Pharmacokinetics, disposition and biotransformation of [14C]-radiolabelled valsartan in healthy male volunteers after a single oral dose

Article abstract of DOI:10.1080/004982597240767

1. The disposition of valsartan, a potent angiotensin II receptor antagonist, was investigated in six healthy male volunteers. They each received a single oral dose of 80 mg of a ¹⁴C-labelled preparation as a neutral buffered solution. 2. Peak

Full text of DOI:10.1080/004982597240767

xenobiotica , 1997, vol. 27, no. 1, 59±71
Pharmacokinetics, disposition and
biotransformation of [14C]-radiolabelled valsartan in
healthy male volunteers after a single oral dose
È
F. WALDMEIER ‹*, G. FLESCH ‹, P. MULLER ‹,
È
T. WINKLER Ž, H.- P. KRIEMLER Ž, P. BUHLMAYER
and M. DE GASPARO
s
s
Pharma Research, Preclinical Safety, Drug Metabolism & Exploratory Toxicology
‹
Ciba-Geigy, CH- 4002 Basle, Switzerland
‹
Switzerland
Pharma Research, Bioanalytics & Pharmacokinetics Ciba-Geigy, CH- 4002 Basle,
‹
Geigy, CH- 4002 Basle, Switzerland
Pharma Medicine & Clinical Development Department, Human Pharmacology Ciba-
Ž
Physics Department, Spectroscopy Ciba-Geigy, CH- 4002 Basle, Switzerland
Pharma Research Ciba-Geigy, CH- 4002 Basle, Switzerland
s
Received 22 July 1996
1. The disposition of valsartan, a potent angiotensin II receptor antagonist, was
investigated in six healthy male volunteers. They each received a single oral dose of 80 mg
"%
of a C-labelled preparation as a neutral buŒered solution.
2. Peak concentrations of radioactivity and valsartan in plasma measured 1 h after
dosing showed rapid onset of absorption. The results of this study combined with other
available data indicate that at least 51 % of the dose was absorbed.
3. Valsartan was the predominant radioactive compound in plasma. Elimination of
b
valsartan and radioactivity was fast and multiexponential. -Half-lives of 6 1 h were
observed. In a terminal elimination phase, low radioactivitylevels decreased with a half-life
³
of81 33 h. A minor, pharmacologicallyinactive metabolite (valeryl-4-hydroxy-valsartan;
³
M1) was detected in the plasma at time points later than 2 h after dosing, representing
AUC
approximately 11 % of the
of plasma radioactivity.
(
h)
#%
4. The bulk of the dose was excreted within 4 days. The total excretion within 7 days
amounted to 99 1% of dose. Faecal excretion was predominant (86 5% of dose).
Valsartan was largely excreted unchanged (81 5% of the dose in the excreta). The
³
³
³
predominant clearance mechanism appeared to be direct elimination via bile.
5. An inactive metabolite, M1, was formed by oxidative biotransformation and
accounted for 9 3% of the dose in the excreta.
³
Introduction
Valsartan (CGP 48933, (S )-N-valeryl-N-[2 -(1H-tetrazol-5-yl)biphenyl- 4-yl)-
«
methyl]- valine; ®gure 1) is an orally active, potent and speci®c angiotensin II
receptor antagonist acting on the AT receptor subtype. It binds to vascular smooth
"
muscle cell membranes with high a nity (K 2 38 nmol l). Binding is saturable and
±
}
i
reversible with a dissociation half-life of 56 min. In contrast to the structurally
related compound losartan, valsartan does not bind to sites unrelated to the AT
"
receptor (de Gasparo and Whitebread 1995). Twelve- month toxicity studies showed
no adverse eŒects at the highest doses of 200 mg kg in rat and 120 mg kg in
marmoset (Criscione et al. 1996).
}
}
In healthy volunteers, valsartan was well tolerated following single oral doses up
to 400 mg and after multiple dosing with 200 mg per day. The basic pharmaco-
* Author for correspondence.
’
0049±8254 97 $12 00
±
1997 Taylor & Francis Ltd
}

60
F. Waldmeier et al.
kinetics of valsartan in healthy volunteers have been studied after oral and
intravenous doses (Flesch et al. 1996). Concentrations of valsartan in plasma and
urine have been measured using hplc (Brunner et al. 1994) and non-renal clearance
was found to be predominant (Flesch et al. 1996). The aim of the present study was
to investigate the absorption, pharmacokinetics, biotransformation and excretion of
valsartan in man. For this purpose, valsartan was administered in a radiolabelled
form.
Materials and methods
Subjects
Six healthy normotensive male Caucasian subjects (33±53 years old, 66±85 kg body weight)
participated in the study and were selected based on a prestudy screen with routine clinical laboratory
tests. All were phenotyped as extensive debrisoquine metabolizers (marker substrate for cytochrome
!
which is near the margin to poor metabolizer status. Subject 1 was phenotyped as a slow acetylator,
P4502D6 isoenzyme activity), with a metabolic ratio
1. Subject 2 showed a metabolic ratio of 5 4,
±
subject 3 as a fast acetylator. All subjects gave written informed consent.
Study design
The study was conducted according to the Declaration of Helsinki (with Venice and Hong Kong
amendments) and Good Clinical (Research) Practice. The study protocol was approved by an Ethics
"%
Review Board. Each fasted volunteer received a single oral dose of 80 mg of a C-radiolabelled
preparation as a drinking solution in buŒer. The 80-mg dose was chosen since this was in the range of the
anticipated therapeutic dosage. Therapeutically , valsartan will be given as a solid dosage form for oral
administration, but a solution was chosen for this study because the preparation process for the normal
solid form could not be scaled down to the size of the radiolabelled batches.
Subjects were fasted for 12 h prior to administration, and light standardized meals were provided 2,
4 and 8 h thereafter. Alcoholwas not allowed throughout the study. The observation period of subject 1
was 168 h (pilot study), and that of the subjects 2±6 was 216 h. Blood was collected (see below) to assess
the kinetics of valsartan and radioactivity, and excreta were collected quantitatively to determine routes
and balance of excretion of valsartan and metabolites.
"%
C-radiolabelled test drug
"%
"%
C]valsartan was synthesized at the Isotope Laboratory of Ciba-Geigy, Basle. The C-radiolabel
[
was located in the carbon atom of the tetrazole ring (®gure 1). Batches with a speci®c radioactivity of
"%
42 7±50 kBq mg were prepared and or puri®ed shortly before use. The radiochemical purity of C-
±
labelled valsartan was 98±99 % (radio-hplc as described below, with isocratic elution).
}
}
Dose administration
After collecting baseline (blank)urine, faeces and blood samples, the test dose was administered. Each
"%
volunteer received an exactly weighed single oral dose of approximately 80 mg C-labelled valsartan
(3 4±4 0 MBq). Each dose was dissolved, shortly before dosing, in 50 ml freshly prepared phosphate
±
±
±
buŒer solution (0 05 mol l K HPO KH PO , pH 7 5). The volunteers drank the trial compound
±
}
}
#
%
#
%
solution, the bottles were rinsed with 50 ml buŒer and the volunteers drank the washing followed by
200 ml tap water.
Sample collectionand storage
During the ®rst 12 h after dosing, venous blood was collected from a forearm vein through a catheter
into heparinized evacuated tubes. At later times, blood was collected by vein puncture. Samples were
collected before dose administration (blank), at 10, 20, 30, 45 min, 1, 2, 3, 4, 5, 6, 8, 12 and 24 h after
administration. Subsequently blood was collected in 24-h intervals up to 168 h (subject 1) or 216 h (nos
2±6). Plasma was separated by centrifugation. Urine and faeces were collected in portions at the following
time intervals : before dose administration (blank), 0±6 h, 6±12, 12±24 and subsequently in 24-h intervals
up to 168 h after dosing. Urine and faeces were collected quantitatively into plastic bottles and boxes
respectively. Twelve hours after dosing the subjects left the clinical site. They returned to the clinical
laboratories for blood collection and to hand over the collected excreta. During the collection intervals,

Disposition of valsartan in man
61
"%
Figure 1. Structural formula of C-radiolabelled valsartan. * Position of the radiolabel.
the samples were kept refrigerated at 2±8 C. In the analyticallaboratory,the samples were processed and
°
°
stored immediately at 20 C until analysis. Samples were shipped to other analyticallaboratoriesunder
dry ice.
®
Chemicals
"
isopropyl-5-[(2 -( -tetrazol-5yl)-biphenyl-4-yl)-methyl]- pyrimidine ; for structure formula, see
n
99 %) and CGP 48791 (6-( -butyl)-4-hydroxy-2-
Non-radiolabelled valsartan (chemical purity
1H
. 1994) used for analytical reference were from Ciba-Geigy. Organic solvents, tetra-
«
et al
hydrofuran (THF) and tri¯ uoroacetic acid (TFA), all of analytical grade, were purchased from
Brunner
l
commercial manufacturers. Silica (Si 60, particle size 40±63 m) was from Merck (Darmstadt,
Germany), solid phase extraction cartridges Chromabond C18 were from Macherey Nagel (Dueren,
Helix pomatia b
-glucuronidase activity
Germany), glucuronidase aryl-sulphatase (crude solution from
;
100 000 Fishman units ml) was purchased from Boehringer Mannheim (Germany).
}
}
Determination of radioactivity concentrations
Processing of biological samples and determination of radioactivity concentrations by liquid
et al
l
l
scintillation counting were as described earlier (Waldmeier
each (four for faeces) were analysed.
. 1991). Three replicates of 400±500
Determination of valsartan in plasma by hplc-¯ uorimetry and in urine by hplc-UV
Concentrations of valsartan in plasma were determined as follows: plasma samples were treated with
1H
n
0 1 ml 1 mol l HCl and internal standard (CGP 48791 ; 6-( -butyl)-4-hydroxy-2-isopropyl-5-[(2 -(
-
«
±
}
etal
tetrazol-5yl)-biphenyl- 4-yl)-methyl]- pyrimidine) was added(Brunner
. 1994). Valsartan and internal
standard were extracted by solid-phase Bond-Elut CH cartridges and eluted with methanol. Concen-
trations of valsartan were determined by high-performance liquid chromatography (hplc) on a
l
250 4 mm i.d. column packed with Nucleosil C18 10 m, with 0 1% TFA in acetonitrile H O (45 55
¬
±
}
}
[v v] ; ¯ ow 1 ml min) and with ¯ uorimetric detection (excitation wavelength 265 nm, # emission
}
}
LOQ
l
) was 0 15 mol l. The method was linear in the
±
wavelength 378 nm). The limit of quantitation (
l
}
concentration range 0 15±12 mol l.
±
Urine samples were processed and analysed asdescribed except thatUV detection at 215 nm was used
}
l
instead of ¯ uorimetry. The method was linear in the concentration range 1 5±40 mol l.
±
}
LOQ
l
1 5 mol l.
±
The methods were validated daily by analysis of drug-free human plasma and urine samples spiked
¯
}
with known amounts of valsartan.
Radio-hplc
Hplc with radiomonitoring (radio-hplc) was performed on a column 250 4 mm i.d. packed with
¬
l
LiChrospher RP18e 5 m (100 A; Merck). A solvent gradient was formed using asolvent delivery system
PU- 980 (Jasco, Tokyo, Japan) from solvent A, 0 1 % aq. TFA, and B, 0 1% TFA in acetonitrile, with a
±
±
¯ ow of 1 ml min. The gradient started with 98 % solvent A and 2% solvent B, changed linearly to 20 %
}
A and 80 % B over 30 min, and ended with 20% A, 80 % B at 35 min. Radioactivitywas detected using
a Ramona D detector (Raytest, Straubenhard, Germany) or a Berthold detector 506C (Berthold,
Wildbad, Germany) or by liquid scintillation counting of collected fractions (0 5 ml). UV- absorption was
±
monitored with a Spectro¯ ow 757 detector (Kratos Analytical Instruments, Ramsey, NJ, USA) at
250 nm.

62
F. Waldmeier et al.
Metabolite patterns in plasma
Selected plasma samples (1 6 ml) were spiked with non-radioactive valsartan. Proteins were
±
precipitated by dropwise additionof6 ml acetonitrile, followed by centrifugation and washing ofthe solid
in vacuo
hplc via a precolumn 25 4 mm i.d. ®lled with LiChrospher RP18e 5 m. The recovery of radioactivity
with 0 5 ml acetonitrile. The combined supernatants were concentrated
l
and applied to radio-
±
¬
from sample processing and hplc was essentially quantitative.
Alternatively, single plasma samples (0 3 ml) were directly subjected to radio-hplc via sample loop
±
without preparatory processing. The hplc column was protected by a4 4 mm LiChrocart cartridge with
l
LiChrospher 300 RP-8 10 m (guard column).
¬
Metabolite patterns in urine and faeces
Samples of pooledurine (0±48 h; 5 ml) were applied to solid phase extraction cartridges Chromabond
C18 (0 5 g). After a water washing, radioactivity was eluted with 5 ml methanol (recovery 95 %). After
in vacuo
, the residue was dissolved in hplc solvent A and subjected to radio-hplc via
±
removing the solvent
a sample loop. The mean total recovery of radioactivity from sample processing and hplc was 93 %.
Samples of 1±3 g pooled faeces homogenates (typical pool period 12±96 h) were spiked with non-
radioactive valsartan and extracted with 5±10 ml methanol by shaking for 30 min followed by
centrifugation. After washing the solid with methanol, the clear supernatants were combined and
in vacuo
concentrated , diluted with water, and subjected to radio-hplc via a sample loop. The recovery
of radioactivity from sample processing and hplc was 92±100 %.
Isolation of metaboliteM1 from faeces
The faeces collected in the period 24±72 h from subject 4 were lyophilized. The residue was
homogenized and extracted with 600 ml methanol and the extract was ®ltered. The ®ltrate was extracted
n
in vacuo
, and applied to a column (4 30 cm) packed with silica Si 60
¬
with -hexane, concentrated
l
(40±63 m). The column was eluted with a stepwise gradient of methanol in chloroform. Radioactivity
was eluted with 30±50 % (v v) methanol chloroform. This process was repeated. Further puri®cation
was performed by serial radio-hplc as described above, except that a 10 250 mm column packed with
l
LiChrospher 5 m RP18e was used with a solvent ¯ ow 4 ml min. This process was repeated twice. A
solvent gradient was used as described for radio-hplc and a modi®ed version with solvents additionally
}
}
¬
}
containing 5% THF. The yield of M1 was approximately 4 mg.
Isolation of metaboliteM1 from urine
The 0±48-h urine of subject 6 (700 ml) was applied to a column (4 30 cm) packed with Amberlite
¬
XAD-2 resin (Rohm & Haas, Philadelphia, PA, USA). The column was eluted with a stepwise gradient
of methanol in water. Radioactivitywas recovered with 50 % (v v) aq. methanol. The main fraction was
separated in serial radio-hplc runs asdescribed above.Further puri®cation was performed using the same
}
l
system except that a column (4 6 250 mm) packed with Supelcosil RP18 ABZ 5 m (Supelco, Gland,
Switzerland) was used and the solvent gradient was shallower (80 % solvent B reached at 75 min). The
±
¬
l
latter step was repeated for ®nal puri®cation of M1. The yield of M1 was approximately 20 g.
UV spectroscopy
UV spectra of puri®ed M1 were recorded during radio-hplc in the wavelength range 190±380 nm
using a diode array UV detector MD910 (Jasco). The spectra were corrected for background which was
estimated as the mean of spectra recorded before and after the hplc-peak.
Mass Spectrometry
M1 isolated from urine and faeces was analysed with a Finnigan MAT 90 mass spectrometer
(Finnigan MAT, Bremen, Germany) equipped with a Xenon gun for fast atom bombardment (FAB-
MS : 8 keV fast atom beam, matrix thioglycerol, ion source temperature 40 C).
°
S
N
N
[(2 -( -tetrazol-5-yl)biphenyl-4-yl)-methyl]- valine) was performed by HR-FAB-MS (high resolution
Accuratemass determination ofthe reference compound CGP 71580 (( )- -(4-hydroxy-valeryl)-
1H
-
«
fast atom bombardment mass spectrometry) applying the software peakmatch method on the M H ion.
The mass spectrometer was a VG 70-SE (Fisons Instruments, Manchester, UK) instrument.
-
"
H-nmr spectrometry
M1 isolated from faeces was dissolved in H -methanol and subjected to proton magnetic resonance
#
%
spectroscopy ( H-nmr) on a Varian Unity 500 spectrometer (500 MHz ; Varian Instruments, Palo Alto,
"
CA, USA) at 25 C. A COSY spectrum (Correlated spectroscopy : homonuclear shift correlation) was
°
measured on the same instrument.

Disposition of valsartan in man
63
Pharmacokinetic evaluations
The peakconcentrations of valsartan and radioactivity in blood and or plasma ( max ) and the time to
C
) were observed values. Using the pharmacokinetics evaluation software Dekin (Ciba-
}
C
Geigy) the following pharmacokinetic parameters were determined : area under the blood plasma
t
reach
(
max
max
}
), calculated by the trapezoidal rule ; area under the
AUC
blood plasma concentration±time curve up to time
concentration±time curve up to 24 h (
(
h)
#%
t
last time point with measurable concentration; apparent half-life of concentrations in the 8±24-h
AUC
(
(t)), calculated by the trapezoidal rule with
}
t
¯
b
t
#
segment ( -phase) of the concentration±time curves ( b), calculated by logarithmic±linear regression
"
c
analysis; apparent terminal half-life of concentrations in the terminal segment ( -phase) of the
concentration±time curves ( c), calculated by logarithmic±linear regression analysis; area under the
t
"
#
AUC
t
concentration calculated for last time point on the regression
AUC
blood plasma concentration±time curve to in®nity (
¢ ), calculated by extrapolation of
to
}
(
)
(t)
C
line established for the determination of
t
C
in®nity by adding (
.
) ln2, with
"
#
¯
t
}
c
t
t
The blood and plasma kinetics data are presented as means SD for
.
"
#
c
n
quantitatively to collect the excreta, therefore excretion and biotransformation data are reported as
6. Subject 2 failed
³
¯
n
means SD for
5. Dose proportions recovered in the individual urine and faeces samples were
³
¯
calculated and the mass balance was established for each individual for 0±168 h. The dose proportions in
the individual metabolite peaks in urine and faeces samples were determined by liquid scintillation
counting of eluate fractions collected after hplc separation, and the total recovery of each metabolite was
calculated for the time period 0±48 h for urine and typically 12±72 h for faeces.
Synthesis of M1
In analogyto the methodology described previously (Buhlmayer
et al
-[(2 -cyano-biphenyl- 4-yl)-methyl]- valine-benzylester (Chem. Abstr. No. 137864-23-4) was acylated
S
. 1994, European Patent), ( )-
N
with 4-oxo-valerylchloride (Helberger 1936, Kanda
«
et al
. 1992) and diisopropylethylamine in
methylenechloride to the keto-derivative ( )- -(4-oxo-valeryl)- -[(2 -cyano-biphenyl-4-yl)-methyl]-
S
valine-benzylester (yield 63%). Transformation of the cyano group to the tetrazole ring was achieved by
N
N
«
n
o
S
N
N
1H
treatment with tri- -butyl tin azide in re¯ uxing -xylene to give ( )- -(4-oxo-valeryl)- -[(2 -( -
«
tetrazol-5-yl)biphenyl- 4-yl)-methyl]- valine-benzylester (41 %). The ester was cleaved by hydrogenation
S
N
N
-
in the presence of palladium on charcoal (10%) in tetrahydrofuran to provide ( )- -(4-oxo-valeryl)-
[(2 -( -tetrazol-5-yl)biphenyl-4-yl)-methyl]- valine (89 %, HR-FAB-MS : (M H)+ observed :
1H
450 2157; calculated: 450 2142 ; C
«
-
H N O ). Subsequent reduction of the keto group with sodium
±
borohydride in methanol gave ( )- -(4-hydroxy-valeryl)- -[(2 -( -tetrazol-5-yl)biphenyl-4-yl)-
±
#%
S
#) & %
N
N
1H
«
methyl]- valine (CGP 71580 ; HR-FAB-MS : (M H)+ observed : 452 2275 ; calculated: 452 2299 ;
±
±
-
C
H N O ; yield : 94 %).
#% $! & %
Binding test to angiotensin II antagonist receptor
Valsartan and the synthesized standard of its hydroxylated metabolite (CGP 71580, ( )- -(4-
in vitro
S
N
hydroxy-valeryl)- -[(2 -( -tetrazol-5-yl)biphenyl-4-yl)-methyl]- valine) were tested for activity in the
N
angiotensin II receptor binding assay as previously reported (Whitebread
1H
«
et al
Whitebread 1995). Rataortic smooth muscle cells (SMC) were homogenized in sodium bicarbonate and
. 1989, de Gasparo and
g
centrifuged at 60 000 for 30 min. The pellet was resuspended in Tris- HCl buŒer, pH 7 4 containing
NaCl, MgCl and EDTA using a tightly ®tting Te¯ on- pestle homogenizer. A cocktail of various
±
#
l
proteinase inhibitors was added. Membranes (20 g protein as measured according to the method of
Bradford [1976] and bovine serum albumin (BSA) as standard) were incubated with 175 pmol l
}
"
)
iodinated (Sar -Ile ) angiotensin II and increasing concentrations of unlabelled CGP 71580 or valsartan
for 120 min in the assay buŒer without BSA. Non-speci®c binding was determined in the presence of
"
mol l unlabelled angiotensin II. Bound and free radioactive (Sar -Ile ) angiotensin II were separated
)
l
by ®ltration through Whatman GF F ®lters. The radioactivity trapped on the ®lter was measured with
1
}
}
c
a
-counter at 70 % e ciency (Pharmacia-LKB, Uppsala, Sweden). The binding data were analysed
et al K i
was calculated
using the four-parameter logistic dose±response method of De Lean
. (1978).
et al
following the Chang±PrussoŒ equation (Criscione
. 1993).
Results
Blood and plasma concentrations
"%
Following oral administration of [ C]valsartan to volunteers, peak concen-
trations of radioactivity were measured in blood and plasma at 1 h (Table 1, Figure
2). Elimination of radioactivity from plasma was multiexponentia l and an apparent
a-half-life of 1 5±2 h was observed in the period 2±6 h, during which absorption may
±

64
F. Waldmeier et al.
"%
Table 1. Pharmacokinetic parameters for concentrations of C-radioactivity in blood and plasma and
n
of valsartan in plasma (
6).
¯
Radioactivity
Radioactivity
Valsartan
concentration concentration concentration
in blood in plasma in plasma
Parameter
t
(median)
(h)
l
( mol l)
(h)
1
6 0 1 7
6 1 0 8
1
10 6 3 0
1
max
Cmax
10 5 3 0
± ±
5 7 1 0
44 2 12 4
± ±
44 2 12 4
24
±
±
±
±
±
±
³
³
³
}
a
b
t
AUC
"
#
6 1 0 6
±
±
±
±
49 6 13 2
± ±
³
³
³
l
( mol l) .h
}
28 2 6 8
±
±
±
³
³
³
AUC (
l
( mol l) .h
(h)
h)
b
#%
b
33 0 8 7
58 8 16 0
±
±
±
± ±
³
³
³
}
(t)
t
t
(range)
48±168
75 3 24 0d
72±216
80 8 33 0d
c
c
"
#
(h)
l
( mol l) .h
}
na
na
±
±
±
±
±
³
³
³
³
39 0 5 1d
66 9 12 1d
± ±
AUC
±
¢
(
)
a
b
c
d
Time period 8±24 h.
t
, last measurable time point.
Typical time period : 48 or 72 h
t
.
®
n
Mean of
5.
¯
!
LOQ
na, Not applicable, concentrations
.
"%
Figure 2. Mean concentrations of total C-radioactivity ( ) and valsartan ( ) in the plasma of six
D
subjects. Upward bars indicate the SD for C-radioactivity, downward bars the SD for valsartan.
+
"%
have been ongoing. A mean a-half-life could not be determined since the number of
time points in the time period was insu cient. A mean apparent b-half-life (t" ) of
b
approximately 6 h was observed in the period 8±24 h and the mean terminal hal#f-life
(t" ) was approximately 80 h. The terminal half-life showed substantial variability,
c
#
so the last time point with measurable radioactivity in plasma varied from 72 to
216 h after the dose. In the terminal elimination phase, the radioactivity concen-
!
trations were
valsartan.
1% of the Cmax , and below the LOQ of the analytical method for
"%
The concentration±time curves of C-radioactivity in blood and plasma were
parallel. The radioactivity concentrations in blood were about half those in plasma,
with the haematocrit ranging from 0 39 to 0 46. Therefore, the association of the
±
±
radioactivity with red blood cells was minimal and radioactivity was essentially
con®ned to plasma.

Disposition of valsartan in man
65
n
Table 2. Cumulative excretion of radioactivity in urine and faeces (
5).
¯
Excretion of radioactivity
(% of dose)
Faeces
Time
interval
(h)
Urine
12 9 3 6
Total
61 4 16 5
0±48
0±96
0±168
48 6 18 6
± ±
±
±
±
±
³
±
93 4 8 1
³
³
13 1 3 8
80 3 11 4
±
±
±
±
85 7 4 5
± ±
³
³
³
13 2 3 8
98 9 1 0
± ±
±
±
±
³
³
³
Valsartan measured by hplc-¯ uorimetry was detected in plasma for 24 h only.
The elimination was biexponential and was comparable with that of plasma
radioactivity. The mean AUC (
of valsartan accounted for 89% of that of plasma
h)
#%
radioactivity, demonstrating that unchanged valsartan was the predominant radio-
active compound in plasma.
Excretion of radioactivity in urine and faeces
Faecal excretion, accounting for 86% of the dose on average, was clearly
predominant (Table 2). Renal excretion was almost complete after 48 h, whereas
substantial faecal excretion proceeded until day 4. Subsequently, excretion of
marginal dose proportions per day continued for several days and the mass balance
after 7 days was essentially quantitative. The terminal excretion in faeces with daily
recoveries of 2% of the dose paralleled the terminal elimination phase (c-phase)
%
in plasma.
Biotransformation
The radioactivity in selected plasma samples was analysed by radio-hplc. In
plasma collected at 1 h (tmax ), unchanged valsartan was the only detectable
radioactive compound. In plasma samples collected at 8 h, in addition to valsartan,
a metabolite was detected which accounted for approximately 10% of the
radioactivity (®gure 3a). This plasma metabolite had the same hplc retention time as
the metabolite M1 in urine and faeces (see below) and was considered to be identical
with M1. Essentially identical chromatograms were obtained following direct
analysis of plasma samples without preparatory protein precipitation. In a
methanolic faeces extract, unchanged valsartan was the main radioactive compound
accounting for approximately 71% of the dose (table 3, ®gure 3b). Metabolite M1
accounted for approximately 8% of the dose and two additional trace peaks (P1, P2)
were detected. In urine, the predominant radioactive compound was valsartan
accounting for 10% of the dose (table 3). Metabolite M1 accounted for 1% of the
dose. Using a speci®c hplc-UV method, the urinary excretion of valsartan in the
period 0±48 h accounted for 10 1 2 6% of the dose (n 5). The mean total of
±
valsartan in urine and faeces amounted to 81% of dose, and that of metabolite M1
±
³
¯
was 9%.
Identi®cation of the metaboliteM1 in urine and faeces
The metabolite M1 was stable in urine under acidic and alkaline conditions (pH
range 1±14, 25±37 C, 16 h) and under treatment with excess glucuronidase
arylsulphatase (pH 5 1, 37 C, 16 h). M1 was isolated from selected faeces and urine
°
°
±
}

66
F. Waldmeier et al.
(a)
(b)
Figure 3. (a) Representative metabolite pro®le by radio-hplc of plasma collected 8 h after dose
administration. Radioactivitywas determined by liquid scintillation counting following collection
of eluate fractions (0 5 ml). (b) Representative metabolite pro®le of faeces (typical pool period
±
12±72 h) by radio-hplc with on-line radioactivity detection.
n
Table 3. Excretion of metabolites in urine and faeces (
5). Data represent minimal values. The
¯
metabolite balance is incomplete due to minor loss of radioactivity during sample processing and
radio-hplc and since the ®nal excreta portions with low levels of radioactivity were not analysed.
Hplc
retention
time
Excretion (% of dose)
Faeces
Hplc peak compound
}
Urine
(0±48 h)
}
metabolite
(min)
(12±72 h)a
Total
P1, metabolite M1
P2, unidenti®ed
P3, unidenti®ed
P4, valsartan
Total of four compounds
Unidenti®ed trace compounds
Total excretiona
20
21
23 5
±
1 1 0 6
± ±
8 0 3 0
±
9 1 3 4
±
±
±
³
³
³
0 3 0 1
1 9 0 7
±
2 1 0 7
±
±
±
±
±
±
± ±
³
³
³
0 3 0 1
0 9 0 3
±
1 2 0 3
± ±
±
±
³
³
³
24 5
±
9 8 2 5
70 8 6 5
±
80 6 5 0
±
±
± ±
³
³
³
11 5 3 2
81 5 4 0
±
93 0 1 4
± ±
±
±
³
³
³
1 4 0 9
1 5 0 3
±
2 9 1 1
±
±
±
± ±
³
³
³
12 9 3 6
83 1 4 0
±
95 9 1 6
± ±
±
±
±
³
³
³
a
For analysis of faecal radioactivity, the typical time period considered was 12±72 h, within which
97% of the total faecal radioactivity was excreted.

Disposition of valsartan in man
67
"
Table 4. H-nmr- spectroscopy data of metabolite M1, isolated from faeces.
Chemical shift
(ppm)
Number of protons,
multiplicity
Proton position group
}
ortho
-disubstituted phenyl ring
para
-substituted phenyl ring
7 5±7 7
±
7 0±7 3
± ±
4
4
±
4 6±4 9
2, AB system and A system
1d
1
5
3
NCH
±
±
4 55 and 4 18
#
NCH#
±
±
H
3 78 and 3 65
OC CH
±
±
H
$H
H
and C (CH )
# $ #
1 4±2 8
COC
C
±
±
1 20 and 1 10
#
H
CHO
$
C
±
±
1 00, 0 84 and 0 80
H
(C ) CH
$ #
6
±
±
±
Figure 4. Structural formula of the metabolite M1 (CGP 71580).
samples which contained comparatively high radioactivity proportions using solvent
extraction and chromatography procedures. The puri®ed M1 was subjected to
"
FAB-MS, UV spectrophotomet ry and H-nmr spectroscopy. The molecular mass
of M1 measured by FAB-MS was 451 (negative ion FAB: m z 450 : M-H ; positive
}
ion FAB: m z 452 : M H, m z 474 : M Na). This mass was greater than that of
-
valsartan by 16 amu, indicating the presence of an additional oxygen atom. The UV
-
}
}
absorption spectra of M1 (kmax : 205 nm ; shoulders at 225, 248 and 279 nm ; ratio
e
e
: 0 32, ratio e
±
e
: 0 52) were comparable with that of valsartan,
±
##&
}
}
#)!
#&!
#&!
demonstrating that the chromophore of valsartan, the aromatic residues, were
"
retained unchanged in M1. Also, the H-nmr spectrum showed the presence of the
unchanged aromatic part of valsartan. Additionally, signals at 1 1 and 1 2 ppm
±
±
showed the presence of a CH (CH) group, whereas signals of a CH (CH )
$ E $E
E
E
group were absent (table 4). In the COSY nmr spectrum, the signals at 1 1 and
#
±
1 2 ppm showed cross peaks to signals at 3 65 and 3 78 ppm respectively. The results
±
±
indicate the presence of a CH CH O moiety and consequently show that M1 was
±
$E
valeryl-4-hydroxy- valsartan (or valeryl-(x-1)-hydroxy- valsartan) (®gure 4). After
E
structure elucidation of M1, the metabolite was synthetized and used as a reference
"
compound (CGP 71580). On hplc, CGP 71580 coeluted with M1 and the H-nmr
spectrum of CGP 71580 was identical with that of metabolite M1, except for the
mixture ratio of the two diastereoisomer s created by the new chiral centre at the
carbon atom of the secondary alcohol group (signals at 1 1 and 1 2 ppm). In M1, the
±
±
ratio was approximately 3:2, while in CGP 71580 it was about 1:1. The absolute
stereochemistry of the diastereoisomer s was not elucidated.

68
F. Waldmeier et al.
Pharmacological activity of M1
The reference compound CGP 71580, which is identical with M1 except for the
slightly diŒerent diastereoisomer mixture ratio, was tested in vitro for a nity to the
angiotensin II AT receptor in the smooth muscle cell binding assay (Whitebread et
"
al. 1989). The IC of CGP 71580 was 0 47 0 03 lmol l (K 0 37 lmol l). The
±
±
±
a nity of CGP 71580 to the AT receptor was 200-fold lower than that of valsartan.
³
}
}
i
&!
"
Discussion
"%
The median tmax of 1 h indicated rapid onset of absorption of C-valsartan
following administration as aneutral-buŒered solution. No conclusion can be drawn
from the presented data on the duration of the absorption process.
The similar plasma concentration±time curves of valsartan and radioactivity
demonstrated that the predominant radioactive compound in the system was the
unchanged valsartan. The metabolite M1 was a minor component in plasma and
became detectable only after 2±3 h. The assay for unchanged valsartan was able to
detect the compound for 24 h after the dose.
Following intravenous administration of 20 mg valsartan to healthy male
volunteers, the mean plasma AUC (
1996). The mean plasma AUC (
administered as solid substance in a capsule was 19 6 6 0 (lmol l). h. With
was 21 6 3 9 (lmol l).h (Flesch et al.
± ±
following an 80 mg single oral dose of valsartan
³
}
h)
#%
h)
#%
±
±
³
}
correction for dose, this plasma AUC ( corresponded to an absolute oral
bioavailability of 23 7% (Flesch et al. 1996). With a non-radioactive solution
h)
#%
³
formulation, the AUC ( was 32 9 8 1 (lmol l). h and the oral bioavailability was
± ±
39 7% (Flesch et al. 1996). In the present study with a solution formulation of the
³
}
h)
#%
³
radiolabelled compound, the mean plasma AUC (
for unchanged valsartan was
44 2 12 4 (lmol l) .h and hence the mean oral bioavailability was 51%, based on
h)
#%
±
±
³
}
i.v. data from the previous study assuming route independent linear kinetics. The
interindividual variability of AUC and bioavailability was substantial. With an
estimated absolute oral bioavailability of about 51%, the total gastrointestinal
absorption amounted to at least 51% of the dose.
Valsartan is a diacid with a carboxylic acid group (pK 3 9) and a tetrazole group
±
(pK 4 7). In water and aqueous acid solid valsartan is poorly soluble whereas it is
a
±
a
well soluble in the normal pH range of the intestine, i.e. 5±7. Absorption of valsartan
administered as a solid appears to be limited by the dissolution rate. When it is
administered into the fasted stomach in a neutral, buŒered solution it can reach the
intestine largely in dissolved, anionic form and then appears to be well absorbed.
This has been substantiated by detailed investigations (Flesch et al., in preparation).
The extent, rate and duration of the faecal excretion indicate elimination with
bile. With the presently available data, a potential ®rst- pass biotransformation
and or ®rst- pass biliary elimination cannot be assessed and hence the extent of
}
absorption cannot be estimated with any more accuracy. With 13% of the dose
recovered in urine, the major part of the absorbed dose, at least about 40% of the
dose, must have been eliminated via bile.
Valsartan was excreted largely as unchanged compound and was only minimally
metabolized. This result may be explained by the fact that valsartan, at physiological
pH, is a hydrophilic di-anion. Therefore it may be a poor substrate for metabolizing
enzymes. The pronounced biliary elimination suggests active involvement of an
anion transporting system in the liver, presumably of the canalicular multispeci®c

Disposition of valsartan in man
69
organ anion transporter (cMOAT), which is responsible for biliary elimination of
many anionic endo- and xenobiotics, particularly of di-anions (Meier 1993, Oude
Elferink and Jansen 1994).
Valsartan was metabolized to a small extent only. The only notable metabolite in
plasma, urine and faeces was the metabolite M1. M1 was neither an N- nor an O-
glucuronic acid conjugate nor a sulphate conjugate. In particular, since alkaline
treatment had no eŒect on M1, an acyl glucuronide could be ruled out. Various
spectroscopic data provided evidence that M1 was a hydroxylated derivative of
valsartan formed by oxidation.
The biotransformation of valsartan diŒers substantially from that of the
structurally and pharmacologically similar compound losartan. The latter has an
alcohol group instead of a carboxylic acid group and undergoes oxidative
biotransformation to the acid metabolite denoted as EXP3174, which shows
markedly higher pharmacological activity than losartan itself (Munafo et al. 1992,
Stearns et al. 1992, 1995, Lo et al. 1996). Therefore EXP3174 is the principal
pharmacologically active agent after administration of losartan. A further metabolic
process observed with losartan in vitro was formation of an N-2-glucuronic acid
conjugate (at N-2 of the tetrazole moiety) when incubated with human liver
microsomes (Huskey et al. 1993).
Valsartan is pharmacologically active in the unchanged form. In contrast with
losartan, the biotransformation of valsartan to M1 does not represent an activation
process, since CGP 71580 (essentially equal to M1) showed only marginal a nity in
the angiotensin II receptor binding assay. Therefore, the biotransformation of
valsartan to M1 represents a minor, additional elimination process. In a study with
losartan in human liver slices and microsomes in vitro, an (x-1)-hydroxylated
metabolite was detected, which is structurally analogous to M1 and which showed
only marginal pharmacological activity (Stearns et al. 1992).
The plasma concentrations of M1 were not determined but can be estimated
from the diŒerence between radioactivity and valsartan concentrations in plasma.
The diŒerence in plasma AUC (
amounted to about 11% of the AUC (
radioactivity. The diŒerence between radioactivity and valsartan plasma levels was
of
h)
h)
#%
#%
highest at 3 h after dosing (tmax , median) amounting to approximately 0 9 lmol l
±
}
(Cmax , mean). The Cmax of M1 represented about 23% (mean) of the valsartan
concentration at the tmax of M1, and 9% of the Cmax of valsartan. By 24 h, the M1
concentration dropped to 8 % of its Cmax . An elimination half-life for M1 in plasma
cannot be determined accurately, but is estimated to be approximately 6 h. Further,
the total M1 excreted in urine and faeces amounted to 9% of the dose, which
represented less than one-®fth of the absorbed dose. Accordingly, the formation and
excretion of M1 (metabolic clearance of valsartan), represents a minor contribution
to the total systemic clearance of valsartan.
If oxygenating cytochrome P450 enzymes are involved in the elimination of a
drug, drug±drug interactions may occur via inhibition or induction of the involved
enzymes. Such interactions are clinically relevant only if the aŒected metabolic
process is a main elimination process for a drug. According to a recent review
(Meyer et al. 1996), a signi®cant interaction potential exists only if one enzyme
contributes at least 40±50 % of the total clearance. In the case of valsartan, the
metabolic clearance is low and therefore the total systemic clearance of valsartanwill
not be aŒected by enzyme inhibition to any signi®cant extent.
The interaction potential in the case of enzyme induction by a co-medication is

70
F. Waldmeier et al.
more di cult to predict. A signi®cant interaction may be expected only if formation
of a major metabolite is exclusively catalysed by one enzyme and if this enzyme is
induced to a marked extent. The structurally similar losartan was found to be
metabolized by both cytochrome P450 enzymes (CYP) 2C9 and 3A4 (Stearns et al.
1995). The involvement of 2C9 can be explained by the established preference of
this enzyme for anionic substrates. At present, it is unknown, which enzyme(s)
metabolize the dianion valsartan, but, in analogy to losartan, 2C9 can be assumed to
be involved. There is little evidence in the literature of induction of the human
CYP2C9 enzyme and of signi®cant drug±drug interactions via induction of CYP2C9
(Parkinson 1996). Therefore, clinically signi®cant interactions of valsartan with
enzyme inducing co-medications are not anticipated, although they cannot be
excluded.
Five of the six subjects were phenotyped as extensive debrisoquin metabolizers
!
showed a metabolic ratio of 5 4, but excreted 9 1% of the valsartan dose in the form
with a metabolic ratio
1. While still being an extensive metabolizer, subject 2
±
±
of M1, which represents the exact average excretion of M1. Therefore it may be
concluded that the debrisoquin- metabolizing enzyme CYP2D6 is not involved in
the metabolism of valsartan, and that interactions of valsartan with co-medications
via 2D6 are unlikely.
The main elimination process for valsartan and M1 is biliary elimination.
Therefore, impairment of hepato-biliary transport functions can have a marked
impact on the clearance of valsartan, as shown in a study in liver patients (Brookman
et al., submitted). On the other hand, since only small amounts of valsartan and M1
are excreted with urine, impaired renal function will not signi®cantly change the
clearance of valsartan.
Acknowledgement s
We wish to thank D. Bagatto, M. Develey and D. Graf for excellent technical
assistance in metabolite analysis and radiometry, H. Mory, P. Leupp and Dr
P. Ackermann for preparation of radiolabelled valsartan, F. Hell for determination
of valsartan concentrations, F. Luthi for clinical assistance, P. Birrer and C. Ritter
for spectroscopic measurements, and St. Whitebread and D. Monna for the binding
assays.
References
radford
B
, M. M., 1976, A rapid and sensitive method for the quanti®cation of microgram quantities of
Analytical Biochemistry
protein utilizing the principle of protein±dye binding.
, 72, 248±254.
loyd lesch
, G.,
rookman
olan
enjamin
, P. E., B , I. S., P
ioufi ullins
, A. and M
almer
yld
, P. J., L
B
, L. J., R
aldmeier
, F., S
, K. R., W
, P., F
W
, F., Pharmacokinetics of valsartan in patients with liver
Clinical Pharmacology and Therapeutics
disease.
runner
, L. A., P
(submitted).
lesch
, G., 1994, A semiautomated analyticalmethod
owell
egen
, M. L., D
B
B
B
, P. and F
for the determination of potential antihypertensive agents (CGP 48933 and or CGP 48369) in
}
Laboratory Robotics and Auto
human plasma using high-performance liquid chromatography.
mation
-
, 6, 171±179.
uret
uhlmayer
riscione
ood
de
asparo
hitebread
chmidlin
, T.,
, P., F
, P., C
, L.,
G
, M., W
, S.,
S
attmann
L
, R. and W
developed from the structurally new amino acid series.
Letters
, J., 1994, Valsartan, a potent orally active angiotensin II antagonist
Bioorganic and Medicinal Chemistry
, 4, 29±34.
stermayer
uhlmayer
chmidlin
, T., Acyl Compounds. European Patent 443983 (Ciba-
, P., O
, F. and S
Geigy).

Disposition of valsartan in man
71
riscione
radley
uhlmayer
hitebread
lazer
, R., L
loyd
uller
, P., M , P.
C
C
, L., B
, W. A., B
, P., W
, S., G
de asparo
and
angiotensin II antagonist.
riscione de asparo
G
, M., 1995, Valsartan: preclinical and clinical pro®le of an antihypertensive
Cardiovascular Drug Reviews
, 13, 230±250.
hitebread amjoue
uhlmayer
ood
, J.,
, L.,
G
, M., B
, P., W
, S., R
1993, Pharmacological pro®le of valsartan: a potent, orally active, nonpeptide antagonist of the
British Journal of Pharmacology
, H. P. R. and W
angiotensin II AT -receptor subtype.
, 110, 761±771.
hitebread
, S., 1995, Binding of valsartan to mammalian angiotensin AT
"
de asparo
G
, M. and W
"
Regulatory Peptides
unson
receptors.
de ean
, 59, 303±311.
odbard
, P. J. and R , D., 1978, Simultaneous analysis of families of sigmoidal
L
, A., M
curves : application to bioassay, radioligand assay, and physiological dose-response curve.
American Journal of Physiology
, 40, 360±367.
egen owald
lesch
uller
ell
ieterle
, W., 1996, Absolute oral
F
, G., M
bioavailability of valsartan in healthy volunteers. Abstract of a poster presented at the 6th
European
, P., H
, F., D
, P., H
, H. and D
European Congress on Biopharmaceutics and Pharmacokinetics, Athens, April 1996.
Journal of Drug Metabolism and Pharmacokinetics
(special issue), 34.
, J. H., 1936, Uber das Verhalten des Lavulinsaurechlorids und der Acetyllavulinsaure bei
Annalen
È
elberger
H
H
der Friedel- Craftsschen Reaktion.
uskey iller hiu
glucuronidation of selected angiotensin II receptor antagonists in hepatic microsomes from rats,
, 522, 269±277.
N
N
-
, S.- E., M
, R. R. and C
, S. H. L., 1993, -glucuronidation reactions. I. Tetrazole
Drug Metabolism and Disposition
dogs, monkeys and humans.
anda rai shizawa
, 21, 792±799.
asai
, M., 1992, New potent mitomycin
orimoto
K
, Y., A , H., A
, T., M
, M. and K
Journal of
derivatives : synthesis and antitumor activity of 7,7-(ethylenedioxy)mitomycins.
Medicinal Chemistry
, 35, 2781±2786.
rea
Pharmacokinetics of losartan, an angiotensin II receptor antagonist, and its active metabolite
o
oldberg
c
u
urtek
jornsson
, C. I. and B , T. D., 1995,
L , M.-W., G
, M. R., M C
, J., L , H., F
Clinical Pharmacology and Therapeutics
Hepatic Transport and Bile Secretion
, P. J., 1993, Canalicular membrane transport processes. In
EXP3174 in humans.
, 58, 641±649.
eier
M
M
,
edited by N. Tavoloni and P. D. Berk (New York: Raven), pp. 587±596.
ichelbaum
eyer
mrein
, U. A., A
alant
aux
, P., L
ertilsson
ikkelsen
uentert
, T. W.,
riest
, B. G., P , R.,
, R., B
ackson
, L. P., B
, G., M
, L., E
eck
, M., G
ollock
enauer
joqvist
H
, S., J
, F. and D
, H., P
, C., P
elini tula
S -S
Acta Psychiatrica Scandinavica
expert group report.
, A., 1996, Antidepressants and drug-metabolizing enzymesÐ
, 93, 71±79.
orland
unafo
B
hristen
, A., C
iollaz
, J. and B
ussberger
hum
ee aeber
, R. M., L , R. J., W , B.,
M
, Y., N
runner
, J., S
, L. Y., B
, H. R., 1992, Drug concentration response relationships in normal
Clinical
volunteers after oral administration of losartan, an angiotensin II receptor antagonist.
Pharmacology and Therapeutics
, 51, 513±521.
ansen
, P. L. M., 1994, The role of the canalicular multispeci®c organic
ude lferink
O
E
, R. P. J. and J
Pharmacology and Therapeutics
anion transporter in the disposal of endo- and xenobiotics.
77±97.
, 64,
arkinson
P
S
, A., 1996, An overview of current cytochrome P450 technology for assessing the safety and
ToxicologicPathology
e cacy of new materials.
tearns iller
S. H. L., 1992, The metabolism of DUP- 753, a nonpeptide angiotensin II receptor antagonist, by
, 24, 45±57.
hakravarty
oss
, R. R., D , G. A., C
osegay
, P. K., R
atto
hiu
,
, R. A., M
, A., G
, G. J. and C
Drug Metabolism and Disposition
hiu
rat, monkey, and human liver slices.
tearns hakravarty hen
, 20, 281±287.
, S. H. L., 1995, Biotransformation of losartan
S
, R. A., C
, P. K., C
, R. and C
Drug Metabolism and
to its active carboxylic acid metabolite in human liver microsomes.
Disposition
, 23, 207±215.
aiser
K. C., 1991, The disposition of [ C]- labelled benazepril HCl in normal adult volunteers after
aldmeier
ckermann
, G., A
aigle
agner
arner
asseter
, A. and L ,
W
W
, F., K
, R., F
, J. W., W
, J., B
"%
Xenobiotica
single and repeated oral dose.
, 21, 251±261.
de asparo
hitebread ele amber
, S.,
characterization of two angiotensin II receptor subtypes.
Communications
M
, M.,
K
, B. and
G
, M., 1989, Preliminary biochemical
Biochemical and Biophysical Research
, 163, 284±291.