IVIVC in Oral Absorption for Fenofibrate Immediate
Release Tablets Using a Dissolution/Permeation System
PHILIPP BUCH,1 PETER LANGGUTH,1 MAKOTO KATAOKA,2 SHINJI YAMASHITA2
1
Department of Pharmaceutical Technology and Biopharmaceutics, Johannes Gutenberg-University, Staudinger Weg 5,
55099 Mainz, Germany
2
Faculty of Pharmaceutical Sciences, Setsunan University, Nagatoge-cho 45-1, Hirakata, Osaka 573-0101, Japan
Received 6 May 2008; revised 7 July 2008; accepted 14 August 2008
Published online 14 October 2008 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21576
ABSTRACT: The usefulness of a dissolution/permeation (D/P) system to predict the
in vivo performance of solid dosage forms containing the poorly soluble drug, fenofibrate,
was studied. Biorelevant dissolution media simulating the fasted and fed state condi-
tions of the human gastrointestinal tract were used in order to simulate the effect of food
on the absorption of fenofibrate. Moreover, the results obtained from the D/P system
were correlated with pharmacokinetic parameters obtained following in vivo studies in
rats. The in vitro parameter (amount permeated in the D/P system) reflected well the
in vivo performance in rats in terms of AUC and Cmax of fenofibric acid. This study thus
demonstrates the potential of the D/P system as valuable tool for absorption screening of
dosage forms for poorly soluble drugs. ß 2008 Wiley-Liss, Inc. and the American Pharmacists
Association J Pharm Sci 98:2001 2009, 2009
Keywords: biopharmaceutics classification system (BCS); Caco-2 cells; dissolution;
permeation; in vitro/in vivo correlations (IVIVC); food-effect-studies
INTRODUCTION dosage forms today is mainly based on in vitro
dissolution testing whereas permeability screen-
The number of lipophilic active substances which ing is done by cell monolayer permeation assays
exhibit poor water solubility is still increasing in (e.g., Caco-2). Although these methods have been
pharmaceutical companies. Poor water solubility used for years they still exhibit shortcomings
is a serious problem which frequently results in because during permeability studies only pure
unpredictable in vivo performance in oral absorp- compounds are tested. Neither dosage form
tion, especially when the formulation is given in excipients nor physiological solubilizing sub-
the fasted state. stances are applied which could alter the perme-
It has always been a challenge for formulation ability of the drug. Furthermore, the high drug
scientists to overcome the issue of low bio- concentrations in the donor solution are far from
availability. In that respect, a screening system mimicking clinical doses in that they tend to be
which can predict or at least rank order in vivo constant over time in contrast to the in vivo
performance would be time- and cost-saving. situation. Also in vitro dissolution tests may not
Determination of in vitro behavior of solid oral reflect the in vivo situation in GI tract for drug
dissolution.
Therefore the necessary step was to create an
Correspondence to: Peter Langguth (Telephone: þ49-6131-
in vitro system to determine a drug s permeability
3925746; Fax: þ49-6131-3925021;
and solubility simultaneously under physiological
E-mail: [e-mail])
conditions for prediction of drug absorption
Journal of Pharmaceutical Sciences, Vol. 98, 2001 2009 (2009)
capacity in vivo. During the development of new
ß 2008 Wiley-Liss, Inc. and the American Pharmacists Association
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009 2001
2002 BUCH ET AL.
chemical entities or new formulations of lipophilic operated under fed and fasted dissolution condi-
drugs an in vitro screening method for predicting tions would predict changes in the food effect
in vivo performance would be helpful. To that depending on the formulation employed.
effect few systems have been developed over the
last decade. A continuous dissolution/Caco-2
permeation system was first evolved by Ginski MATERIALS AND METHODS
et al.,1 Kobayashi et al.,2 and Sugawara et al.3
focused on simulating the pH changes in the GI Materials
tract, whereas Kataoka et al.4 invented a dis-
The Caco-2 cell line was purchased from American
solution/permeation (D/P) system to which they
Type Culture Collection (Rockville, MD) at
applied an amount of drug corresponding to the
passage 17. Dulbecco s modified Eagle medium
clinical dose. Motz et al.5,6 used a flow through
(D-MEM) was purchased from Sigma-Aldrich
setup to determine dissolution and permeation
(St. Louis, MO). Nonessential amino acids
concomitantly. All these systems are based on a
(NEAA), fetal bovine serum (FBS), L-glutamate,
Caco-2 cell monolayer to determine the drug s
trypsin (0.25%)-EDTA (1 mM) and antibiotic-
permeability. Most in vitro systems are equipped
antimycotic mixture (10,000 U/mL penicillin G,
with a pump to deliver the dissolved drug from
10,000 mg/mL streptomycin sulfate and 25 mg/mL
the dissolution chamber to the permeation mod-
amphotericin B in 0.85% saline) were purchased
ule. The application of a pump complicates the
from Gibco Laboratories (Lenexa, KS). Egg-
handling because different flow rates alter the
phosphatidylcholine (lecithin) was purchased
drug s permeated amount. Pressure variations
from NOF Corp. (Tokyo, Japan). Sodium taur-
caused by pulsating pumps may damage the
ocholate (NaTC) and bovine serum albumin (BSA)
pressure sensitive Caco-2 cell monolayer.
were purchased from WAKO Pure Chemical
The equipment by Yamashita and coworkers
Industries, Ltd. (Osaka, Japan). All other chemi-
consists of two half-chambers (one for dissolution,
cals were of analytical reagent grade.
one as a receiver compartment) with a Caco-2
monolayer mounted in between. By using fasted
and fed state simulating intestinal fluids (FaS-
Preparation of Caco-2 Monolayer
SIFmod and FeSSIFmod6.5) as apical solutions, the
effect of food intake on the oral absorption of
Caco-2 cells were cultured in humidified air with
poorly water-soluble drugs has been studied.7 11
5% CO2 at 378C in culture flasks (Nippon Becton
Furthermore, sticking of lipophilic drug to the
Dickinson Co., Ltd., Tokyo, Japan) using D-MEM
tubing material is a frequently observed compli-
supplemented with 10% FBS, 1% L-glutamate, 1%
cation. This phenomenon is often seen with
NEAA and 5% antibiotic-antimycotic solution.
compounds of BCS class II, for which the presence
Cells were harvested with trypsin-EDTA and
of food may lead to an increase in their solubility,
seeded at a density of 3 105 cells/filter onto
dissolution and bioavailability.12
polycarbonate filters (0.3 mm pores, 4.20 cm2
We tested several formulations with the D/P
growth area) inside a cell culture insert (Nippon
system and used FaSSIFmod and FeSSIFmod6.5 as
Becton Dickinson Co., Ltd.). Culture medium
apical media. All formulations contained fenofi-
(1.5 mL in the insert and 2.6 mL in the well) was
brate. Fenofibrate can be classified as BCS class II
changed every 48 h for the first 6 days and every
drug which exhibits a pronounced food effect on its
24 h thereafter. Cells were used for the experi-
bioavailability.
ments between days 18 and 21 postseeding.
The main purpose of the present study was to
validate a D/P system in terms of a possible
correlation between in vitro results and in vivo
D/P System
data from rats. We expected the fasted state to be
most discriminative for the compositions as For these studies the previously described D/P
fenofibrate shows a food effect and is only poorly system was used.4 In short, the Caco-2 monolayer
absorbed in the fasted state.13 Therefore, is fixed between two chambers, the donor and the
attempts were made to correlate the in vitro data receiver chamber. In the donor chamber the
obtained using FaSSIFmod as the apical medium tested formulation dissolves, the dissolved drug
with in vivo data from fasted rats. A further permeates across the Caco-2 monolayer, and from
objective was to evaluate whether the D/P system the receiver chamber samples are taken for
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009 DOI 10.1002/jps
IVIVC FOR IR TABLETS 2003
determining the concentration change over time. To estimate the in vivo absorption the in vitro
As master solution Hank s balanced salts solution permeated amount of both fenofibrate and feno-
including 19.45 mM glucose and 10 mM HEPES fibric acid (% of dose/2 h) in the D/P system was
was used (transport medium, TM). used in the following equation:
FaSSIFmod and FeSSIFmod6.5 were based on
Absmax PAg
TM supplemented with NaTC 3 mM and Lecithin
Predicted absorption ð%Þ Åº
0.75 mM or five times higher concentrations PAg þ PAg
50
of both substances, respectively.14,15 The
donor chambers contained 8 mL FaSSIFmod or Absmax is the maximum absorption (defined as
FeSSIFmod6.5, both adjusted to pH 6.5. 5.5 mL of 100%), PA is the cumulative in vitro permeated
TM containing 4.5% w/v BSA (pH 7.4) was used as amount in the D/P system (% of dose/2 h), PA50 is
receiver medium in each receiver chamber.16 All the permeated amount, which corresponds to
the solutions and chambers were preheated to 50% of the absorption in vivo, and g is a Hill
378C and maintained at that temperature. The coefficient. PA50 and g were obtained by a fitting
solutions in both chambers were stirred at procedure from the permeated amount (PA) of
200 rpm with magnetic stirrers (8 mm 2 mm drugs in the D/P system and their oral absorption
in size). in human by using MULTI program developed by
Yamaoka et al.17 The values for PA50 and g were
0.334 0.096 (SD) and 0.883 0.178 in the fasted
Permeability Studies in the D/P System
state, and 0.121 0.021 and 1.334 0.198 in the
Before the Caco-2 monolayer was fixed between fed state, respectively.
the chambers of the D/P system, it was incubated
for 20 min in the culture well with TM (pH 6.5) for
the apical and TM containing 4.5% w/v BSA Analysis of In Vitro Samples by HPLC
(pH 7.4) for the receiver side. After positioning the
All samples were analyzed using a reversed-phase
Caco-2 monolayer with support filter in the D/P
HPLC system (LC-10A, Shimadzu Co., Kyoto,
system FaSSIFmod or FeSSIFmod6.5 were filled in
Japan) with a binary pump, a vacuum degasser,
the apical chamber and receiver medium was
an autosampler, a column oven and a variable
filled in the receiver chamber. Drug products or
wavelength ultraviolet detector (SPD-10A,
pure drugs were added to the apical side as
Shimadzu Co.). The mobile phase was composed
powder. The fenofibrate dose tested in vitro
of 50 mM phosphate buffer (pH 2.5) (A) and
amounted to 1.45 mg, which corresponds to 1/
acetonitrile (B) and the flow rate was 1.0 mL/min.
100 of the clinically taken dose of 145 mg. Samples
All drugs were trapped and eluted from the
(200 mL) from the receiver solution were taken
analytical column (J sphere ODS-H80 75
over 2 h and replaced by fresh medium. For HPLC
4.6 mm ID, YMC, Kyoto, Japan) held at 408C
analysis only the deproteinized supernatants
using the following gradient conditions: 0 2.0 min,
(1.3 mL) of the samples containing BSA were
50% A; 2.0 3.0 min, 50 30% A; 3.0 10.0 min, 30%
used which were obtained by adding 1.4 mL of
A; and 10.0 13.0 min, 50% A. Fenofibrate and
methanol followed by centrifugation (15,000 rpm,
fenofibric acid were detected at a wavelength of
20 min, 208C). After all samples from the receiver
288 nm.
compartment were gathered, apical solution was
taken and filtered through cellulose acetate filter
(pore size, 0.2 mm) for determining the final
Animals
concentration of the dissolved drug. Fenofibrate
and fenofibric acid concentrations were both All animal experiments were approved by
determined in all samples as Caco-2 cells meta- the Ethical Review Committee of Setsunan
bolized only a fraction of fenofibrate. University. Wistar male rats weighing 250 g were
The transepithelial electric resistance (TEER) deprived of food with free access to water for 18 h
values of the cell monolayers were monitored before the experiments. Before the formulations
before and after the experiments by Millicell-ERS were administered to the rats they were pulveriz-
(MILLIPORE, Billerica, MA). The experiment s ed in a mortar if necessary and suspended in
conditions had no effect on the integrity of the cell 0.5% methyl cellulose aqueous solution. Each rat
layers, as TEER-values did not decrease signifi- was administered a volume of 1 mL containing
cantly and were higher than 300 V cm2. fenofibrate in a concentration which corresponds
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009
2004 BUCH ET AL.
to the clinical dose (145 mg/70 kg). Each formu- polyvinyl alcohol, titanium dioxide, talc, soybean
lation was administered to three rats. lecithin and xanthan gum. It was used as
At 0.5, 1, 2, 3, 4, 8, 24 h after administration, the obtained.
rats were anesthetized using ether to collect blood Formulation C: This formulation contains
samples (0.6 mL) from the jugular vein which was micronized fenofibrate capsules (HEXAL1
visualized through an incision on the shoulder. 160 mg Hartkapseln) containing also dimethi-
Blood samples were centrifuged and plasma was cone-emulsion, gelatin, hypromellose, cornstarch,
frozen at 708C for subsequent measurement of sodium lauryl sulfate, simethicone-emulsion,
fenofibric acid. sucrose, talc, iron oxides and titanium dioxide.
Formulation D: Consists of lipid microparticles
which were produced by a combination of atomiz-
Analysis of Plasma Samples ing and congealing.20 22 A UP200S ultrasonic
processor (Hielscher, Teltow, Germany) was used
All plasma samples were pretreated with 1.4 mL
in a temperature controlled environment (48C).
acetonitrile and centrifugation (15,000 rpm,
The ultrasonic processor was placed upside down
20 min and 208C) and the deproteinized super-
so that the sonotrode was on top. Softisan 138 and
natant (1.3 mL) was used to determine the
fenofibrate (15% w/w) were melted at 908C under
amount of fenofibric acid. All samples were
stirring.23 29 Thereafter, the melt was pumped
analyzed with a reversed-phase HPLC system
(feeding rate 2.7 mL/min) through a jacketed pipe
(LC-VP, Shimadzu Co.) consisting of a binary
at 808C onto the oscillating top of the sonotrode.
pump, a vacuum degasser, an autosampler, a
The top of the metal pipe was equipped with a
column oven and a mass-spectrometer equipped
metal ring which slightly touches the sonotrode s
with the electrospray ionization interface (LCMS-
surface. When the liquid touched the sonotrode,
2010A, Shimadzu Co.). The mobile phase was
the melt was atomized by ultrasound energy into
composed of 0.1% formic acid in water (A) and
small droplets, which subsequently solidify upon
acetonitrile (B) and the flow rate was 0.5 mL/min.
cooling. The metal ring which protrudes over the
Fenofibric acid was trapped and eluted from
top of the sonotrode s surface forced the small
the analytical column (J sphere ODS-H80
droplets to fall to the ground and not to be sprayed
75 4.6 mm ID, YMC) held at 408C using the
onto the pipe or any other part of the instrument.
following gradient conditions: 0 1.0 min, 95% A;
Eventually, the particles were sieved into a size
1.0 1.5 min, 95 5% A; 1.5 4.0 min, 5% A; and 4.0
fraction of 100 160 mm and stored at 48C.
6.0 min, 95% A. Selected ion monitoring was used
Formulation E: Consisted of micronized fenofi-
for detection of protonated molecules of fenofibric
brate (5 mm average particle size) which was
acid (m/z 319.00). Data processing was carried out
purchased from Labochim (Milan, Italy).
using LCMS Solution software (Shimadzu Co.).
Formulation F: Consisted of bulk fenofibrate
Standard noncompartmental pharmacokinetic
(28 mm average particle size) which was pur-
parameters were calculated using the pharmaco-
chased from Labochim.
kinetic program Topfit.18
Formulations Employed RESULTS AND DISCUSSION
Formulation A: This formulation was based on the
In Vitro Results
solid dispersion principle (MeltDose1-technique)
and was used as delivered by the manufacturer.19 The fraction dissolved in the donor medium is
It contains fenofibrate dispersed in a hydrophilic given in Table 1. The rank order of the fractions
polymer and compressed to tablets following of dose dissolved from the tested formulations
mixing with magnesium stearate, Avicel PH 200 was as follows: formulation B > formulation
and Polyplasdone XL. A > formulation C > formulation D > formulation
Formulation B: This formulation is a fenofibrate E > formulation F for FaSSIFmod as well as for
nanoparticulate tablet formulation (TriCor1), FeSSIFmod6.5.30,31 In the case of dissolution in
containing fenofibrate, hypromellose 2910, doc- FaSSIFmod, formulations A and B exhibited equal
usate sodium, sucrose, sodium lauryl sulfate, results. In FeSSIFmod6.5 more fenofibrate was
lactose monohydrate, silicified microcrystalline dissolved after 2 h from formulation B compared
cellulose, crospovidone, magnesium stearate, to A. This could be due to the inability of the filter
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009 DOI 10.1002/jps
IVIVC FOR IR TABLETS 2005
Table 1. Fraction of Fenofibrate Dose Dissolved After mined in the D/P system using FaSSIFmod and
2 h (mean SD, n 3) in the Apical Chamber of the D/P
FeSSIFmod6.5 as apical media (Fig. 1a and b). The
System
food effect of fenofibrate34 which is seen in
humans was particularly visible in the permea-
FaSSIFmod (%) FeSSIFmod6.5 (%)
tion profiles of the pure powders obtained by the
Formulation A 7.49 0.46 17.92 0.76 D/P system.
Formulation B 7.67 0.52 25.47 0.87
The calculated absorptions of fenofibrate from
Formulation C 5.33 0.36 11.76 0.94
tested formulations are summarized in Table 2.
Formulation D 1.54 0.35 8.67 3.85
Formulations E and F resulted in >3 times higher
Formulation E 0.74 0.07 5.22 1.23
absorption when tested in FeSSIFmod6.5 compared
Formulation F 0.71 0.22 4.65 0.72
to FaSSIFmod. This is in accordance with phar-
macokinetic data published previously (the extent
materials to separate nanoscopic material from of absorption varies from 30 50% to 60 90% when
dissolved material. For formulation B fenofibrate the drug is taken in the fasting state or when it is
powder is downsized by wet-milling.32 In the final given after a meal, respectively).13
tablet the particle size is in the nanometer range, The smaller the ratio between the predicted
such that some of the particles may pass through absorptions using FaSSIFmod and FeSSIFmod6.5,
the cellulose acetate filter with a pore size of the more effective should the formulation reduce
0.2 mm following tablet disintegration. This could the food effect. The performances of formulations
explain the equal or in the case of FeSSIFmod6.5 B and A showed that these formulations could
better dissolution results for formulation B diminish this ratio significantly, as seen in
compared to A, and thus, most likely may Table 2. As expected, all formulations revealed
contribute to an overestimation of the true higher calculated absorption when FeSSIFmod6.5
fraction dissolved of formulation B in the apical was used as apical solution in the D/P system.
chamber of the D/P system. These results also With formulation A the best results not only in
clearly demonstrate the advantage of the D/P minimizing the ratio but also in increasing the
system in which the oral absorption of drugs from permeated amount of fenofibrate and fenofibric
various formulations can be evaluated not only acid in FaSSIFmod (Fig. 1a) and FeSSIFmod6.5
from the dissolution but also from the permeation (Fig. 1b) were achieved. Both formulations D and
process without the necessity of having to C performed as poor as the pure fenofibrate
separate colloidal particles. By taking the solubi- powders concerning the goal of reducing the food
lity of fenofibrate into consideration (14.3 effect, but they raised the calculated absorption
1.05 mg/mL in FaSSIFmod and 36.79 0.39 mg/ compared to the pure powders. The rank order of
mL in FeSSIFmod6.5, respectively, at 378C) it the drug formulations with respect to permeated
becomes obvious that all experiments were run amount of fenofibrate and fenofibric acid followed
under nonsink conditions.33 almost completely the rank order in terms of
The permeation profiles of fenofibrate from fraction dissolved in the donor medium except for
all drug products and pure powders were deter- the A/B pair (explanation see above).
Table 2. Fraction of Fenofibrate Dose Permeated After 2 h (Mean, n 3) Into the Receiver Chamber of the D/P
System, the Predicted Absorptions and the Ratio between the Predicted Absorptions, All Based on the Data Obtained
from the D/P System Using FaSSIFmod and FeSSIFmod6.5
Permeated Amount Predicted Absorption
(% of Dose/2 h) (% of Dose)
FaSSIFmod FeSSIFmod6.5 FaSSIFmod FeSSIFmod6.5 Ratio (FeSSIFmod6.5/FaSSIFmod)
Formulation A 0.304 0.625 47.94 88.95 1.86
Formulation B 0.177 0.379 36.32 82.09 2.26
Formulation C 0.077 0.276 21.5 75 3.49
Formulation D 0.043 0.153 14.64 57.71 3.94
Formulation E 0.031 0.087 10.97 39.29 3.58
Formulation F 0.02 0.062 7.82 29.22 3.74
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009
2006 BUCH ET AL.
Figure 2. Average ( SEM) fenofibric acid plasma
concentrations in rats (n ź 3) after oral administration
of different fenofibrate formulations. Key: formulation
A(~), formulation B (!), formulation C (*), formula-
tion D (&), formulation E ( ), formulation F (*).
Waal attraction forces between nonpolar mole-
cules could lead to aggregation and agglomeration
during the process of homogenization. This pheno-
menon of insufficient wettability can explain a
slower dissolution rate leading to lower bioavail-
ability of formulation E.35 38
Formulation A gave the highest peak concen-
tration (Cmax) for the fenofibric acid in rats. That is
in accordance with the D/P system s permeation
Figure 1. Effect of different formulations on permea-
profile obtained from formulation A. In FaSSIFmod
tion of fenofibrate in the D/P system with FaSSIFmod (a)
the sum of permeated fenofibrate and fenofibric
and FeSSIFmod6.5 (b) as apical medium. Each data point
acid was already after 15 min significantly higher
is the mean SEM of three to six independent experi-
( p < 0.05) for formulation A compared to the
ments. Key: formulation A (~), formulation B (!),
permeation values of the other formulations.
formulation C (*), formulation D (&), formulation E
( ), formulation F (*).
In Vitro In Vivo Correlation
Pharmacokinetics in Plasma
Formulation A showed the best results both based
The plasma concentrations versus time profiles of on calculated absorption and in vivo AUC0 24 h
fenofibric acid following oral administration are values in rats, respectively. Comparing the results
shown in Figure 2. The pharmacokinetic para- obtained in the D/P system using FaSSIFmod with
meters following noncompartmental analysis are the in vivo results in fasted rats (calculated
listed in Table 3. In agreement with the present absorptions vs. AUC0 24 h values illustrated in
in vitro results, formulation A presented the best Fig. 3) the ranking of the drug products were the
intestinal absorption enhancement. The rank same. However, the spread of the predicted extent
order of the formulations based on the AUC0 24 h of absorption from the D/P system (14.64 47.94%
values was formulation A > formulation B > in FaSSIFmod) for the fenofibrate drug products
formulation C > formulation D > formulation F > tested was much larger than the spread of the
formulation E. Except for the pure powders this exposure data in vivo (38.98 100%, relative to
reflects the results obtained with the D/P system. formulation A). This could be due to the fact that
Formulation F had a higher AUC0 24 h value than the rat has no gall bladder and thus secrets bile
formulation E. The formulations were fed to the constantly.39 44 Zwart et al.45 compared bile
rats as a suspension. The higher surface energy of flow and total bile salt rate throughout the day
fine particles and subsequent stronger van der in human and in rat (hepatic bile flow: in humans
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009 DOI 10.1002/jps
IVIVC FOR IR TABLETS 2007
Table 3. Mean Pharmacokinetic Parameters ( SD, n ź 3) Following Oral
Administration of Fenofibrate Formulations at Equal Doses of 2.1 mg/kg
Cmax (ng/mL) tmax (h) AUC0 24 h (ng h/mL)
Formulation A 9635.3 2829.87 2.33 1.15 64,135.45 15,613.33
Formulation B 6265.64 1823.49 1 52,224.76 10,958.09
Formulation C 7179.1 1004.52 2.33 1.53 45,892.51 4306.44
Formulation D 2226.47 424.47 3.33 0.58 24,997.64 9493.6
Formulation E 737.86 177.29 1.67 1.15 8109.88 2333.59
Formulation F 1496.7 472.49 2.33 0.58 18,673.64 3162.45
2 22 mL/day/kg, in rats 48 92 mL/day/kg). CONCLUSIONS
Following their collected data the rat might not
be the right model to predict absorption in the In conclusion, it appears that the D/P system in
fasted state since the rat releases bile independent combination with biorelevant media can be used
of food intake and in a higher amount which to predict relevant biopharmaceutic properties of
makes the fasted situation in rat and in human different fenofibrate formulations in rats. Assays
incomparable. This could be an explanation of the in the D/P system with modified FaSSIF and
decreased differences between the drug products FeSSIF should be useful in formulation develop-
in the fasted rat compared to the differences in the ment, especially for lipophilic drugs which are
in vitro data obtained with FaSSIFmod in the known to show a positive food effect.
present study. In Figure 3, correlation between Further work will be necessary in order to
the calculated absorptions using FeSSIFmod6.5 and challenge the D/P system for its predictability of
AUC0 24 h values in vivo in fasted rats was also formulation performance in humans.
incorporated. These results might indicate that,
from the point of view of dissolution of poorly
soluble drugs, GI tract conditions in fasted rats
ACKNOWLEDGMENTS
are close to that in human under fed state rather
than under fasted state. This would be the next
P.B. received a scholarship from the German
issue for our research to investigate the species
Academic Exchange Service (DAAD) during his
differences in the food effect on oral drug
stay in Hirakata. The authors thank Dr. Per Holm
absorption.
(LifeCycle Pharma) and Dr. Dieter Scherer
(ApisPharma) for constructive comments and for
the supply of the RD 24 formulation.
REFERENCES
1. Ginski MJ, Taneja R, Polli JE. 1999. Prediction of
dissolution-absorption relationships from a contin-
uous dissolution/Caco-2 system. AAPS PharmSci
1:E3.
2. Kobayashi M, Sada N, Sugawara M, Iseki K, Miya-
zaki K. 2001. Development of a new system for
prediction of drug absorption that takes into
account drug dissolution and pH change in the
gastro-intestinal tract. Int J Pharm 221:87 94.
Figure 3. Plot of predicted absorptions of all fenofi- 3. Sugawara M, Kadomura S, He X, Takekuma Y,
brate formulations calculated from D/P data obtained Kohri N, Miyazaki K. 2005. The use of an in vitro
with FaSSIFmod and FeSSIFmod6.5 versus AUC0 24 h dissolution and absorption system to evaluate oral
values for all fenofibrate formulations. Key: formulation absorption of two weak bases in pH-independent
A(~), formulation B (!), formulation C (*), formula- controlled-release formulations. Eur J Pharm Sci
tion D (&), formulation E (þ), formulation F (*). 26:1 8.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009
2008 BUCH ET AL.
4. Kataoka M, Masaoka Y, Yamazaki Y, Sakane T, 18. Heinzel G, Woloszczak R, Thomann P. TopFit:
Sezaki H, Yamashita S. 2003. In vitro system to Pharmacokinetic and Pharmacodynamic Data
evaluate oral absorption of poorly water-soluble Analysis System for the PC, G. Fischer. 1993.
drugs: Simultaneous analysis on dissolution and 19. Holm P, Norling T. 2006. A tablet comprising a
permeation of drugs. Pharm Res 20:1674 1680. fibrate. WO 2006084475.
5. Motz SA, Klimundová J, Schaefer UF, Balbach S, 20. Cavallari C, Rodriguez L, Albertini B, Passerini N,
Eichinger T, Solich P, Lehr CM. 2007. Automated Rosetti F, Fini A. 2005. Thermal and fractal ana-
measurement of permeation and dissolution of pro- lysis of diclofenac/Gelucire 50/13 microparticles
pranolol HCl tablets using sequential injection ana- obtained by ultrasound-assisted atomization. J
lysis. Anal Chim Acta 581:174 180. Pharm Sci 94:1124 1134.
6. Motz SA, Schaefer UF, Balbach S, Eichinger T, 21. Passerini N, Perissutti B, Moneghini M, Voinovich
Lehr CM. 2007. Permeability assessment for solid D, Albertini B, Cavallari C, Rodriguez L. 2002.
oral drug formulations based on Caco-2 monolayer Characterization of carbamazepine-Gelucire 50/13
in combination with a flow through dissolution cell. microparticles prepared by a spray-congealing
Eur J Pharm Biopharm 66:286 295. process using ultrasounds. J Pharm Sci 91:699
7. Charman WN, Porter CJH, Mithani S, Dressman 707.
JB. 1997. Physicochemical and physiological mech- 22. Rodriguez L, Passerini N, Cavallari C, Cini M,
anisms for the effects of food on drug absorption: Sancin P, Fini A. 1999. Description and preliminary
The role of lipids and pH. J Pharm Sci 86:269 282. evaluation of a new ultrasonic atomizer for spray-
8. Custodio JM, Wu CY, Benet LZ. 2008. Predicting congealing processes. Int J Pharm 183:133 143.
drug disposition, absorption/elimination/transpor- 23. Hauss DJ. 2007. Oral lipid-based formulations. Adv
ter interplay and the role of food on drug absorption. Drug Deliv Rev 59:667 676.
Adv Drug Deliv Rev 60:717 733. 24. Humberstone AJ, Charman WN. 1997. Lipid-based
9. Fleisher D, Li C, Zhou Y, Pao LH, Karim A. 1999. vehicles for the oral delivery of poorly water soluble
Drug, meal and formulation interactions influenc- drugs. Adv Drug Deliv Rev 25:103 128.
ing drug absorption after oral administration. 25. Jannin V, Musakhanian J, Marchaud D. 2008.
Clinical implications. Clin Pharmacokinet 36:233 Approaches for the development of solid and
254. semi-solid lipid-based formulations. Adv Drug
10. Persson EM, Gustafsson AS, Carlsson AS, Nilsson Deliv Rev 60:734 746.
RG, Knutson L, Forsell P, Hanisch G, Lennernas H, 26. Jaspart S, Piel G, Delattre L, Evrard B. 2005. Solid
Abrahamsson B. 2005. The effects of food on the lipid microparticles: Formulation, preparation,
dissolution of poorly soluble drugs in human and in characterisation, drug release and applications.
model small intestinal fluids. Pharm Res 22:2141 Expert Opin Drug Deliv 2:75 87.
2151. 27. O Driscoll CM, Griffin BT. 2008. Biopharmaceutical
11. Welling PG. 1996. Effects of food on drug absorp- challenges associated with drugs with low aqueous
tion. Annu Rev Nutr 16:383 415. solubility The potential impact of lipid-based for-
12. Dressman JB, Amidon GL, Reppas C, Shah VP. mulations. Adv Drug Deliv Rev 60:617 624.
1998. Dissolution testing as a prognostic tool for 28. Porter CJ, Pouton CW, Cuine JF, Charman WN.
oral drug absorption: Immediate release dosage 2008. Enhancing intestinal drug solubilisation
forms. Pharm Res 15:11 22. using lipid-based delivery systems. Adv Drug Deliv
13. Dollery C. 1991. Therapeutic drugs, Vol. 1. Edin- Rev 60:673 691.
burgh: Churchill Livingstone. 29. Pouton CW. 2006. Formulation of poorly water-
14. Kataoka M, Masaoka Y, Sakuma S, Yamashita S. soluble drugs for oral administration: Physicochem-
2006. Effect of food intake on the oral absorption of ical and physiological issues and the lipid formula-
poorly water-soluble drugs: In vitro assessment of tion classification system. Eur J Pharm Sci 29:278
drug dissolution and permeation assay system. 287.
J Pharm Sci 95:2051 2061. 30. Hanafy A, Spahn-Langguth H, Vergnault G, Gre-
15. Patel N, Forbes B, Eskola S, Murray J. 2006. Use of nier P, Tubic Grozdanis M, Lenhardt T, Langguth
simulated intestinal fluids with Caco-2 cells and rat P. 2007. Pharmacokinetic evaluation of oral fenofi-
ileum. Drug Dev Ind Pharm 32:151 161. brate nanosuspensions and SLN in comparison to
16. Yamashita S, Furubayashi T, Kataoka M, Sakane conventional suspensions of micronized drug. Adv
T, Sezaki H, Tokuda H. 2000. Optimized conditions Drug Deliv Rev 59:419 426.
for prediction of intestinal drug permeability using 31. Shepherd J. 1994. The fibrates in clinical practice:
Caco-2 cells. Eur J Pharm Sci 10:195 204. Focus on micronised fenofibrate. Atherosclerosis
17. Yamaoka K, Tanigawara Y, Nakagawa T, Uno T. 110:55 63.
1981. A pharmacokinetic analysis program (multi) 32. Peterson ML, Hickey MB, Zaworotko MJ, Almars-
for microcomputer. J Pharmacobiodyn 4:879 885. son Ö. 2006. Expanding the scope of crystal form
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009 DOI 10.1002/jps
IVIVC FOR IR TABLETS 2009
evaluation in pharmaceu tical science. J Pharm LX. 2006. Why is it challenging to predict intestinal
Pharm Sci 9:317 326. drug absorption and oral bioavailability in human
33. Dai WG, Pollock-Dove C, Dong LC, Li S. 2008. using rat model. Pharm Res 23:1675 1686.
Advanced screening assays to rapidly identify solu- 40. Chiou WL, Barve A. 1998. Linear correlation of
bility-enhancing formulations: High-throughput, the fraction of oral dose absorbed of 64 drugs
miniaturization and automation. Adv Drug Deliv between humans and rats. Pharm Res 15:1792
Rev 60:657 672. 1795.
34. Corsini A, Bellosta S, Davidson MH. 2005. Phar- 41. DeSesso JM, Jacobson CF. 2001. Anatomical and
macokinetic interactions between statins and physiological parameters affecting gastrointestinal
fibrates. Am J Cardiol 96:44 49. absorption in humans and rats. Food Chem Toxicol
35. Adam A, Schrimpl L, Schmidt PC. 2000. Some 39:209 228.
physicochemical properties of mefenamic acid. 42. Fagerholm U. 2007. Prediction of human pharma-
Drug Dev Ind Pharm 26:477 487. cokinetics Gastrointestinal absorption. J Pharm
36. Perrut M, Jung J, Leboeuf F. 2005. Enhancement of Pharmacol 59:905 916.
dissolution rate of poorly-soluble active ingredients 43. Merle L, Dangoumau J, Balabaud C. 1978. Effect of
by supercritical fluid processes Part I: Microniza- food and light schedule on bile flow in the rat. Cell
tion of neat particles. Int J Pharm 288:3 10. Mol Life Sci 34:764 765.
37. Rasenack N, Müller BW. 2002. Dissolution rate 44. Zhao YH, Abraham MH, Le J, Hersey A, Luscombe
enhancement by in situ micronization of poorly CN, Beck G, Sherborne B, Cooper I. 2003. Evalua-
water-soluble drugs. Pharm Res 19:1894 1900. tion of rat intestinal absorption data and correla-
38. Swanepoel E, Liebenberg W, de Villiers MM, Dek- tion with human intestinal absorption. Eur J Med
ker TG. 2000. Dissolution properties of piroxicam Chem 38:233 243.
powders and capsules as a function of particle size 45. Zwart LL, Rompelberg CJM, Sips A, Welink J,
and the agglomeration of powders. Drug Dev Ind Engelen JGM. 1999. Anatomical and physiological
Pharm 26:1067 1076. differences between various species used in studies
39. Cao X, Gibbs ST, Fang L, Miller HA, Landowski CP, on the pharmacokinetics and toxicology of xenobio-
Shin HC, Lennernas H, Zhong Y, Amidon GL, Yu tics. Rev Lit, RIVM report 623860010.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 6, JUNE 2009