this could introduce bias in the comparative
analysis. Indeed, the rat datasets present a greater
percentage of compounds with high CL as compared
with the human dataset (Appendix; see Supplemen-
tary information). Moreover, Table 2 shows that in the
rat dataset obtained from the literature the relative
predictivity of all methods is comparable, as antici-
pated for high CL drugs as illustrated in Table 4. The
rate of transporter-mediated uptake and efflux deter-
mines the rate of hepatobiliary elimination, and con-
siderable species variation in the mechanisms of bil-
iary excretion is seen in the literature.
26,43–45
Trans-
porters are, therefore, important determinants of CL
in the body.
26–28
Furthermore, a molecular weight
threshold was observed for biliary excretion, which
is different in rats and humans, particularly for an-
ionic compounds.
44
Consequently, the literature re-
ports lower predictivity of CL for drugs that poten-
tially undergo biliary excretion.
45
Accordingly, this
present study reports a difficulty in predicting drug
CL in rat of five drugs that potentially undergo bil-
iary excretion. Lam and Benet
27
also demonstrated
the effect of hepatic uptake processes on the CL es-
timate of digoxin using microsomal and hepatocytes
data. The direct scaling method has an advantage
when CL in vivo of a drug is unexpectedly high; how-
ever, careful attention should be given when the direct
scaling method is used because this method naturally
trends toward overestimation of the true metabolic
CL (AFE value > 1), as proven in this study and the
literature.
1,2,7,14
Biopharmaceutics drug disposition classification
system and/or BCS classification may help to un-
derstand the interspecies difference in the predic-
tion accuracy. Overall, only 5%–10% of drugs differ
in classification between the BCS and BDDCS. How-
ever, for class 1 drugs, the difference in classifica-
tion between BCS and BDDCS is estimated to occur
for about 40% of drugs.
16
Nevertheless, the present
study observed a relevant difference in the predic-
tion accuracy between class 1 and 2 compounds, clas-
sified according to either the BDDCS or BCS, and
this difference is much more important in the rat
dataset as compared with the human dataset. Ume-
hara and Camenisch
46
also reported differences in
CL estimations in rats across the BDDCS classes.
Thus, it appears that there are interspecies differ-
ences in the mechanisms of CL under in vivo condi-
tions, which may have influenced the current compar-
ative assessment. Alternatively, one should consider
that the BCS/BDDCS classification is based on hu-
man data and it is conceivable that classification may
change if it were based on rat data. BDDCS and BCS
class 2 compounds are expected to be governed by
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 11
COMPARATIVE ANALYSIS OF IVIVE METHODS 4319
transport effects in addition to liver metabolism, in
contrast to class 1 compounds for which transport ef-
fects should be minimal.
16
We conclude that the po-
tential interplay of transporters and enzymes must
be considered in defining the CL
int
of the liver par-
ticularly for class 2 compounds. In other words, our
hypothesis is that microsomes alone and the IVIVE
methods in their current form are insufficient to es-
timate total drug CL because hepatic transporter ef-
fects and/or other pathways (e.g., glucuronidation) are
probably not taken into account.
46,47
However, cou-
pled with the IVIVE method proposed by Umehara
and Camenisch,
46
who quantified the impact of trans-
port processes in the prediction of CL, the novel IVIVE
method for metabolism proposed by Poulin et al.
14
takes into consideration various mechanistic deter-
minants of hepatic CL.
Wan et al.
1
also reported some inaccuracies while
predicting drug CL in rat. These authors suggested
changing the value of hepatic blood flow rate in IVIVE
to correct for poor IVIVE. The CL in rats was pre-
dicted by using a blood flow rate of 100 mL/(min kg).
If we had used a value of 100 mL/(min kg) instead of
65 mL/(min kg) in the current study, higher predicted
CL values would have been seen, which would have
been beneficial for only some drugs (a high blood flow
rate generally decreased the prediction accuracy in
this study; data not shown). Nevertheless, we agree
that this aspect can be a source of discrepancy for the
rat dataset in addition to the aforementioned expla-
nations.
CONCLUSION
The analyses described here provide insight into
IVIVE application with regard to the impact of bind-
ing terms on CL prediction performance. This has im-
portant implications for the development of a mecha-
nistic and generic IVIVE method for the evaluation of
compounds in drug discovery and development. This
study confirms the trends observed in the literature
for the IVIVE methods; no systematic underpredic-
tion or overprediction of CL for the Poulin et al.
14
method, systematic underprediction of CL for the con-
ventional and Berezhkovskiy methods, and overpre-
diction of CL for the direct scaling method.
1,2,7,14,15
The results corroborate that incorporation of fu
liver
in
IVIVE leads to significant statistical improvement in
the method’s predictivity, especially for drugs bound
to AL and highly bound in plasma as well as drugs
with low CL in vivo. Data quality is crucial for
the prediction accuracy, and therefore, the preclini-
cal datasets chosen for further comparative analyses
should include only drugs for which liver metabolism
is the main route of elimination to avoid any bias.
It is expected that predicting metabolic CL may face
interspecies differences, particularly for BDDCS and/
or BCS class 2 compounds. However, consideration of
other datasets of drugs is required to further support
the conclusions of this study. Finally, a recent study
has developed a generic and mechanistic microsome
composition-based model used to predict fu
inc
from
physicochemical data only.
31
Consequently, we con-
sider that a combination of this generic model with
this present study can provide a meaningful physio-
logical model for mechanism-based predictions of CL
from IVIVE calculations. Overall, this study will help
to identify the most promising IVIVE methods avail-
able and will highlight related problems.
ACKNOWLEDGMENTS
This work represents an initiative undertaken as a
part of Dr Poulin’s research program supported by
Genentech, Inc. Professor Haddad is supported by a
Discovery Grant from the National Sciences and En-
gineering Research Council of Canada. The authors
wish to thank BinQing Wei for database mining at
Genentech, Ning Liu for generating microsomal sta-
bility data, Emile Plise and Jonathan Cheng for gen-
erating blood–plasma partitioning and MDCK per-
meability data, and Ronitte Libedinsky for editorial
help.
APPENDIX
This appendix contains the experimental settings
used for the Genentech compounds and two tables
for the drug-specific input parameters used in this
study.
MATERIALS AND METHODS USED FOR THE
GENENTECH COMPOUNDS
Potassium buffered saline (PBS) was prepared by Me-
dia Preparation Facility at Genentech, Inc. (South
San Francisco, California). Acetonitrile (ACN) and
water were purchased from Honeywell Burdick &
Jackson (Muskegon, Michigan). Propranolol, quini-
dine, tolbutamide, metoprolol, procaine, sodium hy-
droxide, phosphoric acid, human alpha1-acid glyco-
protein (AAG) were obtained from Sigma–Aldrich (St.
Louis, Missouri). All other chemicals were obtained
from commercial sources and were of the highest
qualities available. Male human plasma (K
2
EDTA)
was purchased from Bioreclamation LLC (Westbury,
New York). Male rat liver microsomes were purchased
from BD Gentest (Bedford, Massachusetts). Human
albumin (AL) was purchased from Calbiochem (San
Diego, California). The single-use rapid equilibrium
dialysis (RED) plate was purchased from Thermo
Scientific (Rockford, Illinois). The automatic 96-well
plasma stability and plasma protein binding assay
were performed using an Agilent BioCel 1200 system
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 12
4320 POULIN ET AL.
Table A1. Datasets of Drugs used for Prediction of plasma CL
Drug Name Class pKa
a
BDDCS
(BCS)
b
fu
p
a
fu
inc
a
R
BP
a
Main Binding
Protein
c
Scaled Microsomal
CL
int
(mL/min kg)
d
Plasma CL
In Vivo (mL/min
kg)
e
References
Human dataset
1 Chlorpromazine B 9.7 1 0.05 0.11 0.78 AAG 25 8.6 Obach
2
2 Propafenone B 9.74 2 0.04 0.26 0.7 AAG 166 13 Obach
2
3 Verapamil B 8.5 1 0.1 0.43 0.77 AAG 122 14.5 Obach
2
4 Diphenhydramin B 8.98 1 0.22 0.29 0.65 AAG 2.1 6.2 Obach
2
5 Lorcainide B 9.5 1 0.15 0.52 0.77 AAG 50 14 Obach
2
6 Diltiazem B 7.7 1 0.22 0.76 1.03 AAG 15 12 Obach
2
7 Amitriptyline B 9.4 1 0.05 0.15 0.86 AAG 14 10 Obach
2
8 Desipramine B 10.3 1 0.18 0.21 0.96 AAG 17 12 Obach
2
9 Imipramine B 9.5 1 0.1 0.18 1.1 AAG 19 13 Obach
2
10 Ketamine B 7.5 1 0.88 0.49 0.82 AAG 27 16 Obach
2
11 Quinidine B 10 1 0.13 0.32 0.92 AAG 3.4 2.5 Obach
2
12 Clozapine B 7.7 2 0.05 0.13 0.87 AAG 4.6 2.5 Obach
2
13 Dexamethazone N 0.32 1 0.93 AL 3 3.5 Obach
2
14 Prednisone N 2 0.25 0.2 0.83 AL 2.7 4.1 Obach
2
15 Diazepam N 1 0.013 0.28 0.71 AL 2.3 0.44 Obach
2
16 Midazolam N 1 0.05 0.88 0.53 AL 160 6.4 Obach
2
17 Alprazolam N 1 0.29 0.66 0.78 AL 1.6 1.2 Obach
2
18 Triazolam N 1 0.1 0.78 0.62 AAG 19 2.9 Obach
2
18 Zolpidem N 1 0.08 0.58 0.76 AL 2.8 4.3 Obach
2
20 Diclofenac A 4 1 0.005 1 0.55 AL 189 4.2 Obach
2
21 Ibuprofen A 4.4 2 0.01 0.84 0.55 AL 8.8 0.78 Obach
2
22 Tolbutamide A 5.27 2 0.04 0.95 0.55 AL 0.9 0.2 Obach
2
23 Warfarin A 5 2 0.01 0.47 0.55 AL 0.51 0.045 Obach
2
24 Tenidap A 3.5 0.0007 0.32 0.56 AL 8.3 0.058 Obach
2
25 Tenoxicam A 1.1; 5.5 1 0.009 0.78 0.67 AL 1.7 0.038 Obach
2
26 Amobarbital A 7.5 1 0.39 0.78 1.5 AL 0.94 0.53 Obach
2
27 Hexobarbital A 7.5 1 0.53 0.81 1 AL 2.3 3.6 Obach
2
28 Methohexital A 7.5 1 0.27 0.86 0.7 AL 49 11 Obach
2
29 PhRMA 48 A 7.2 (2) 0.0016 0.029 0.7 AL 79.2 5.1 Poulin et al.
6,14
30 PhRMA 55 N (2) 0.001 0.029 0.65 AL 559 6 Poulin et al.
6,14
31 PhRMA 56 N 0.114 0.9 0.59 AL 124.2 8.9 Poulin et al.
6,14
32 PhRMA 57 B 10; 6.5 (1) 0.0009 0.24 0.74 AL 48.6 5.3 Poulin et al.
6,14
33 PhRMA 58 N 0.5 1 0.7 AL 6.3 2.2 Poulin et al.
6,14
34 PhRMA 62 B 8.7 (1) 0.425 0.98 1 AAG 6.28 4.2 Poulin et al.
6
35 PhRMA 64 B 9.9 (1) 0.275 1 0.87 AAG 23 5.4 Poulin et al.
6
36 PhRMA 63 B 6.5 (2) 0.04 0.64 0.7 AAG 78.6 5 Poulin et al.
6,14
37 PhRMA 65 B 8.4 (1) 0.12 0.62 0.54 AAG 53.1 9 Poulin et al.
6
38 PhRMA 91 B 10.1 (2) 0.025 0.46 1 AAG 23.4 2.5 Poulin et al.
6,14
39 PhRMA 92 B 10.2 (2) 0.048 0.12 1.2 AAG 17.1 8.2 Poulin et al.
6,14
Continued
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 13
COMPARATIVE ANALYSIS OF IVIVE METHODS 4321
Table A1. Continued
Drug Name Class pKa
a
BDDCS
(BCS)
b
fu
p
a
fu
inc
a
R
BP
a
Main Binding
Protein
c
Scaled Microsomal
CL
int
(mL/min kg)
d
Plasma CL
In Vivo (mL/min
kg)
e
References
40 JNJ 1 (Lorcainide) B 9.5 1 0.15 0.45 0.7 AAG 32 17.00 DeBuck et al.
4
41 JNJ 2 (domperidone) B 7.89; 2.5 2 0.061 0.34 0.74 AAG 69.3 8.17 DeBuck et al.
4
42 JNJ 5 (Alfentanil) B 6.5 1 0.079 0.97 0.63 AAG 190 5.05 DeBuck et al.
4
43 JNJ 6 (sulfentanil) B 8.1 0.075 0.87 0.74 AAG 184 11.81 DeBuck et al.
4
44 JNJ 7 (Ketanserin) B 7.5 2 0.049 0.32 0.7 AAG 31.5 5.85 DeBuck et al.
4
and
Reimann et al.
25
45 JNJ 8 (Ritanserin) B 8.2; 2.1 2 0.008 0.45 0.65 AAG 4.91 0.51 DeBuck et al.
4
46 JNJ 9 (Sabeluzole) B 7.8; 3.4 0.016 0.06 0.84 AAG 5.1 4.05 DeBuck et al.
4
47 JNJ16 B 7.2; 3.1 0.034 0.08 0.75 AAG 20.3 8.57 DeBuck et al.
4
48 JNJ 18 (Risperidone) B 8.24; 3.1 1 0.1 0.34 0.67 AAG 7.96 5.62 DeBuck et al.
4
Dog dataset
49 PhRMA 39 N (4) 0.01 0.478 0.71 AL 494 21.8 Poulin et al.
6
50 PhRMA 49 B 9.4; 7.8 (3) 0.894 0.80 1.5 AAG 29 12.23 Poulin et al.
6
51 PhRMA 51 A 4.25 (2) 0.0246 0.27 0.56 AL 9.29 1.2 Poulin et al.
6
52 PhRMA 55 B 4.07 (2) 0.001 0.012 0.5 AL 545.7 17.53 Poulin et al.
6
53 PhRMA 56 N 0.249 0.83 1.18 AL 104.9 26.79 Poulin et al.
6
54 PhRMA 57 N (1) 0.006 0.24 0.76 AL 192.9 15.2 Poulin et al.
6
55 PhRMA 58 N 0.83 1 0.96 AL 18 5.25 Poulin et al.
6
56 PhRMA 62 B 8.65 (1) 0.45 0.98 0.77 AAG 32.5 9.77 Poulin et al.
6
57 PhRMA 63 B 6.5 (2) 0.14 0.64 0.81 AAG 84.8 11.56 Poulin et al.
6
58 PhRMA 64 B 9.86 (1) 0.48 1 1.6 AAG 59.2 18.9 Poulin et al.
6
59 PhRMA 65 B 8.4; 3 (1) 0.27 0.73 1.05 AAG 152.8 23.1 Poulin et al.
6
60 Pfizer 3 N 0.065 0.83 1 AL 78.8 28.8 Hosea et al.
5
61 Pfizer 5 N 0.178 0.73 1.1 AL 120.6 33.6 Hosea et al.
5
62 Pfizer 6 N 0.003 0.01 1 AL 172.2 12 Hosea et al.
5
63 Pfizer 12 N 0.196 0.84 1 AL 23.5 12 Hosea et al.
5
64 Pfizer 31 N 0.03 1 1 AL 20 4.3 Hosea et al.
5
Monkey dataset
65 PhRMA 51 A 4.25 (2) 0.01 0.27 0.55 AL 58.7 9.95 PhRMA
6,7
66 Pfizer 3 N 0.073 1 1 AL 470.7 30.3 Hosea et al.
5
67 Pfizer 9 N 0.023 0.73 1 AL 322.6 43 Hosea et al.
5
68 Pfizer 11 N 0.0003 0.06 0.61 AL 111.1 12 Hosea et al.
5
69 Pfizer 12 N 0.247 0.85 1 AL 39.5 6 Hosea et al.
5
70 Pfizer 31 N 0.05 1 1 AL 50 12 Hosea et al.
5
Rat dataset (literature)
71 PhRMA 38 N 0.03 0.413 0.86 AL 83.0932 29.3 Poulin et al.
6
72 PhRMA 39 N (4) 0.03 0.478 0.86 AL 8135.6 43.03 Poulin et al.
6
73 PhRMA 41 B 10.2 (1) 0.35 0.349 1.8 AAG 613.0252 87.35 Poulin et al.
6
74 PhRMA 49 B 9.4; 7.8 (3) 0.854 0.803 2.2 AAG 90.644 68.27 Poulin et al.
6
75 PhRMA 51 A 4.25 (2) 0.0191 0.27 0.63 AL 23 14.03 Poulin et al.
6
76 PhRMA 55 N (2) 0.001 0.056 0.68 AL 1246 4.48 Poulin et al.
6
77 PhRMA 56 N 0.132 1 0.79 AL 229 16 Poulin et al.
6
78 PhRMA 57 B 10; 6.5 (1) 0.001 0.24 0.78 AL 177.16 22.1 Poulin et al.
6
Continued
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 14
4322 POULIN ET AL.
Table A1. Continued
Drug Name Class pKa
a
BDDCS
(BCS)
b
fu
p
a
fu
inc
a
R
BP
a
Main Binding
Protein
c
Scaled Microsomal
CL
int
(mL/min kg)
d
Plasma CL
In Vivo (mL/min
kg)
e
References
79 PhRMA 58 N 0.81 1 0.94 AL 48.16 20.29 Poulin et al.
6
80 PhRMA 63 B 6.5 (2) 0.05 0.64 0.9 AL 425.184 53.43 Poulin et al.
6
81 PhRMA 64 B 9.86 (1) 0.41 1 1.56 AAG 691.44 125.1 Poulin et al.
6
82 PhRMA 65 B 8.4; 3 (1) 0.11 0.76 0.98 AAG 569.32 64.1 Poulin et al.
6
83 PhRMA 91 B 10.1 (2) 0.235 0.546 1.2 AAG 1454 35.7 Poulin et al.
6
84 PhRMA 92 B 10.2 (2) 0.03 0.166 1.25 AAG 1218 38.23 Poulin et al.
6
85 PhRMA 93 N (4) 0.008 0.016 1.26 AL 13 13.14 Poulin et al.
6
86 PhRMA 94 B 8.62 (2) 0.036 0.484 1.08 AAG 1419 35.2 Poulin et al.
6
87 PhRMA 95 B 9.65 (4) 0.008 0.491 1 AAG 4852.12 64.44 Poulin et al.
6
88 PhRMA 96 N (2) 0.102 0.722 0.8 AL 213.28 15.96 Poulin et al.
6
89 Alprazolam N 1 0.35 1 1.6 AL 150 76.8 Jones and Houston
3
90 Chlordiazepoxide N 1 0.15 1 1.4 AL 48 10 Jones and Houston
3
91 Clobazam N 1 0.21 1 1.3 AL 135 33 Jones and Houston
3
92 Clonazepam N 1 0.16 0.76 1.4 AL 85 19 Jones and Houston
3
93 Diazepam N 1 0.15 0.93 1.2 AL 800 67 Jones and Houston
3
94 Flunitrazepam N 1 0.25 1 1.2 AL 240 47 Jones and Houston
3
95 Midazolam N 1 0.07 0.63 1 AL 370 40.00 Jones and Houston
3
96 Triazolam N 1 0.28 0.84 1.5 AAG 850 84.00 Jones and Houston
3
97 JNJ 1 (Lorcainide) B 9.44 1 0.26 0.45 1.2 AAG 624 103.00 DeBuck et al.
4
98 JNJ 2 (Domperidone) B 7.89; 2.5 2 0.092 0.34 1.3 AAG 178 86.67 DeBuck et al.
4
99 JNJ 5 (Alfentanil) B 6.5 1 0.16 0.97 0.69 AAG 416 30.93 DeBuck et al.
4
100 JNJ 6 (Sufetanil) B 8.1 0.07 0.87 0.74 AAG 250 69.34 DeBuck et al.
4
101 JNJ 7 (Ketanserin) B 7.5 2 0.012 0.32 0.65 AAG 10 3.83 DeBuck et al.
4
102 JNJ 8 (Ritanserin) B 8.2; 2.1 2 0.02 0.45 0.74 AAG 139 26.67 DeBuck et al.
4
103 JNJ 9 (Sabeluzole) B 7.8; 3.4 1 0.02 0.06 0.84 AAG 43 35.87 DeBuck et al.
4
104 JNJ16 B 7.2; 3.1 0.036 0.08 0.78 AAG 28.2 24.70 DeBuck et al.
4
105 JNJ 18 (Risperidone) B 8.24; 3.1 1 0.118 0.34 0.85 AAG 250 64.10 DeBuck et al.
4
106 JNJ 21 B 7.27 0.015 0.23 1.5 AAG 35.6 27.30 DeBuck et al.
4
107 JNJ 22 A 8.2 0.001 0.9 0.74 AL 156 11.70 DeBuck et al.
4
108 JNJ 23 B 7; 3.1 0.082 0.06 0.8 AAG 208 33.30 DeBuck et al.
4
109 Pfizer 3 N 0.021 1 1 AL 412 16 Hosea et al.
5
110 Pfizer 5 N 0.028 0.86 1.1 AL 353.9 27 Hosea et al.
5
111 Pfizer 6 N 0.0015 0.01 1 AL 88 13 Hosea et al.
5
112 Pfizer 12 N 0.181 0.88 1 AL 28.1 14.1 Hosea et al.
5
113 Pfizer 31 N 0.03 0.87 1 AL 40 20 Hosea et al.
5
Rat dataset (Genentech)
114 Gen 1 B 7; 2.2 (2) 0.051 0.40 1.24 AL 66.00 42.4 This study
115 Gen 2 B 4.7; 8.1 (2) 0.084 0.27 1.69 AAG 24.00 20.6 This study
116 Gen 3 B 8.6 (3) 0.132 0.68 0.70 AL 37.00 34.6 This study
117 Gen 4 B 4.6; 9 (3) 0.635 0.66 1.22 AAG 16.00 42.8 This study
118 Gen 5 B 1.5; 7.5 (3) 0.172 0.37 1.87 AL 173.00 26.8 This study
119 Gen 6 B 5.6; 8.7 (2) 0.065 0.40 0.97 AAG 185.00 9.5 This study
120 Gen 7 B 8.4; 10.2 (1) 0.35 0.77 0.80 AAG 41.00 26.7 This study
121 Gen 8 B 7.7; 9.9 (2) 0.098 0.66 1.62 AL 33.00 22.9 This study
122 Gen 9 B 7; 9.6 (2) 0.148 0.70 1.51 AL 37.00 26.6 This study
Continued
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 15
COMPARATIVE ANALYSIS OF IVIVE METHODS 4323
Table A1. Continued
Drug Name Class pKa
a
BDDCS
(BCS)
b
fu
p
a
fu
inc
a
R
BP
a
Main Binding
Protein
c
Scaled Microsomal
CL
int
(mL/min kg)
d
Plasma CL
In Vivo (mL/min
kg)
e
References
123 Gen 10 B 4.3; 6.1 (2) 0.186 0.80 1.16 AL 24.00 22.2 This study
124 Gen 11 B 6.7; 7.8 (2) 0.01 0.49 0.74 AL 31.00 1.3 This study
125 Gen 12 B 4.5; 6.3 (2) 0.646 0.77 1.06 AAG 29.00 40.6 This study
126 Gen 13 B 2.4; 7.3 (2) 0.018 1.00 0.68 AAG 35.00 2.1 This study
127 Gen 14 B 5.6 (2) 0.554 0.53 0.97 AAG 28.00 40.1 This study
128 Gen 15 B 1.4; 4.9 (2) 0.068 0.27 1.05 AL 35.00 42.7 This study
129 Gen 16 B 5 (2) 0.054 0.70 0.86 AL 58.00 21.03 This study
130 Gen 17 B 4.5 (2) 0.122 0.25 1.32 AL 35.00 44.4 This study
131 Gen 18 B 5.3; 8.9 (2) 0.01 0.78 0.75 AL 27.00 2.2 This study
132 Gen 19 B 4.2 (1) 0.609 0.32 1.30 AAG 31.00 21.6 This study
133 Gen 20 B 3.3 (2) 0.026 0.50 0.70 AL 24.00 1.8 This study
134 Gen 21 B 1.9; 5.6 (2) 0.63 0.86 1.51 AAG 47.00 44.4 This study
AL, albumin; AAG, alpha1-acid glycoprotein; BDDCS, Biopharmaceutics Drug Disposition Classification System; BCS, Bioclassification System; CL, clearance; CL
int
, intrinsic clearance; fu
p
, unbound
fraction in plasma; fu
inc
, unbound fraction in incubations; PhRMA, Pharmaceutical Research and Manufacturers of America; R
BP
, blood-to-plasma concentration ratio.
a
Experimentally determined.
b
BDDCS class. If the BDDCS class is not available, the BCS class is provided in parenthesis.
6,7,14,16
For the Genentech compounds, it is an estimate of BCS because we used in vitro data not human
measured data.
c
Observed from the literature
1–7,13,16–18
or as described in the Appendix for the Genentech compounds.
d
Reported from the original sources. Scaled to in vivo condition as described in the Methods section.
e
Clearance refers to plasma kinetics; it was assumed that clearance is due mainly to hepatic metabolic clearance.
Table A2. Reduced Datasets of Drugs
Drugs Names Class pKa1
a
BDDCS
(BCS)
b
fu
a
p
fu
a
inc
R
a
BP
Main Binding
Protein
c
Scaled Microsomal CL
int
(mL/min kg)
d
Plasma CL In Vivo
(mL/min kg)
e
References
Rat dataset
135 PhRMA 37 N (1) 0.0029 0.028 0.57 AL 12.36 26.4 Poulin et al.
6
136 PhRMA 40 B 10.3(3)0.004 0.192 1.3 AAG 5.37 68 Poulin et al.
6
137 PhRMA 43 A 3.4(1)0.005 0.85 1 AL 14.43 6.9 Poulin et al.
6
138 JNJ 7 (Ketanserin) B 7.520.012 0.32 0.65 AAG 10 3.83 DeBuck et al.
4
139 Digoxin N 3 0.75 0.97 1 AL 6.42 37 Lam and Benet
27
AL, albumin; AAG, alpha1-acid glycoprotein; BDDCS, Biopharmaceutics Drug Disposition Classification System; BCS, Bioclassification System; CL, clearance; CL
int
, intrinsic clearance; fu
p
, unbound
fraction in plasma; fu
inc
, unbound fraction in incubations; PhRMA, Pharmaceutical Research and Manufacturers of America; R
BP
, blood-to-plasma concentration ratio.
a
Experimentally determined.
b
BDDCS class. If the BDDCS class is not available, the BCS class is provided in parenthesis.
6,7,14,16
c
Estimated as described in the Methods section.
d
Reported from the original sources. Scaled to in vivo condition as described in the Methods section.
e
Clearance refers to plasma kinetics; it was assumed that clearance is due mainly to hepatic uptake processes.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 16
4324 POULIN ET AL.
running the VWorks
TM
software package (Agilent
Technologies, Santa Clara, California) that integrates
multiple devices such as a liquid handler, a plate ho-
tel, a centrifuge, a plate sealer, and two incubators.
Stability in Human Plasma and Rat Liver
Microsomes
All test materials [10 mM in 100% dimethyl sul-
foxide (DMSO)] were diluted with DMSO to obtain
0.5 mM stock solutions. Four test compounds were
cassetted into one solution at 0.125 mM per com-
pound in DMSO. Human plasma and rat liver micro-
somes were thawed and diluted with PBS at pH 7.4 to
0.5 mg/mL of rat liver microsomes, 40 mg/mL human
AL, and 1.0 mg/mL human AAG. The pH of human
plasma, human AL, human AAG, and rat liver mi-
crosomes were measured and, if necessary, adjusted
to pH 7.4 with either sodium hydroxide or phospho-
ric acid. Aliquots (396 :L) of human plasma, human
AL, human AAG, and rat liver microsomes were dis-
pensed individually into 0.5 mL Axygen plates (Axy-
gen Inc. Union City, California) and spiked with 4 :L
of 0.5 mM working solutions. The final incubation con-
centration of test materials was 1.25 :Mwith0.1%
DMSO. After mixing, samples were incubated at 37
C,
and samples (40 :L) were withdrawn at 0, 1, 2, and
3 h and diluted with 40 :L PBS. The samples were
quenched with 150 :L ACN containing 2.5 nM pro-
pranolol (internal standard) to stop the reaction. Sam-
ples were centrifuged at 1000g for 10 min. Super-
natants (100 :L) were transferred to a 96-well analy-
sis plate and diluted with 100 :L of water prior to
liquid chromatography–tandem mass spectrometry
(LC–MS/MS) analysis. Procaine and metoprolol were
used as positive and negative controls, respectively.
Determination of Protein Binding Using 96-Well
RED
Human plasma (100%), human AL (40 mg/mL), hu-
man AAG (1 mg/mL), and rat liver microsomes
(0.5 mg/mL) were thawed. Human AL, human AAG,
and rat liver microsomes (0.5 mg/mL) were prepared
in PBS at pH 7.4. Aliquots (396 :L) of human plasma,
human AL, human AAG, and rat liver microsomes, as
well as PBS at pH 7.4, were dispensed individually
into 0.5 mL Axygen plates (Axygen Inc.) and spiked
with 4 :L of 0.5 mM stock solutions in DMSO. The
final incubation concentration of test materials (com-
pounds) was 5 :M with 0.1% DMSO. After mixing,
aliquots (300 :L) of mixture were placed into the sam-
ple chamber of the single-use RED plate as the donor
(indicated by the red ring), and aliquots (500 :L) of
PBS were dispensed into the adjacent chamber as the
receiver (indicated by the white ring), both in dupli-
cate. The plate was covered with a clear plate cover
top and incubated with a Liconic shaker at 37
Cfor
4h.
The binding ratio of human AAG to human AL
was determined. An aliquot (396 :L) of human AL
and human AAG were dispensed into 0.5 mL Axy-
gen plates (Axygen Inc.) and spiked with 4 :Lof
0.5 mM stock solutions in DMSO. After mixing,
aliquots (300 :L) of mixture (AL spiked with com-
pound) were placed into the sample chamber of the
single-use RED plate as the donor (indicated by the
red ring), and aliquots (500 :L) of AAG (spiked with
compound) were dispensed into the adjacent cham-
ber as the receiver (indicated by the white ring),
both in duplicate. After the 4-h incubation, 40 :L
aliquots from the receiver (white chamber) and 4 :L
from the donor (red chamber) were withdrawn and
added to ACN (150 :L) containing 2.50 nM propra-
nolol (internal standard). An equal volume of blank
human plasma, 40 mg/mL human AL, 1.0 mg/mL hu-
man AAG, 0.5 mg/mL of rat liver microsomes, or PBS
was added to the receiver wells, and 36 :Lofblank
human plasma, 40 mg/mL human AL, 1.0 mg/mL hu-
man AAG, 0.5 mg/mL of rat liver microsomes, or PBS
was added to donor wells to create analytically iden-
tical sample matrices to minimize matrix effect. Sam-
ples were centrifuged at 1000g for 10 min, and super-
natants (100 :L) were transferred to a 96-well anal-
ysis plate together. Water (100 :L) was added to the
samples prior to LC–MS/MS analysis. Tolbutamide
and quinidine were used as controls.
LC–MS/MS Analysis
The samples were analyzed using a Cohesive LX-
2 Transcend Multiplexing system with Agilent
1100 series High-performance liquid chromatogra-
phy (HPLC) pumps from Agilent Technologies and an
HTS PAL autosampler from CTC Analytics (Carrboro,
North Carolina) connected to a Sciex API4000-QTrap
mass spectrometer (Foster City, California) in the pos-
itive ion mode, with a turbo ion spray source. Multiple
reaction monitoring was used to quantify test materi-
als. Aliquots (10 :L) of the samples were injected onto
a Hypersil Gold C
18
Column (1.9 :m, 2.1 × 20 mm
2
)
from Thermo Electron Corporation (San Jose, Cal-
ifornia). Chromatographic separation was achieved
using a solvent system consisting of mobile phase A
(HPLC-grade water containing 5 mM ammonium ac-
etate with 0.1% acetic acid) and mobile phase B (ACN
containing 0.1% acetic acid). The initial flow rate was
0.5 mL/min. The initial gradient condition of 1% mo-
bile phase B was held for 0.2 min, then ramped to 98%
mobile phase B in 0.8 min, and held at 98% mobile
phase B for 1.4 min at a flow rate of 0.7 mL/min. The
gradient was returned to initial conditions of 1% mo-
bile phase B at a flow rate of 0.5 mL/min and held at
these conditions for 0.6 min to equilibrate the system.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 17
COMPARATIVE ANALYSIS OF IVIVE METHODS 4325
Data Processing and Analysis
Data collection was conducted with Analyst 1.5 soft-
ware, and data processing was performed with Multi-
quant 2.0. The percentage of protein binding or micro-
somal binding in each matrix was calculated from the
area ratio of the analyte detected (normalized to inter-
nal standard) from the receiver side responses to the
donor side multiplied by 100. Then 100% binding was
used to calculate the unbound fraction (fu). The bind-
ing ratio of AAG to AL was calculated from the area
ratio of the analyte detected (normalized to internal
standard) from the receiver side (AAG) responses to
the donor side (AL). When AAG/AL is dialysed against
each other, a ratio greater than 0.6 suggests AAG is
preferred.
REFERENCES
1. Wan H, Bold P, Larsson LO, Ulander J, Peters S, Lofberg B,
Ungell AL, Nagard M, Llinas A. 2010. Impact of input param-
eters on the prediction of hepatic plasma clearance using the
well-stirred model. Curr Drug Metab 11:583–594.
2. 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.
3. Jones HM, Houston JB. 2004. Substrate depletion approach
for determining in vitro metabolic clearance: Time dependen-
cies in hepatocyte and microsomal incubations. Drug Metab
Dispos 32:973–982.
4. De Buck SS, Sinha VK, Fenu LA, Nijsen MJ, Mackie CE,
Gilissen R. 2007. Prediction of human pharmacokinetics us-
ing physiologically-based pharmacokinetics modeling: A ret-
rospective analysis of 26 clinically tested drugs. Drug Metab
Dispos 35:1766–1780.
5. Hosea NA, Collard WT, Cole S, Maurer TS, Fang RX, Jones
H, Kakar SM, Nakai Y, Smith BJ, Webster R, Beaumont K.
2009. Prediction of human pharmacokinetics from preclinical
information: Comparative accuracy of quantitative prediction
approaches. J Clin Pharm 49:513–33.
6. Poulin P, Jones HM, Jones RDO, Yates JWT, Gibson CR, Chien
JY, Ring BJ, Adkison KK, He H, Vuppugalla R, Marathe P,
Fischer V, Dutta S, Sinha VK, Bj¨ornsson T, Lav´e T, Ku MS.
2011. PhRMA CPCDC initiative on predictive models of hu-
man pharmacokinetics. 1. Goals, properties of the PhRMA
dataset and comparison with literature datasets. J Pharm Sci
100:4050–4073.
7. Ring B, Chien JY, Adkison KK, Jones HM, Rowland M, Jones
RDO, Yates JWT, Ku MS, Gibson CR, He H, Vuppugalla R,
Marathe P, Fischer V, Dutta S, Sinha VK, Bj
¨
ornsson T, Lav
´
e
T, Poulin P. 2011. PhRMA CPCDC initiative on predictive
models of human pharmacokinetics. 3. Comparative assess-
ment of prediction methods of human clearance. J Pharm Sci
100:4090–4110.
8. Tamaki S, Komura H, Kogayu M, Yamada S. 2011. Compar-
ative assessment of empirical and physiological approaches
on predicting human clearances. J Pharm Sci 100:1147–
1155.
9. Pelkonen O, Turpeinen M. 2007. in vitro–in vivo extrapo-
lation of hepatic clearance: Biological tools, scaling factors,
model assumptions, and correct concentrations. Xenobiotica
37:1066–1089.
10. Beaumont K, Gardner I, Chapman K, Hall M, Rowland
M. 2011. Towards an integrated human clearance pre-
diction strategy that minimizes animal use. J Pharm Sci
100:4518–4535.
11. Foster JA, Houston BJ, Halifax D. 2011. Comparison of in-
trinsic clearances in human liver microsomes and suspended
hepatocytes from the same donor livers: Clearance-dependent
relationship and implications for prediction of in vivo clear-
ance. Xenobiotica 41:124–136.
12. Berezhkovskiy LM. 2010. The corrected traditional equations
for calculation of hepatic clearance that account for the dif-
ference in drug ionisation in extracellular and intracellular
tissue water and the corresponding corrected PBPK equation.
J Pharm Sci 100:1167–1183.
13. Berezhkovskiy LM, Liu N, Halladay JS. 2011. Consistency
of the novel equation for determination of hepatic clearance
and drug time course in liver that account for the difference in
drug ionization in extracellular and intracellular tissue water.
J Pharm Sci 101:516–518.
14. Poulin P, Kenny JR, ECA Hop, Haddad S. 2012. in vit-
ro–in vivo extrapolation of clearance: Modeling hepatic
metabolic clearance of highly bound drugs and comparative
assessment with existing methods. J Pharm Sci 101:838–
851.
15. Halifax D, Houston BJ. 2012. Evaluation of hepatic clearance
prediction using in vitro data: Emphasis of fraction unbound
in plasma and drug ionisation using a dataset of 107 drugs.
J Pharm Sci 101:2645–2652.
16. Benet LZ, Broccattelli F, Oprea TI. 2011. BDDCS applied to
over 900 drugs. AAPS J 32:1311–1336.
17. Paillard M, Sraer JD, Claret M, Favier MP. 1970. Mea-
surement of intracellular pH in several tissues by using the
micro-electrode specifically sensible. J Physiol (Paris) 62(Suppl
3):423–425.
18. Rothe KF, Heisler N. 1986. Correction of metabolic alkalosis
by HCL and acetazolamide; effects on extracellular and in-
tracellular acid-base status in rats in vivo. Acta Anaesthesiol
Scand 30:566–570.
19. Rothe VF. 1984. New aspects of acid-base balance: Influences
of plasma pH variation on intracellular tissue pH in vivo.
Fortschr Med 46:158–157.
20. Rodgers T, Rowland M. 2006. Physiologically based phar-
macokinetic modelling 2: Predicting the tissue distribution of
acids, very weak bases, neutrals and zwitterions. J Pharm Sci
95:1238–1257.
21. Peyret T, Poulin P, Krishnan K. 2010. A unified algorithm
for predicting partition coefficients for PBPK modeling of
drugs and environmental chemicals. Toxicol Appl Pharmacol
249:197–207.
22. Gordon AH, Humphrey JH. 1961. Measurement of intracellu-
lar albumin in rat liver. Biochem J 78:551–556.
23. Park R, Leach WJ, Arieff AL. 1980. Measurement of the
liver extracellular space in vivo in dogs. Horm Metab Res
12:680–684.
24. Poulin P, Theil FP. 2009. Development of a novel method for
predicting human volume of distribution at steady-state of ba-
sic drugs and comparative assessment with existing methods.
J Pharm Sci 98:4941–4961.
25. Reimann IW, Okonkwo PO, Klotz U. 1990. Pharmacokinetics
of ketanserin in man. Eur J Clin Pharmacol 25:73–76.
26. Meuldermans W, Hendrickx J, Lauwers W, Swysen E, Hurk-
mans R, Knaeps F, Woestenborghs R, Heykants J. 1984. Ex-
cretion and biotransformation of ketanserin after oral and in-
travenous administration in rats and dogs. Drug Metab Dispos
12:772–781.
27. Lam JL, Benet LZ. 2004. Hepatic microsome studies
are insufficient to characterize hepatic metabolic clearance
and metabolic drug–drug interactions: Studies of digoxin
metabolism in primary rat hepatocytes versus microsomes.
Drug Metab Dispos 32:1311–1336.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012
Page 18
4326 POULIN ET AL.
28. Shitara Y, Horie T, Sugiyama Y. 2006. Transporters as a
determinant of drug clearance and tissue distribution. Eur J
Pharm Sci 27:425–446.
29. Sohlenius-Sternbeck AK, Jones C, Ferguson D, Middleton BJ,
Projean D, Floby E, Bylund J, Afzelius L. 2012. Practical use
of the regression offset approach for the prediction of in vivo in-
trinsic clearance from hepatocytes. Xenobiotica. [Epub ahead
of print.] DOI: 10.3109/00498254.2012.669080.
30. Nagilla R, Frank KA, Jolivette LJ, Ward KW. 2006. Investi-
gation of the utility of published in vitro intrisinc clearance
data for prediction of in vivo clearance. J Pharmacol Toxicol
Methods 53:106–116.
31. Poulin P, Haddad S. 2011. Microsome composition-based
model as a mechanistic tool to predict nonspecific binding of
drugs in liver microsomes. J Pharm Sci 100:4501–4517.
32. Wattanachai N, Polasek TM, Heath TM, Uchaipichat V,
Wongwiwat T, Tassaneeyakul W, Miners JO. 2011. in vit-
ro–in vivo extrapolation of CYP2C8-catalyzed paclitaxel 6"-
hydroxylation: Effects of albumin on in vitro kinetic parame-
ters and assessment of interindividual variability in predicted
clearance. Eur J Clin Pharmacol 67:815–824.
33. Burczynski FJ, Wang GQ, Elmhadoun B, She YM, Roberts MS,
Standing KG. 2001. Hepatocyte palmitate uptake: Effect of
albumin surface charge modification. Can J Physiol Pharmacol
79:868–875
34. Qin M, Nilsson M, Øie S. 1994. Decreased elimination of
drug in the presence of alpha1-acid-glycoprotein is related
to a reduced hepatocytes uptake. J Pharmacol Exp Ther
269:1176–1181.
35. Bilello JA, Biello PA, Stellrecht K, Leonard J, Norbeck DW,
Kempf DJ, Robins T, Drusano GL. 1996. Human serum alpha
acid glycoprotein reduces uptake, intracellular concentration,
and antiviral activity of A-80987, an inhibitor of the human
immunodeficiency virus type 1 protease. Ant Microb Agent
Chemother 40:1491–1497.
36. Osterloh K, Ewert U, Pries AR. 2002. Interaction of al-
bumin with the endothelial cell surface. Am J Physiol
283:H398–H405.
37. Weisiger RA. 1985. Dissociation from albumin: A potentially
rate-limiting step in the clearance of substances by the liver.
Proc Natl Acad Sci U S A 82:1563–1567.
38. Gabellec MM, Steffan AN, Dodeur M, Durang G, Kirn A, Rebel
G. 1983. Membrane lipids of hepatocytes, kupffer cells and
endothelial cells. Biochem Biophys Res Commun 113:845–853.
39. Mitchell SJ, Huizer-Pajkos A, Cogger VC, McLachlan AJ, Le
Couteur DG, Hilmer SN. 2011. Poloxamer 407 increases the
recovery of paracetamol in the isolated perfused rat liver. J
Pharm Sci 100:334–340.
40. Boswell AC, Tesar DB, Mukhyala K, Theil FP, Fielder PJ,
Khawli LA. 2010. Effects of charge on antibody tissue dis-
tribution and pharmacokinetics. Bioconjug Chem 21:2153–
2163.
41. Igawa T, Tsunoda H, Tachibana T, Maeda A, Mimoto F,
Moriyama C, Nanami M, Sekimori Y, Nabuchi Y, Aso Y, Ha-
torri K. 2010. Reduced elimination of IgG antibodies by en-
gineering the variable region. Protein Eng Des Sel 23:385–
392.
42. Halifax D, Turlizzi E, Zanelli U, Houston BJ. 2012. Clearance-
dependent underprediction of in vivo intrinsic clearance from
human hepatocytes: Comparison with permeabilities from ar-
tificial membrane (PAMPA) assay, in silico and caco-2 assay,
for 65 drugs. Eur J Pharm Sci 45:570–574.
43. Levine WG. 1978. Biliary excretions of drugs and other xeno-
biotics. Ad Rev Pharmcol Toxicicol 18:81–96.
44. Yang X, Gandhi YA, Duignan DB, Morris ME. 2009. Predic-
tion of biliary excretion in rats and humans using molecular
weight and quantitative structurepharmacokinetic relation-
ships. AAPS J 11:511–524.
45. Mahmood I, Sachajwalla C. 2002. Interspecies scaling of
biliary excreted drugs. J Pharm Sci 91:1908–1914.
46. Umehara K, Camenisch G. 2012. Novel in vitro–in vivo ex-
trapolation (IVIVE) method to predict hepatic organ clearance
in rat. Pharm Res 29:603–617.
47. Gardiner P, Paine SW. 2011. The impact of hepatic uptake on
the pharmacokinetics of organic anions. Drug Metab Dispos
39:1930–1938.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 101, NO. 11, NOVEMBER 2012 DOI 10.1002/jps
Page 19
  • Source
    • "Equation (3) represents the intrinsic in vitro clearance CL int,LM , Equation (4) scales CL int,LM to the full liver, while Equation (5) assumes the direct scaling approach as representative of in vivo hepatic clearance (Poulin et al., 2012), although this approach may overestimate the in vivo clearance in case of extensive plasma protein binding. In these equations, the following scaling factors were used: 34 mg/g microsomal proteins per g liver, 0.71 mL/min/g liver for hepatic clearance (Q H ) and a relative liver mass of 2.6 g liver/kg bodyweight (Lipscomb and Poet, 2008). "
    [Show abstract] [Hide abstract] ABSTRACT: Tris(1-chloro-2-propyl) phosphate (TCIPP) is an emerging contaminant which is ubiquitous in the indoor and outdoor environment. Moreover, its presence in human body fluids and biota has been evidenced. Since no quantitative data exist on the biotransformation or stability of TCIPP in the human body, we performed an in vitro incubation of TCIPP with human liver microsomes (HLM) and human serum (HS). Two metabolites, namely bis(2-chloro-isopropyl) phosphate (BCIPP) and bis(1-chloro-2-propyl) 1-hydroxy-2-propyl phosphate (BCIPHIPP), were quantified in a kinetic study using HLM or HS (only BCIPP, the hydrolysis product) and LC-MS. The Michaelis–Menten model fitted best the NADPH-dependent formation of BCIPHIPP and BCIPP in HLM, with respective VMAX of 154 ± 4 and 1470 ± 110 pmol/min/mg protein and respective apparent Km of 80.2 ± 4.4 and 96.1 ± 14.5 µM. Hydrolases, which are naturally present in HLM, were also involved in the production of BCIPP. A HS paraoxonase assay could not detect any BCIPP formation above 38.6 ± 10.8 pmol/min/µL serum. Our data indicate that BCIPP is the major metabolite of TCIPP formed in the liver. To our knowledge, this is the first quantitative assessment of the stability of TCIPP in tissues of humans or any other species. Further research is needed to confirm whether these biotransformation reactions are associated with a decrease or increase in toxicity.
    Full-text · Article · Oct 2015 · Chemosphere
  • Source
    • "The simulations can be done using different approaches, such as well-stirred model, parallel tube and dispersion models. Currently, new models extend to more comprehensive predictions that take into account protein binding, transport phenomena and concentration gradients in the liver247248249250. Overall, clearance predictions can be used to predict drug concentrations in plasma, particularly at steady state, but these in vivo translation methods have not yet been used for the data from experiments with stem cell derived or 3D hepatocyte cultures. "
    [Show abstract] [Hide abstract] ABSTRACT: Cultured cells are widely used in the evaluation of new drugs and drug delivery systems. Cells can be grown at different levels of complexity ranging from simple reductionist models to complex organotypic models. The models are based on primary, secondary or stem cell derived cell cultures. Generation of tissue mimics with cultured cells is a difficult task, because the tissues have well-defined morphology, complex protein expression patterns and multiple inter-linked functions. Development of organotypic cell culture models requires proper biomaterial matrix and cell culture protocols that are able to guide the cells to the correct phenotype. This review illustrates the critical features of the cell culture models and, then, selected models are discussed in more detail (epidermal, corneal epithelial, retinal pigment epithelium, and hepatocyte models). The cell models are critically evaluated paying attention to the level of characterization and reliability of in vivo translation. Properties of the cell models must be characterized in detail using multiple biological assays and broad sets of model drugs. Robust in vivo predictions can be achieved with well-characterized cell models that are used in combination with computational methods that will bridge the gap between in vitro cell experiments and physiological situation in vivo in the body.
    Full-text · Article · Jun 2014 · Journal of Controlled Release
  • Source
    • "Typically, the well stirred model is used for estimation of hepatic clearance (Houston, 1994). The Poulin method showed better accuracy (Poulin et al., 2012a) compared with other in vitro-in vivo extrapolation methods for hepatic clearance estimation, especially for low-clearance compounds highly bound to albumin. Faldaprevir fits this profile. "
    [Show abstract] [Hide abstract] ABSTRACT: Faldaprevir is an HCV protease inhibitor which effectively reduces viral load in patients. Since faldaprevir exhibits slow metabolism in vitro and low clearance in vivo, metabolism was expected to be a minor clearance pathway. The human [(14)C]ADME study revealed that two mono-hydroxylated metabolites (M2a and M2b) were the most abundant excretory metabolites in feces, constituting 41% of the total administered dose. In order to de-convolute formation and disposition of M2a and M2b in human and determine why the minor change in structure (addition of 16 amu) produced chemical entities that were excreted and not present in circulation, multiple in vitro test systems were employed. The results from these in vitro studies clarified the formation and clearance of M2a and M2b. Faldaprevir is metabolized primarily in liver by CYP3A4/5 to form M2a and M2b, which are also substrates of efflux transporters (P-gp and BCRP). The role of transporters is considered important for M2a and M2b as they demonstrate low permeability. It is proposed that both metabolites are efficiently excreted via bile into feces and do not enter the systemic circulation to an appreciable extent. If these metabolites permeate to blood, they can be readily taken up into hepatocytes from the circulation by uptake transporters (likely OATPs). These results highlight the critical role of drug metabolizing enzymes and multiple transporters in the process of the formation and clearance of faldaprevir metabolites. Faldaprevir metabolism also provides an interesting case study for metabolites which are exclusively excreted in feces, but are of clinical relevance.
    Preview · Article · Dec 2013 · Drug metabolism and disposition: the biological fate of chemicals
Show more