Comparison of celecoxib metabolism and excretion in mouse, rabbit, dog, cynomolgus monkey and rhesus monkey

Article abstract of DOI:10.1080/00498250050078039

1. The metabolism and excretion of celecoxib, a specific cyclooxygenase 2 (COX-2) inhibitor, was investigated in mouse, rabbit, the EM (extensive) and PM (poor metabolizer) dog, and rhesus and cynomolgus monkey. 2. Some sex and species differences were ev

Full text of DOI:10.1080/00498250050078039

xenobiotica, 2000, vol. 30, no. 7, 731±744
Comparison of celecoxib metabolism and excretion
in mouse, rabbit, dog, cynomolgus monkey and
rhesus monkey
S. K. PAULSON‹*, J. Y. ZHANG‹, S. M. JESSENŒ, Y. LAWALŒ,
N. W. K. LIU, C. M. DUDKOWSKI‹, Y.-F. WANG‹,
M. CHANG‹, D. YANG‹, J. W. A. FINDLAY‹, M. A. BERGEŽ,
C. S. MARKOS‹, A. P. BREAU‹, J. D. HRIBAR and J. YUAN‹
‹
Pharmacokinetics, Bioanalytical & Radiochemistry, ŒRegulatory Aåairs and Physical
Methodology, G. D. Searle & Co., 4901 Searle Parkway, Skokie, IL, USA
Metabolic Chemistry, Covance Laboratories, Inc., Madison, WI, USA
Ž
Received 2 February 2000
. The metabolism and excretion of celecoxib, a speci®c cyclooxygenase 2 (COX-2)
1
inhibitor, was investigated in mouse, rabbit, the EM (extensive) and PM (poor metabolizer)
dog, and rhesus and cynomolgus monkey.
2
. Some sex and species diåerences were evident in the disposition of celecoxib. After
intravenous (i.v.) administration of ["%C]celecoxib, the major route of excretion of
radioactivity in all species studied was via the faeces: EM dog (80.0%), PM dog (83.4%),
cynomolgus monkey (63.5%), rhesus monkey (83.1%). After oral administration, faeces
were the primary route of excretion in rabbit (72.2%) and the male mouse (71.1%), with
the remainder of the dose excreted in the urine. After oral administration of ["%C]celecoxib
to the female mouse, radioactivity was eliminated equally in urine (45.7%) and faeces
46.7%).
(
3
. Biotransformation of celecoxib occurs primarily by oxidation of the aromatic methyl
group to form a hydroxymethyl metabolite, which is further oxidized to the carboxylic acid
analogue.
4. An additional phase I metabolite (phenyl ring hydroxylation) and a glucuronide
conjugate of the carboxylic acid metabolite was produced by rabbit.
5
. The major excretion product in urine and faeces of mouse, rabbit, dog and monkey
was the carboxylic acid metabolite of celecoxib.
Introduction
Celecoxib, 4-[5-(4-methylphenyl)-3-(tri¯uoromethyl)-1H-pyrazol-1-yl] ben-
zenesulphonamide, is a speci®c inhibitor of the inducible form of cyclooxygenase
(
COX-2) and does not inhibit COX-1 at plasma concentrations that are therapeutic
in arthritis (Penning et al. 1997, Isakson et al. 1998, Simon et al. 1998). Constitutive
COX-1 is widely expressed in most all tissues and appears responsible for most of
the physiological prostaglandin production associated with gastric lining cyto-
protection (O’Neill and Ford-Hutchinson 1993, Seibert et al. 1994). COX-2
expression is induced in in¯ammation (Croåord et al. 1994, Seibert et al. 1994). It
has been hypothesized that COX-2 inhibitors will have the bene®ts of classical non-
steroidal anti-in¯ammatory drugs (NSAID) without the risk of gastrointestinal
toxicity (Isakson et al. 1998). Celecoxib was recently approved in the USA for the
treatment of the signs of symptoms of rheumatoid arthritis and osteoarthritis.
*
Author for correspondence: e-mail: Susan.K.Paulson!monsanto.com
Xenobiotica ISSN 0049-8254 print}ISSN 1366-5928 online ’ 2000 Taylor & Francis Ltd
http:}}www.tandf.co.uk}journals

7
32
S. K. Paulson et al.
Figure 1. Proposed metabolic pathway for celecoxib.
The chemical structure of celecoxib is given in ®gure 1. Celecoxib is extensively
metabolized in humans via a single metabolic pathway that involves hydroxylation
of the aromatic methyl group of celecoxib followed by further oxidation of the
hydroxymethyl metabolite to the carboxylic acid (Paulson et al. 2000a). The major
excretion product of celecoxib in humans is the carboxylic acid metabolite.

Disposition of ["%C]celecoxib in several species
733
Glucuronide conjugation of the carboxylic acid metabolite is a minor pathway of
elimination of the carboxylic acid metabolite in humans.
In dog, celecoxib is metabolized in vitro to the hydroxymethyl metabolite with
subsequent oxidation to the carboxylic acid analogue. There is a polymorphism in
the canine metabolism of celecoxib associated with the formation of the hydroxy-
methyl metabolite (Paulson et al. 1999). At least two distinct populations of dog that
have been described designated PM for `poor metabolizers’ and EM for `extensive
metabolizers’ of celecoxib. An equal distribution of the two phenotypes, EM and
PM, exists within a population of the beagle dog. The polymorphism appears due to
a diåerence in the rate of metabolism of celecoxib by liver cytochrome P450. The
speci®c cytochrome(s) responsible for the polymorphism is not understood.
The present study investigated the metabolism and excretion of celecoxib in
several animal species: mouse, rabbit, dog, cynomolgus monkey and rhesus monkey.
Routes of excretion were determined; metabolites were isolated from urine and
faeces and their structures elucidated.
Materials and methods
Chemicals
Celecoxib and radiolabelled celecoxib, 4-[5-(4-methylphenyl)-3-(tr i¯ uoromethyl)-1H-pyrazol-1-yl-
-"%C], were synthesized at Searle (Skokie, IL, USA). The chemical purity of celecoxib was " 99%;
5
chemical purity of the celecoxib metabolites was " 95%. Radiochemical purity of the radiolabelled
celecoxib was " 96% and the speci®c activity of the radiolabelled celecoxib was 56 mCi}mmol. The
hydroxymethyl and carboxylic acid metabolites of celecoxib as well as the 1-O-glucuronide conjugate of
the carboxylic acid metabolite of celecoxib were synthesized at Searle and MS and NMR (®gure 1)
con®rmed the structures. The hydroxymethyl metabolite was synthesized from celecoxib by photo-
catalyzed bromination (NBS) of the methyl group followed by hydrolysis of the bromomethyl products.
Sodium borohydride reduction of the hydrolysis mixture converted the mixture of hydroxymethyl and
aldehyde products to the desired compound. Oxidation of the hydroxymethyl metabolite with Jones’s
chromic acid in acetone aåorded the carboxylic acid metabolite. The carboxylic acid metabolite was
condensed with the benzyl ester of TBDMS-protected glucal epoxide followed by protecting group
removal to aåord the glucuronide. All other chemicals and reagents were analytical grade and
commercially available.
Animal studies
Mouse. CD-1 mice (n¯ 33}sex) with body weights 20±40 g (Charles River Breeding Laboratory,
Portage, MI, USA) were administered ["%C]celecoxib in a solution of polyethylene glycol 400:water
(
2:1 w}v) by oral gavage at 35 mg}kg (175 lCi}kg). Mice were fasted overnight before dosing and for at
least 4 h afterwards. They were housed in individual metabolism cages designed for the separation and
collection of urine and faeces. Blood samples (C 1 ml) were collected from the auricular artery or
marginal ear vein (n¯ 30}sex) into chilled, heparinized vacutainers at 0.5, 2, 4, 8, 24, 48, 72, 96, 120 or
1
68 h after dosing. Three blood samples were collected for each time point and only one sample was
collected from each mouse. All urine and faeces were collected from mice (n¯ 3}sex) at ®18 to 0 h
predose and at 24-h intervals for 7 consecutive days after dose administration. Urine was collected into
containers surrounded by dry ice. Faeces were collected at ambient temperature into Stomacher bags and
frozen immediately on dry ice at the end of each collection interval. Urine and faeces were stored at about
®
20 °C until analyses for radioactivity and metabolic pro®le.
Rabbit. Three female New Zealand White (Hra:SPF) rabbits with body weights of 2.8±3.2 kg
Hazelton Research Products, Denver, PA, USA) were administered ["%C]celecoxib in a solution of
(
polyethylene glycol 400:water (2:1 w}v) by oral gavage at 10 mg}kg (60 lCi}kg). Rabbits were fasted
overnight before dosing and for at least 4 h afterwards. They were housed in individual metabolism cages
designed for the separation and collection of urine and faeces. Blood samples (C 5 ml) were collected
from the auricular artery or marginal ear into chilled, heparinized vacutainers at 0.5, 2, 8 and 24 h after
dosing. All urine and faeces were collected at ®18 to 0 h predose and at 24-h intervals for 7 consecutive
days after dose administration. Urine was collected into containers surrounded by dry ice. Faeces were
collected into at ambient temperature into Stomacher bags and frozen immediately on dry ice at the end

7
34
S. K. Paulson et al.
of each collection interval. Urine and faeces were stored at about ®20 °C until analyses for radioactivity
and metabolic pro®le.
Monkey. One female cynomolgus and one female rhesus monkey were administered i.v. ["%C]celecoxib
in a solution of polyethylene glycol 400:water (2:1 w}v) at 1 mg}kg (20 lCi}kg) via the jugular vein.
Monkeys were fasted overnight before dosing and for at least 4 h afterward. They were housed in
individual metabolism cages designed for the separation and collection of urine and faeces. Blood samples
(
C 5 ml) were collected from the femoral vein into chilled, heparinized vacutainers at 5, 15, 30 min, 2, 3,
4
and 6 h for the cynomolgus monkey, and at 3 and 4 h after dosing for the rhesus monkey. All urine and
faeces were collected from the monkeys at ®18 to 0 h predose and at 24-h intervals for 7 consecutive days
after dose administration. Urine was collected into containers surrounded by dry ice. Faeces were
collected at ambient temperature and frozen immediately on dry ice at the end of each collection interval.
Urine and faeces were stored frozen at about ®20 °C until analyses for radioactivity and metabolic
pro®le.
Dog. Six pure-bred beagle dogs (n¯ 2 male, 4 female) (Hazelton Research Product, Inc., Cumberland,
VA, USA) weighing 10±14 kg were characterized as PM or EM as described (Paulson et al. 1999). Brie¯y,
dogs were administered a single 5 mg}kg i.v. bolus injection of celecoxib in a solution of polyethylene
glycol 400:water (2:1 w}v). Venous blood samples were collected into chill tubes containing sodium
heparin at C 6, 12 and 24 h post-dose. Plasma was analyzed for celecoxib concentrations. Dogs with a 24-
h plasma celecoxib concentration below the limit of detection of the assay (0.010 lg celecoxib}ml plasma)
were EM of celecoxib; dogs with a 24-h plasma celecoxib concentration above the limit of detection of
the assay were PM of celecoxib. Dogs were allowed at least 1 week to recover from the pre-screen
procedure. After recovery, dogs were administered i.v. ["%C]celecoxib in a solution of polyethylene glycol
4
00:water (2:1 w}v) at 5 mg}kg (20 lCi}kg) via the cephalic vein. The dogs (n¯ 3 EM, 3 PM) were
fasted overnight before dosing and for at least 4 h afterward. They were housed in individual metabolism
cages designed for the separation and collection of urine and faeces. Blood samples (C 5 ml) were
collected from the jugular vein into chilled, heparinized vacutainers predose and at 5, 15, 30 min, 1, 2, 4,
8
, 24, 48, 72 and 96 h after dosing. All urine and faeces were collected from the dogs at ®18 to 0 h predose
and at 24-h intervals for the 7 consecutive days after dose administration. Urine was collected into
containers surrounded by dry ice. Faeces were collected at ambient temperature and frozen immediately
on dry ice at the end of each collection interval. Urine and faeces were stored at about ®20 °C until
analyses for radioactivity and metabolic pro®le.
Radioactivity analyses
Plasma. Blood samples were centrifuged at C 1000 g at 4 °C for 10 min. Aliquots of plasma (0.1±0.5 ml)
were placed in 20-ml glass scintillation vials along with liquid scintillation cocktail (Ultima Gold XR,
Radiomatic, Meriden, CT; or Aquassure4, Packard Instrument Co., Downers Grove, IL, USA) and
radioactivity was determined in a Model 1500 or 1900 TF liquid scintillation counter (LSC) (Packard).
Urine. Total volumes of the urine samples were recorded. Aliquots of urine were placed in liquid
scintillation vials along with Aquassure4 or Ultima Gold XR liquid scintillation cocktail and radioactivity
was determined by LSC.
Faeces. Faecal samples were thawed at ambient temperature and weighed. About two-to-three times the
sample weight of methanol:water (50:50 or 80:20 v}v) was added to the faecal samples and the samples
were homogenized either using a probe-type homogenizers or in a Stomacher4 bag using a Stomacher4
Laboratory Blender 400 (A. J. Seward, London, UK). Duplicate or triplicate aliquots (0.3±0.6 g) of each
faecal homogenate were placed in a Combustocone4 with pads (Packard) and allowed to dry overnight.
Samples were combusted using a Packard Sample Oxidizer. The combustion products were trapped in
7
ml Carbosorb4 and mixed with 9 ml Perma¯uor4 (Packard) for determination of radioactivity by LSC.
The percentage of dose excreted in each urine and faecal sample was calculated as: (total radioactivity
recovered in each urine and faecal sample divided by total radioactivity dosed)¬100.
Quantitative metabolite pro®ling
Plasma. The mouse, rabbit and monkey plasma samples (1 ml) were acidi®ed with C 0.2 vol. 1 N
phosphoric acid and extracted by a 1-ml 100-mg C18 Bond Elut4 solid-phase extraction (SPE) column
(
Varian, Harbor City, CA, USA) preconditioned with 2-ml aliquots of acetonitrile, methanol and water.
After loading the plasma sample, the column was washed with 2 ml water, then eluted with 2 ml
acetonitrile. The eluate was dried under a stream of nitrogen and reconstituted into 0.2 ml solvent A of
the mobile phase used in the HPLC analysis. The recoveries of radioactivity following extraction were
9
1.4, 87.4, 98.1 and 90.0 % for mouse, rabbit, cynomolgus monkey and rhesus monkey respectively.

Disposition of ["%C]celecoxib in several species
735
Dog plasma samples (C 1 ml) were extracted with an equal volume of acetonitrile, mixed by vortexing
and then centrifuged at 1000 g for 10 min. Supernatants were removed to a separate tube and 1 ml
acetonitrile:water (50:50 v}v) was added to the pellet. The pellet and acetonitrile:water were vortexed
and centrifuged as indicated above. Supernatants from the two extractions were combined. The
recoveries of radioactivity following extractions were 108% for EM and 105% for PM dog.
Urine. The metabolic pro®le of ["%C]celecoxib in urine was determined by direct injection of urine
samples (100 ll) onto the HPLC system. The percentage of the dose excreted as each metabolite in urine
was calculated as: % dose recovered in the urine sample¬the fraction of total HPLC radioactivity of each
metabolite peak.
The urine samples from each dog (5% of the 0±72-h samples) were pooled. An aliquot (10 ml) of each
pooled urine sample was concentrated under nitrogen to 1.5 ml. Acetonitrile (200 ll) was added to the
concentrated urine. Samples were vortexed, sonicated for 10 min and then centrifuged at 1000 g for
1
5 min. The supernatant was removed and water:acetonitrile (80 :20 v}v, 300 ll) was added to the pellet.
The mixture was sonicated for 10 min and then centrifuged as above. The supernatants were removed
and combined with the previous supernatant. The recoveries of radioactivity following extractions were
9
7.1% for EM and 95.3% for PM dog.
Faeces. To determine the distribution of radioactivity in faeces from mouse, rabbit and monkey,
aliquots (C 0.2 g) of faecal homogenates were pooled on a percentage weight basis and extracted with
1
5 ml methanol by end-over-end rotation for 1.5 h at room temperature. The faecal extracts were
centrifuged at C 2000 g for 10 min at 5 °C using a Sorvall RT6000D centrifuge (Dupont Co.,
Wilmington, DE, USA). The supernatant volume was measured, aliquots taken and radioactivity
determined by LSC. The pellet was resuspended in 15 ml methanol, vortexed brie¯y and extracted as
described above. The extracts were combined and evaporated in a heated water bath under a stream of
nitrogen. The residues were reconstituted in 3.0 ml methanol and centrifuged atC 2000 g for 10 min at
4
°C. Aliquots of the reconstituted residues were evaporated in a heated water bath, under a stream of
nitrogen, and reconstituted in 15% acetonitrile containing 0.025 m sodium acetate, pH 4.5, and directly
injected onto the HPLC system. The recovery of radioactivity from the extraction procedures were" 95,
8
0, C 100 and C 100% for mouse, rabbit, cynomolgus monkey and rhesus monkey respectively.
Dog faecal homogenate samples (C 5 g) were acidi®ed with 0.1 N HCl and extracted several times
with acetonitrile or methanol at a ratio of 1:1. Samples were vortexed to mix, sonicated for 10 min and
then centrifuged at 1000 g for 15 min at 5 °C. Supernatants from each extraction were combined. The
recovery of radioactivity from the extraction procedure was 99.2% for EM and 99.3% for PM dog.
The percentage of the dose excreted as each metabolite in faeces was calculated as: % of dose
recovered in the faecal sample¬ the fraction of total HPLC radioactivity of each metabolite peak.
HPLC
HPLC analyses of plasma, urine and faecal samples was performed using an HP1050 autosampler and
HP1050 pump (Hewlett Packard, Wilmington, DE, USA) equipped with a Novapak C18 column
(
3.9¬150 mm, 4 lm, Waters, Marlborough, MA, USA) and a Waters Novapak C18 guard column. A
linear gradient was employed from 20% acetonitrile in 0.025 m sodium acetate buåer, pH 4.5 (mobile
phase A) to 60% acetonitrile in 0.025 m sodium acetate buåer, pH 4.5 (mobile phase B) over 20 min and
held at mobile phase B for 5 min before being re-equilibrated at mobile phase A for 20 min. The ¯ow rate
of mobile phase was 1 ml}min. Eluates from the HPLC column following injection of urine samples and
faecal extracts were mixed with Ultima Flo4 (Packard) at 1:3 (v}v) and analysed for radioactivity using
a Flo-One± Beta A-500 radioactivity detector (Packard). Plasma extracts were injected on the HPLC and
0.5-min fractions were collected using a Foxy 200 (ISCO, Inc., Lincoln, NE, USA) fraction collector.
Each fraction was mixed with 10 ml Ultima Gold scintillation cocktail (Radiomatic, Meriden, CT, USA)
and "%C content was determined by LSC.
LC-MS}MS
Liquid chromatography-mass spectrometry}mass spectrometry (LC-MS}MS) analyses were per-
formed using a series 200 LC autosampler (Perkin-Elmer, Norwalk, CT, USA), and two LC-10A HPLC
pumps (Shimadzu, Tokyo, Japan) coupled to a API-III Plus triple quadrupole mass spectrometer (PE
Sciex, Concord, Ont., Canada). Separation was performed on a Waters NovaPak C18 HPLC column
(
3.9¬150 mm, 4 lm, Marlborough, MA, USA) that was eluted isocratically at 1.0 ml}min, with mobile
phase C (acetonitrile: 0.025 m ammonium acetate, 20:80, pH 4.5) for 5 min, then programmed with a
linear gradient system to mobile phase D (acetonitrile:0.025 m ammonium acetate, 60:40, pH 4.5) over
2
0 min and held at mobile phase D for another 5 min. About 20% of the ¯ow was diverted to the mass
spectrometer using a 1:4 T-splitter. The mass spectral analyses were performed in a negative-ion turbo-
ionspray ionization mode set at 400 °C with the turbo-ionspray interface and ori®ce voltages maintained
at ®3700 and ®65 V respectively. The nitrogen gas for nebulization was set at 50 psi, whereas the

7
36
S. K. Paulson et al.
nitrogen auxiliary and curtain gases were ¯ow controlled at 2 and 1.8 l}min respectively. Collision-
induced dissociation (CID) studies were performed using a collision energy of ®30 eV and at a collision
gas thickness of 250¬1013 molecules}cm# argon.
Results
Radioactivity in urine and faeces
The excretion data after i.v. administration of ["%C]celecoxib to dog and monkey
are given in tables 1 and 2 respectively. The total recovery of radioactivity in urine
and faeces collected 0±168 h after dosing to EM dog, PM dog, cynomolgus monkey
and rhesus monkey was 85.3³1.67, 87.3³0.87, 91.4 and
dog, the majority of the radioactivity was excreted in the faeces (80.0³1.1% for
" 100% respectively. For
!
EM; 83.4³0.7% for PM), with 6% excreted in the urine (5.10³0.77% for EM;
.47³0.41% for PM). A greater percentage of the radioactivity was excreted in
urine of monkey (27.9% for cynomolgus; 24.9% for rhesus) with the remainder of
3
the dose excreted in faeces (63.5% for cynomolgus; 83.1% for rhesus).
Data on the excretion of radioactivity after oral administration of ["%C]celecoxib
to mouse and rabbit are given in tables 3 and 4 respectively. Total mean (³SEM)
percentage of administered radioactivity recovered was 93.5³1.9% for the male
mouse (22.4³5.4% in urine; 71.1³3.7% in faeces) and 92.4³1.3% for the female
mouse (45.7³0.8% in urine; 46.7³0.7% in faeces). In rabbit, total mean (³SEM)
percentage of administered radioactivity recovered was 97.2³0.4%, with
2
5.0³4.4% excreted in the urine and 72.2³4.6% excreted in the faeces.
Metabolites of celecoxib present in plasma, urine and faeces
Plasma. The plasma metabolite pro®les for the rabbit revealed the presence of
two metabolic pathways for celecoxib: oxidation of the methyl group or hy-
droxylation in the phenyl ring of celecoxib to form hydroxymethyl (M3) and
hydroxylphenyl (M4) metabolites respectively (®gure 1 and 2, and table 5). A
carboxylic acid metabolite (M2), formed from further oxidation of M3, as well as its
glucuronide (M1) were also in rabbit plasma. The plasma metabolic pro®les of
[
"%C]celecoxib diåered between the male and female mouse (table 5). In the male
mouse, celecoxib was the major circulating radioactive component and only minor
amounts of M3 (! 11%) and M2 (!
7%) metabolites were in plasma. In the female
mouse, celecoxib represented 54.4% (0.5 h) and 26.6% (8 h) of the radioactive
material in plasma. The female mouse had higher circulating levels of M2 and M3
than the male. Following i.v. administration of ["%C]celecoxib, M2 was the major
circulating radioactive component for both the cynomolgus and rhesus monkey after
1
5 min (table 6). The plasma metabolic pro®les were diåerent between the EM and
PM dog. The plasma of the PM dog contained primarily celecoxib with only minor
amounts of M3, whereas the plasma of the EM dog contained M1±4 as well as
celecoxib. Although, the plasma HPLC pro®le from the EM dog had radioactivity
eluting at a position similar to M4 (hydroxyphenyl metabolite of rabbits), the
structure was not con®rmed by LC}MS. There was no evidence for phenyl ring
hydroxylation in mouse, the PM dog or monkey.
Urine and faeces. Following i.v. administration of radiolabelled celecoxib to dog
and monkey, M2 was the major excretion product accounting for 96.2, 71.4, 79.3

Disposition of ["%C]celecoxib in several species
737
Table 1. Excretion of radioactivity in urine and faeces after a single i.v. dose of 5 mg}kg ["%C]celecoxib
to the EM and PM dog.
Radioactivity (% of administered dose)
Collection
interval
EM (n¯ 3)
PM (n ¯ 3)
(h)
Urine
Faeces
Urine
Faeces
0
±24
4±48
4.52³0.56
0.44³0.23
0.07³0.02
0.03³0.01
0.01³0.00
0.02³0.01
0.02³0.00
5.10³0.77
50.0³8.96
24.6³8.28
4.31³3.90
0.97³0.87
0.08³0.04
0.07³0.05
0.02³0.01
80.0³1.10
2.63³0.33
0.57³0.15
0.13³0.01
0.05³0.00
0.05³0.01
0.03³0.01
0.01³0.00
3.47³0.41
55.6³5.73
23.1³4.00
3.30³1.42
1.04³0.65
0.25³0.14
0.08³0.03
0.03³0.00
83.4³0.70
2
4
7
9
8±72
2±96
6±120
1
1
20±144
44±168
0
Cage wash
Total
±168
0.21
85.3³1.67
0.42
87.3³0.87
EM, extensive metabolizer of celecoxib; PM, poor metabolizer of celecoxib.
Table 2. Excretion of radioactivity in urine and faeces after a single i.v. dose of 1 mg}kg ["%C]celecoxib
to cynomologus and rhesus monkey.
Radioactivity (% of administered dose)
Collection
interval
Cynomolgus monkey (n ¯ 1)
Rhesus monkey (n¯ 1)
(h)
Urine
Faeces
Urine
Faeces
0
±24
4±48
18±9
8.1
0
0.6
0.1
0.1
0.1
27.9
0.02
6.4
34.2
13.0
7.9
1.8
0.2
63.5
21.6
0.5
2.6
0.1
0
0
0.1
24.9
18.5
56.8
7.2
0.5
0.1
0
2
4
7
9
8±72
2±96
6±120
1
1
20±144
44±168
0
Total
0
83.1
±168
91.4
" 100
Table 3. Excretion of radioactivity in urine and faeces after a single oral dose of 35 mg}kg ["%C]celecoxib
to mouse.
Radioactivity (% of administered dose)
Collection
interval
Male (n¯ 3)
Female (n¯ 3)
(h)
Urine
Faeces
Urine
Faeces
0
±24
4±48
19.0³6.8
1.1³0.3
0.43³0.17
0.07³0.01
0.17³0.04
0.03³0.01
0.27³0.06
62.8³4.3
6.4³3.2
0.9³0.5
0.3³0.2
0.4³0.1
0.1³0.0
0.1³0.0
40.2³0.9
2.7³0.7
0.9³0.1
0.2³0.1
0.2³0.1
0.05³0.01
0.4³0.3
42.0³1.1
3.3³0.5
0.6³0.2
0.3³0.2
0.2³0.1
0.1³0.04
0.1³0.05
2
4
7
9
8±72
2±96
6±120
1
1
20±144
44±168
Cage wash
±168
Total
0.17³0.04
1.04³0.3
0
22.4³5.4
71.1³3.7
45.7³0.8
46.7³0.7
93.5³1.9
92.4³1.3

7
38
S. K. Paulson et al.
Table 4. Excretion of radioactivity in urineand faeces after a single oral dose of 10 mg}kg ["%C]celecoxib
to female rabbit (n ¯ 3).
Radioactivity (% of
administered dose)
Time
(
h)
Urine
Faeces
0±24
4±48
8±72
2±96
19.0³1.5
27.0³10.2
2
4
7
9
1
1
5.19³3.09 38.2³6.4
0.46³0.20 6.05³1.20
0.26³0.09 0.69³0.06
6±120 0.03³0.00 0.17³0.02
20±144 0.01³0.01 0.04³0.01
44±168 0.01³0.00 0.02³0.01
0
±168 25.0³4.4
72.2³4.60
Total 97.2³0.4
Figure 2. Representative radiochromatogram of an extract of plasma obtained from a female rabbit
administered 10 mg}kg ["%C]celecoxib.

Disposition of ["%C]celecoxib in several species
739
Table 5. Metabolite pro®les in mouse and rabbit plasma after oral administration of ["%C]celecoxib.
Metabolite (%)
Sampling time (h)
Celecoxib
M1
M2
M3
M4
Female mousea
0
2
4
8
4
.5
54.4
33.7
48.5
26.6
na
nd
nd
nd
nd
na
13.9
22.0
17.9
39.6
na
30.0
39.2
30.4
26.6
na
nd
nd
nd
nd
na
2
Male mousea
0
2
4
8
4
.5
87.9
86.0
87.6
84.0
na
nd
nd
nd
nd
na
2.32
1.66
1.74
6.48
na
7.3
10.4
8.93
2.77
na
nd
nd
nd
nd
na
2
Female rabbitb
0
2
4
8
4
.5
51.5
37.8
na
28.9
4.93
10.8
33.1
na
7.40
37.5
37.7
25.4
na
33.0
48.2
nd
0.71
na
nd
1.92
nd
2.85
na
9.38
7.40
2
a Values are from pooled samples from n 3 animals.
¯
b Values for the 0.5-, 4-, 8- and 24-h samples are from pooled samples from
n¯ 3 animals. The value for the 2-h sample represents mean of samples from
n¯ 3 animals.
na, Sample was not analyzed due to low levels of radioactivity in mouse
plasma at 24 h. A 4-h plasma sample was not collected for rabbit.
nd, no radioactivity detected in peak.
Table 6. Metabolite pro®les in plasma after i.v. administration of 1 mg}kg ["%C]celecoxib to rhesus and
cynomolgus monkey and i.v. administration of 5 mg}kg ["%C]celecoxib to EM and PM dog.
Metabolite (%)
Sampling time (h)
Celecoxib
M1
M2
M3
M4
Cynomolgus monkey
0
0
0
2
3
4
6
.083
.25
.5
72.6
34.4
28.6
34.2
20.5
24.7
9.2
nd
nd
nd
nd
nd
nd
nd
27.4
64.0
70.1
64.8
79.5
75.3
90.1
0
1.57
1.31
0.624
0
nd
nd
nd
nd
nd
nd
nd
0
0
Rhesus monkey
3
4
EM dog
36.7
30.2
nd
nd
59.1
69.8
0
0
nd
nd
0
4
.5
78.4
50.3
2.24a
2.36a
5.30
5.25
9.47
27.5
0.36b
nd
PM dog
0
4
.5
94.0
93.8
nd
nd
nd
nd
1.77
5.78
nd
nd
a Metabolite elutes at a similar position to M1.
b Metabolite elutes at a similar position to M4.
nd, no radioactivity detected in peak.

7
40
S. K. Paulson et al.
Table 7. Percentage of the dose excreted as celecoxib or metabolites following a single i.v. dose given
to dog, cynomolgus and rhesus monkey.
Metabolite (%)
Collection
Species
interval (h)
Celecoxib
M1
M2
M3
M4
% of the Radioactive doses in urine
Dog (EM)
Dog (PM)
0±72
0±72
0±24
0±24
nd
nd
nd
nd
0.15b
0.07b
nd
nd
4.6
3.0
18.7
21.1
nd
0.03
nd
0.07a
0.07a
nd
nd
Cynomolgus monkey
Rhesus monkey
of the Radioactive dose in faeces
Dog (EM)
nd
%
0±72
0±72
0±96
0±48
nd
nd
nd
nd
nd
nd
nd
nd
78.5
76.3
52.7
75.1
1.1
2.0
nd
nd
nd
nd
nd
nd
Dog (PM)
Cynomolgus monkey
Rhesus monkey
EM, extensive metabolizer of celecoxib; PM, poor metabolizer of celecoxib.
a Metabolite elutes at a similar position on the radiochromatogram as M4.
b Metabolite elutes at a similar position on the radiochroamtogram as M1.
nd, no radioactivity detected in peak.
Table 8. Percentage of the dose excreted as celecoxib or metabolites following oral administration of a
single dose to mouse and rabbits.
Metabolite (%)
Collection
Species
interval (h)
Celecoxib
M1
M2
M3
M4
%
of the Radioactive dose in urine
Mouse (M)
Mouse (F)
Rabbit
0±48
0±48
0±48
nd
nd
nd
nd
nd
1.6
19.6
42±1
21.9
nd
nd
nd
nd
nd
nd
%
of the Radioactive dose in faeces
Mouse (M)
Mouse (F)
Rabbit
0±48
0±48
0±48
8.3
0.4
0.5
nd
nd
0.1
58.4
42.9
63.9
1.2
0.7
nd
nd
nd
nd
nd, no radioactivity detected in peak.
and 83.1% of the dose in urine and faeces of rhesus and cynomolgus monkey, and
PM and EM dog respectively (table 7). Minor amounts of M3 (!
2%) and the M1
1%) were excreted in dog faeces. Dog, cynomolgus or rhesus monkey excreted
(!
no unchanged drug.
M2 was also a major excretion product of male and female mouse and rabbit,
accounting for
respectively (table 8). Minor amounts (C
C 78.0, 85.0 and 85.8% of the dose excreted in the ®rst 48 h
1%) of M3 was excreted in faeces of both
the male and female mouse. The female mouse diåered from the male regarding the
pathway of excretion of celecoxib and its metabolites. In the male, M2 represented
1
9.6% of the dose excreted in urine and 58.4% of the dose excreted in faeces,
whereas in the female, 42.1% of the dose excreted in urine and 42.9% of the dose
excreted in faeces was M2. Celecoxib (8.3% of the dose) was excreted unchanged in
the faeces of the male, representing either unabsorbed dose or biliary excretion of the
unchanged drug. Only 0.4% of the dose was excreted as celecoxib following oral
administration to the female.

Disposition of ["%C]celecoxib in several species
Structure elucidation of metabolites
741
The identity of celecoxib and its metabolites was elucidated by MS}MS.
Because the rabbit produced all of the metabolites of celecoxib, details of the
structure elucidation for only that species are presented. Representative radio-
chromatograms of plasma and urine samples from a rabbit administered
[
"%C]celecoxib are shown in ®gures 2 and 3. Ionspray ionization mass spectral data
for celecoxib metabolites are shown in table 9.
Celecoxib. The HPLC retention time of the celecoxib peak was identical to that of
an authentic celecoxib standard. The deprotonated molecular ion at 380 daltons in
its negative-ion mass spectrum was consistent with that of the celecoxib standard.
The CID spectrum of the negative-ion at m}z 380 generated a series of product ions
identical to those in the CID spectrum of the celecoxib standard.
M1. The HPLC retention time of M1 was identical to that of an authentic
standard of the 1-O-glucuronide conjugate of M1. The mass spectrum of M1 was
identical to the mass spectrum of the synthetic standard of carboxylic acid celecoxib
glucuronide (®gure 1 and table 9). The HPLC and MS data indicate that M1
consisted of glucuronide conjugates of the carboxylic acid metabolite of celecoxib.
The four peaks of glucuronide conjugates of M1 observed in the HPLC pro®les of
rabbit urine are likely to be positional isomers (Spahn-Langguth and Benet 1992,
Sidelmann et al. 1996).
M2. M2 had the same HPLC retention time as the authentic standard of the
carboxylic acid metabolite of celecoxib. The deprotonated molecular ion of M2 at
m}z 410 in its negative-ion mass spectrum, 30 daltons higher than the parent
compound celecoxib, suggested that M2 was a carboxylic acid metabolite. The CID
spectrum of the negative-ion at m}z 410 generated a series of product ions similar to
those in the CID spectrum of the synthetic carboxylic acid celecoxib (®gure 1 and
table 9). These data indicate that M2 was a carboxylic acid metabolite of celecoxib.
M3. M3 had the same HPLC retention time as the hydroxymethyl celecoxib
standard. The deprotonated molecular ion of M3 at m}z 396 in its negative-ion mass
spectrum was consistent with that of the synthetic hydroxymethyl metabolite of
celecoxib. The CID spectrum of M3 (m}z 396) generated a series of product ions
matching those in the CID spectrum of the synthetic hydroxymethyl celecoxib
(
®gure 1 and table 9). These results con®rmed that M3 was a monohydroxylated
metabolite of celecoxib in which the hydroxylation occurred on the methyl moiety
of the 5-(4-methyl)phenyl group.
M4. The HPLC retention time of the M4 peak was
C 2 min longer than that of
M3. The deprotonated molecular ion of M4 at m}z 396 is 16 daltons higher than that
of celecoxib, indicating that another monohydroxylated metabolite of celecoxib,
diåerent from M3, was formed. The CID spectrum of M4 (m}z 396) produced a
series of product ions at m}z 332, 317, 312, 302, 297, 277, 178, 160 and 69, which was
also diåerent from the CID spectrum of M3. The sequential losses of 64 (SO ), 15
(
#
NH), 20 (HF) and 20 (HF) daltons from the m}z 396 produced the product ions at
m}z 332, 317, 297 and 277 respectively. The base peak at m}z 302 was formed by a
loss of 30 (CH ) daltons from the m}z 317 ion, suggesting that the hydroxyl group
$

7
42
S. K. Paulson et al.
Figure 3. Representative radiochromatogram of urine from a female rabbit administered 10 mg}kg
"%C]celecoxib.
[
Table 9. Ionspray inonization mass spectral data for celecoxib metabolites.
Metabolite
M1
[M±H]­
CID mass spectral data, m}z (%)
586
410 (28), M-H-C H O ;
’
) ’
3
3
2
1
1
66 (60), M-H-C H 0 -CO ;
) ’
02 (28), M-H-C H O -CO -SO ;
) ’
82 (7), M-H-C H O CO -SO -HF;
) ’
’
#
’
#
#
’
#
#
75 (5), C H O ;
( ’
13 (100), C H O -CH OHCH OH.
( ’
’
’
#
#
M2
410
366 (32), M-H-CO ;
#
3
2
2
2
1
1
6
02 (100), M-H-CO -SO ;
# #
82 (74), M-H-CO -SO -HF;
62 (98), M-H-CO -SO -2HF;
33 (52), M-H-CO -SO -CF ;
79 (48), M-H-SO NH-OCH C H C-HF;
# ’
59 (26), M-H-SO NH-OCH C H C-2HF;
# ’
#
#
#
#
#
#
$
#
%
#
%
9 (82), CF .
$
M3
396
396 (12), M-H
3
3
2
2
2
1
1
6
66 (6), M-H-CH O;
02 (100), M-H-CH O-SO ;
#
#
82 (48), M-H-CH O-SO -HF;
#
#
62 (45), M-H-CH O-SO -2HF;
#
#
33 (25), M-H-CH O-SO -CF ;
#
#
#
$
79 (23), M-H-SO NH-OCH C H C-HF;
# ’
59 (10), M-H-SO NH-OCH C H C-2HF;
# ’
#
%
#
%
9 (35), CF .
$
M4
396
396 (9), M-H
3
3
3
3
2
2
1
1
6
32 (11), M-H-SO ;
17 (35), M-H-SO -NH;
#
#
12 (11), M-H-SO -HF;
#
02 (100), M-H-SO -CHOH;
#
97 (24), M-H-SO -CH -HF;
#
77 (8), M-H-SO -CH -2HF;
$
#
$
78 (15), M-H-SO NH-CH C H OHC-HF;
$ ’
60 (10), M-H-SO NH-CH C H OHC-2F;
$ ’
#
$
#
$
9 (23), CF .
$

Disposition of ["%C]celecoxib in several species
743
was at the phenyl moiety of the 5-(4-methyl)phenyl group. The product ion at m}z
9 was related to the CF group. The MS results suggested that M4 was a
monohydroxylated metabolite of celecoxib with the hydroxylation occurring on the
6
$
phenyl moiety of the 5-(4-methyl)phenyl group (®gure 1).
Discussion
The metabolism and excretion of celecoxib was investigated following i.v.
administration to dog, cynomolgus and rhesus monkey, and after oral administration
to mouse and rabbit. Celecoxib was extensively metabolized by all species studied.
Only minor species diåerences in the metabolic pathway for celecoxib were
observed. A gender diåerence in the disposition of celecoxib was observed in mouse.
Two phase I oxidative pathways for the metabolism of celecoxib were demonstrated
(
®gure 1). The major phase I pathway that occurred in mouse, rabbit, cynomolgus
and rhesus monkey, and dog involved oxidation of the aromatic methyl group of
celecoxib to form a hydroxymethyl metabolite (M3). M3 underwent further
oxidation to the carboxylic acid metabolite (M2). M2 was produced by all the species
studied. Oxidation of the aromatic methyl group of celecoxib was the metabolic
pathway for elimination of celecoxib in humans and rat (Paulson et al. 2000a, b). A
minor phase I pathway was the ring hydroxylation of the methylphenyl group of
celecoxib to form M4, which was shown to occur only in rabbit and, possibly,
EM dog.
In rabbit, the carboxylic acid metabolite of celecoxib (M2) was conjugated with
glucuronic acid to form the O-glucuronide conjugate (M1). M1 was only a minor
metabolite of celecoxib in the rabbit representing
was also only a minor metabolite in humans accounting for
! 2% of the dose excreted. M1
! 2% of the dose
excreted (Paulson et al. 2000a). Conjugation of the hydroxy and the carboxylic acid
metabolite of celecoxib also occurs in rat. Rat also produced a glucuronide conjugate
of the hydroxymethyl metabolite (Paulson et al. 2000b).
There are two populations of dogs that have been reported as being PM or EM
of celecoxib (Paulson et al. 1999). Qualitatively the metabolism of celecoxib was
similar between the two populations with the primary metabolic pathway being
hydroxylation of the aromatic methyl group. The polymorphism was associated
primarily with the formation of M3 catalysed via liver microsomal CYP450.
However, interestingly the EM dog contains a metabolite in plasma that elutes at a
position similar to M4 (ring-hydroxylation) on a HPLC radiochromatogram that
was not observed in plasma of the PM dog. Greater amounts of metabolites were
found circulating in the EM dog than in the corresponding PM animal.
A gender diåerence in the disposition of celecoxib was observed in mouse.
Although the metabolic pathway for celecoxib was the same for the male and
female, metabolism was slightly more extensive in the female, with 0.4% of the dose
excreted as unchanged celecoxib compared with 8.3% excreted as unchanged drug
in the male. The male also had higher circulating levels of metabolites than the
female.
Monkeys metabolize celecoxib via a single pathway with the formation of M3
and M2. The carboxylic acid metabolite (M2) appears to be the major circulating
component of plasma and the major excretion product. However, since only a single
cynomolgus monkey and rhesus monkey were studied, there can be no assessment of
interanimal variability in this species.

7
44
Disposition of ["%C]celecoxib in several species
After i.v. dosing to monkey and dog, excretion of the radioactivity was near
complete in rhesus monkey and EM and PM dog by 48 h after dose administration.
In the cynomolgus monkey, most of the dose was excreted by 5 days after dose
administration. Most of the radioactivity was excreted in the faeces, suggesting that
for dog and monkey the major route of elimination for celecoxib is metabolism
followed by excretion of the metabolites in bile. Radiolabeled celecoxib was
administered by oral gavage to mouse and rabbit. Excretion of radioactivity was near
complete by 48 h in these species. Higher urinary excretion occurred in female
(
45.7%) compared with male (22.4%) mouse.
In conclusion, celecoxib was eliminated primarily by a single metabolic pathway,
i.e. hydroxylation of the aromatic methyl group of celecoxib and further oxidation
of the hydroxylated metabolite to a carboxylic acid metabolite. The major route of
elimination of celecoxib was metabolism followed by excretion of the metabolites in
the faeces, with the remainder of the metabolites excreted by the kidney.
Acknowledgements
Radiolabeled compounds were synthesized by Scott Harring at the Radio-
chemistry Department at G. D. Searle.
References
Crofford, L. J., Wilder, R. L., Ristimaki, A. P., Sano, H., Remmers, E. F., Epps, H. R. and Hla, T.,
1994, Cyclooxygenase-1 and -2 expression in rheumatoid synovial tissues: eåects of interleukin-
1b, phorbol ester, and corticosteroids. Journal Clinical Investigation, 93, 1095±1101.
Isakson, P., Zweifel, B., Masferrer, J., Koboldt, C., Seibert,. K, Hubbard, R., Geis, S. and
Needleman, P. 1998, Speci®c COX-2 inhibitors: from bench to bedside. In J. Vane and J.
Botting (eds), Selective COX-2 Inhibitors: Pharmacology, Clinical Eåects and Therapeutic
Potential (London: Kluwer and William Harvery), 127±133.
O’Neill, G. P. and Ford-Hutchinson, A. W., 1993, Expression of messenger mRNA for cyclo-
oxygenase-1 and cyclooxygenase-2 in human tissues. FEBS Letters, 330, 156±160.
Paulson, S. K., Engel, L., Reitz, B., Bolten, S., Burton, E. G., Maziasz, T. J., Yan, B. and
Schoenhard, G. L., 1999, Evidence for polymorphism in the canine metabolism of the
cyclooxygenase 2 inhibitor, celecoxib. Drug Metabolism Disposition, 27, 1133±1142.
Paulson, S. K., Hribar, J. D., Liu, N. W. K., Hajdu, E., Bible, R. H., Piergies, A. and Karim, A.,
000a, Metabolism and excretion of ["%C] celecoxib in healthy male volunteers. Drug Metabolism
2
Disposition, 28, 308±314.
Paulson, S. K., Zhang, J., Breau, A. P., Hribar, J. D., Liu, N. W. K., Jessen, S. M., Lawal, Y. M.,
Cogburn, J. N., Gresk, C. J., Markos, C. S., Schoenhard, G. L. and Burton, E. G., 2000b,
Pharmacokinetics, tissue distribution, metabolism and excretion of celecoxib in rats. Drug
Metabolism and Disposition, 28, 514±521.
Penning, T. D., Talley, J. J., Bertenshaw, S. R., Carter, J. D., Collins, P. W., Docter, S.,
Graneto, M. J., Lee, L. F., Malecha, J. W., Miyashiro, J. M., Rogers, R. S., Rogier, D. J.,
Yu, S. S., Anderson, G. D., Burton, E. G., Cogburn, J. N., Gregory, S. A., Koboldt, C. M.,
Perkins, W. E., Seibert, K., Veenhuizen, A. W., Zhang, Y. Y. and Isakson, P. C., 1997,
Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors:
identi®cation of 4-[5-(4-methylphenyl)-3-(tr i¯ uoromethyl)-1H-pyrazol-1-yl]bezenesulfonamide
(
SC-58635, celecoxib). Journal Medical Chemistry, 40, 1347±1365.
Seibert, K., Zhang, Y., Leahy, K., Hauser, S., Masferrer, J., Perkins, W., Lee, L. and Isakson,
P. C., 1994, Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in
in¯ammation and pain. Proceedings of the National Academy of Sciences, USA, 91, 12013±12017.
Sidelmann, U. G., Lenz, E. M., Spraul, M., Hofmann, M., Troke, J., Sanderson, P. N., Lindon,
J. C., Wilson, A. D. and Nicholson, J. K., 1996, 750 MHz HPLC-NMR spectroscopic studies
on the separation and characterisation of the positional isomers of the glucuronides of 6,11-
dihydro-11-oxodibenz[b,e]oxepin-2-acetic acid. Analytical Chemistry, 68, 106±110.
Simon, L. S., Lanza, F. L., Lipsky, P. E., Hubbard, R. C., Talwalker, S., Schwartz, B. D., Isakson,
P. C. and Geis, G. S., 1998, Preliminary study of the safety and eæcacy of SC-58635, a novel
cyclooxygenase 2 inhibitor, Arthritis Rheumatology, 41, 1591±1602.
Spahn-Langguth, H. and Benet, L., 1992, Acyl glucuronides revisited: Is the glucuronidation process
a toxi®cation as well as a detoxi®cation mechanism? Drug Metabolism Reviews, 241, 5±48.