Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Nonsteroidal antiinflammatory drugs are widely used for the treatment of pain, inflammation, and fever. Their mechanism of action involves the inhibition of COX¹, a hemeprotein that exists in two forms (COX-1 and COX-2) and converts arachidonic acid to pro-inflammatory prostaglandins and their subsequent metabolic products (Donnelly and Hawkey, 1997; Jouzeau et al., 1997; Lane, 1997; Lipsky and Isakson, 1997). Because the products of COX-1 are cytoprotective, selective inhibition of COX-2 can reduce inflammation with less of the gastrointestinal side effects characteristic of nonselective nonsteroidal antiinflammatory drugs. Therefore, much attention has been focused on the design of potent and selective COX-2 inhibitors like etoricoxib (5-chloro-6'-methyl-3-[4-(methylsulfonyl)phenyl]-2,3'-bipyridine; ARCOXIA² ), which has been developed for acute and chronic analgesia, the treatment of the signs and symptoms of osteoarthritis, rheumatoid arthritis, dysmenorrhea, and gouty arthritis (Chauret et al., 2001; Ouellet et al., 2001; Riendeau et al., 2001; Sorbera et al., 2001).

Etoricoxib pharmacokinetics are linear at clinically relevant doses and the pharmacokinetic half-life, and pharmacological response, supports oral once-a-day dosing (Agrawal et al., 2001, 2002). In addition, the results of in vitro metabolism studies have indicated that etoricoxib is metabolized via 6'-methyl hydroxylation and 1'-N-oxidation (Fig. 1). These metabolites, which do not inhibit COX-1 and do not contribute significantly to the inhibition of COX-2, are formed by P450s (Chauret et al., 2001; Kassahun et al., 2001). Although the in vitro metabolism of etoricoxib has been described, its absorption, disposition, and metabolism in man have yet to be reported. Therefore, the study presented here describes the absorption and disposition of [¹⁴C]etoricoxib in human subjects. The specific objectives of this study were 1) to investigate the routes of elimination of etoricoxib in healthy subjects following single oral and i.v. doses of [¹⁴C]etoricoxib, 2) to quantitate concentrations of total radioactivity and etoricoxib in plasma after single oral and i.v. doses of radiolabeled etoricoxib, 3) to examine the metabolism of etoricoxib in humans, 4) to demonstrate mass balance for [¹⁴C]etoricoxib in humans, and 5) to estimate the absolute bioavailability of etoricoxib administered as an oral solution.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Proposed metabolic pathways of etoricoxib in healthy human volunteers. Bolded arrows indicate major pathways. The asterisk denotes the site of the carbon-14 label.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. [¹⁴C]Etoricoxib (>99% radiochemical purity) was synthesized by the Labeled Compound Synthesis Group (Merck Research Laboratories, Rahway, NJ). The Carbon-14 label was incorporated at the methyl group of the methane sulfonyl phenyl moiety (Fig. 1). Standards of the 1'-N'-oxide, 6'-hydroxymethyl, 6'-hydroxymethyl-1'-N'-oxide, and 6'-carboxy metabolites of etoricoxib were provided by Drs. Y. LeBlanc and P. Roy (Merck-Frosst Center for Therapeutic Research, Quebec, Canada). Unlabeled etoricoxib was synthesized by the Department of Process Research (Merck Research Laboratories). -Glucuronidase (Helix pomatia, Type H-5) was obtained from Sigma-Aldrich (St. Louis, MO). All other commercially available reagents and solvents were of either analytical or HPLC grade.

Human Studies. The study protocol was approved by an Ethics Review Board. All subjects understood the procedures and agreed to participate in the study by giving written informed consent (Declaration of Helsinki).

Measurement of Total Radioactivity. Plasma and urine samples were thawed at room temperature and mixed thoroughly. After centrifugation of plasma to remove the fibrin, aliquots (0.5 or 1.0 ml) were transferred to polyethylene vials, and 15 ml of scintillation cocktail (Ready-Safe; Beckman Coulter, Inc., Fullerton, CA) was added for counting in a Beckman LS 5000CE liquid scintillation spectrometer (Beckman Coulter Inc.). Urine samples were thawed overnight at 4°C, and 1-ml aliquots were transferred to vials for counting as described for plasma.

Preparation of Plasma Samples for Etoricoxib Assay. Concentrations of etoricoxib in plasma samples were determined using previously described methodology (Matthews et al., 2001). In brief, the method involves solid-phase extraction in the 96-well format of etoricoxib from plasma followed by HPLC with postcolumn photochemical cyclization and fluorescence detection of the plasma extracts. The method is linear over the concentration range of 5 to 500 ng/ml etoricoxib. Based on the replicate analyses of spiked standards, the within-day precision was better than 7% at all points on the calibration curve, whereas within-day accuracy was within 5% of nominal at all standard concentrations. The lower limit of quantitation was 5.0 ng/ml.

Preparation of Plasma, Urine, and Fecal Samples for Metabolite Profile Analysis.

Plasma Samples collected at 0.25, 0.5, 0.75, 1, 2, 4, 6, 8, 10, 12, 15, 18, 21, and 24 h after intravenous (25 mg) administration, and at 0.5, 1, 2, 4, 6, 8, 10, 12, 15, 18, 21, and 24 h after oral (100 mg) dosing, were thawed at room temperature. For each subject, aliquots (0.5 ml, i.v.; 1.0 ml, p.o.) from each sample were pooled for a total volume of 7 ml (i.v.) and 12 ml (p.o.) plasma. An aliquot (0.5 ml) was taken from each pooled sample for liquid scintillation counting. The plasma proteins in the remaining samples were precipitated by the addition of four volumes of 1.5% glacial acetic acid in acetonitrile/methanol (2:1) while vortexing vigorously. After centrifugation, the supernatants were transferred to clean tubes, and an aliquot (1 ml) was taken for liquid scintillation counting. Recovery of radioactivity after extraction was 90%. The supernatants were evaporated to dryness in the SpeedVac concentrator. The residues were reconstituted in 400 µl of 25 mM ammonium acetate (pH 7), and the solutions were transferred to autosampler vials for radiochromatography. The injection volume was 300 µl. The remaining material was submitted for evaluation by mass spectrometry.

Urine. Samples from 0- to 2-, 2- to 4-, 4- to 6-, 6- to 9-, 9- to 12-, 12- to 18-, and 18- to 24-h collection intervals after i.v. and oral administration (all subjects) were thawed at room temperature and mixed vigorously. Aliquots from each time interval were combined in proportion to their respective volumes to produce a representative 0- to 24-h sample for each subject. A 200-µl aliquot of the pooled mixture was taken for scintillation counting. To concentrate the urine samples prior to radiochromatography, aliquots (5 ml) were transferred to glass centrifuge tubes, and trifluoroacetic acid (15 µl) was added to acidify the samples to approximately pH 3. Acetonitrile (20 ml) was added with mixing, and particulate material was removed by centrifugation (~3000 rpm, 10 min). The supernatants were transferred to 50-ml glass tubes, and the solvent was removed in a centrifugal vacuum concentrator (SpeedVac, Savant Instruments, Holbrook, NY). The residues were reconstituted in 300 µl of 20% acetonitrile in 25 mM ammonium acetate, pH 7 (v/v), and the solutions were sonicated and transferred to autosampler vials for radiochromatography (injection volume of 200 µl). The remaining material of selected samples was submitted for evaluation by mass spectrometry. Urine samples from 24- to 36- and 36- to 48-h intervals were combined in proportion to their respective volumes to provide a representative 24- to 48-h sample. The samples (5 ml each) were extracted and prepared for chromatography as above. Selected samples were submitted for evaluation by mass spectrometry.

Feces. Homogenized fecal samples (selected subjects) were thawed at room temperature and mixed thoroughly. Duplicate 1-ml aliquots were transferred to glass tubes and 4 ml of 1.5% glacial acetic acid in acetonitrile/methanol (2:1) was added while mixing. The samples were homogenized with a hand-held tissue homogenizer (Tissue Tearor, Biospec Products Inc., Bartlesville, OK) and sonicated for 2 min, and this homogenization-sonication step was then repeated. After centrifugation, the supernatants were transferred to clean tubes, and aliquots (100 µl) were taken for liquid scintillation counting. Recovery of radioactivity was good (>95%). The supernatants were evaporated to dryness in the SpeedVac concentrator, and the residues were reconstituted in 250 µl of 25 mM ammonium acetate (pH 7) followed by the addition of 50 µl of acetonitrile. The samples were transferred to autosampler vials for radiochromatography (200 µl injected), and the remaining reconstituted material was submitted for evaluation by mass spectrometry.

Conditions for Radiochromatography of Plasma, Urine, and Fecal Extracts. The above samples were subjected to chromatography on a HP1100 chromatograph system (Hewlett Packard, Palo Alto, CA). The metabolites were separated on a Zorbax Eclipse XDS C₈ column (4.6 × 250 mm) eluted at a flow rate of 1 ml/min using a linear gradient starting at 20% acetonitrile in 25 mM ammonium acetate (pH 7) and increasing to 50% acetonitrile at 1%/min for a total run time of 30 min. The column was maintained at 40°C in an oven compartment. A postrun time of 5 min under starting mobile phase conditions permitted re-equilibration of the column. The effluent was monitored by UV at 280 and 236 nm (photodiode array detector) and by an in-line radiochemical detector (-RAM model 2B, IN/US, Tampa, FL), using a 3 ml/min flow rate for the scintillation cocktail (flow rate of 4 ml/min through the radiochemical detector).

Fractionation of Pooled Urine for Metabolite Identification by LC-MS/MS. Aliquots (~20 ml) of pooled 0- to 24-h urine were acidified to pH 3 with trifluoroacetic acid, and 5 × 1.1 ml was transferred to autosampler vials for injection (2 × 500 µl per vial) onto the HPLC column using the conditions described above. The effluent was collected with the aid of a fraction collector (ISCO Foxy-200; Isco Inc., Lincoln, NE) as 1-min fractions (1 ml/min). A total of 10 injections were made. An aliquot (0.5 ml) from each fraction (10 ml of total volume) was taken for liquid scintillation counting, and a histogram was constructed. Those fractions containing 2% of the total radioactivity were evaporated to dryness using a rotary evaporator. The residues were dissolved in acetonitrile, or a 2:1 mixture of acetonitrile and water, and the solutions were transferred to autosampler vials. Solvent was evaporated to dryness (N₂, 44°C), and the residues were submitted for mass spectrometry.

Conditions for LC-MS/MS. Chromatography was conducted on a HP1090 chromatograph (Hewlett Packard) interfaced to a Finnigan TSQ tandem mass spectrometer (Thermo Finnigan, San Jose, CA). Separation of metabolites was carried out on a Phenomenex Luna C₁₈ (2) column (2.0 × 250 mm, 5 µm; Phenomenex, Torrance, CA) using a mobile phase consisting of 0.1% formic acid in Milli-Q water (solvent A) and 0.1% formic acid in acetonitrile (solvent B) at a constant flow rate of 0.2 ml/min. The gradient conditions were as follows: isocratic conditions starting at 10% B were maintained for 3 min followed by a linear increase to 25% B over 2 min. The concentration of B was slowly increased to 35% over the next 20 min, followed by a rapid increase to 80% B over 2 min. The column was washed at 80% B for 5 min, and conditions were returned to 10% B over 3 min. The system was equilibrated for 10 min at 10% B prior to each injection. Total run time was 27 min prior to washing and equilibration of the column.

Chromatographic Conditions to Separate Two Coeluting Metabolites. The 6'-carboxylic acid metabolite and the O--D-glucuronide of 6'-hydroxymethyl-etoricoxib were found to coelute using the neutral pH chromatographic conditions described above. Changing the elution conditions to an acidic mobile phase permitted separation of the two components. Under these conditions, the amount of the 6'-hydroxymethyl-etoricoxib glucuronide could be estimated.

Incubation of Urine and Fecal Homogenates with -Glucuronidase. Two-milliliter aliquots of urine (0-24 h) and 0.9 ml of fecal homogenates were transferred to glass tubes. One milliliter of -glucuronidase solution (2 mg/ml in 0.2 M sodium acetate, pH 5.2) was added to each tube. A second set of samples was prepared to which 1 ml of the sodium acetate buffer was added to serve as control. All the samples were incubated overnight at 37°C in a shaking water bath. The reaction was stopped with the addition of 4 volumes of 1.5% acetic acid in acetonitrile/methanol (2:1). After centrifugation, the supernatants were transferred to clean tubes, and the solvent was evaporated to dryness in the SpeedVac concentrator. The residues were reconstituted in 300 µl (urine) and 500 µl (feces) of 20% acetonitrile in 25 mM ammonium acetate, pH 7 (v/v), and the solutions were sonicated and transferred to autosampler vials for radiochromatography under neutral pH conditions. Injection volumes were 200 µl (urine) and 400 µl (feces).

Identification and Quantitation of Metabolites. Authentic standards of etoricoxib ([M + H]⁺ ion at m/z 359) and its phase I metabolites (the 6'-hydroxymethyl, 6'-carboxylic acid, 1'-N-oxide, and 6'-hydroxymethyl-1'-N-oxide derivatives) were available to support the characterization of etoricoxib metabolites by similarity of chromatographic retention time and their respective UV spectra. Further support of their identification was provided by mass spectroscopy (6'-hydroxymethyl, [M + H]⁺ ion at m/z 375; 6'-carboxylic acid, [M + H]⁺ ion at m/z 389; 1'-N-oxide, [M + H]⁺ ion at m/z 375; and 6'-hydroxymethyl-1'-N-oxide, [M + H]⁺ ion at m/z 391). The structure of the glucuronide conjugate of 6'-hydroxymethyl-etoricoxib was verified by mass spectroscopy. The [M + H]⁺ ion at m/z 551 was consistent with a glucuronic acid conjugate of a mono-oxygenated metabolite of etoricoxib. The product ion spectrum generated upon collision-induced dissociation of the [M + H]⁺ ion was consistent with a structural assignment of an O--D-glucuronide of the 6'-hydroxymethyl metabolite.

Pharmacokinetic Methods. Plasma concentrations of etoricoxib and radioactivity, and actual sampling times relative to the etoricoxib dose, were used to estimate pharmacokinetic parameters for each treatment in each subject (AUC, t_1/2, Cl, and V_dss for the i.v. dose; and AUC, C_max, T_max, and apparent t_1/2 for the oral dose). The C_max and T_max were obtained by inspection of the concentration-time data. The terminal t_1/2 was estimated from the best-fit parameters of a single exponential to the log-linear portion of the plasma concentration-time curve. The best-fit parameters were obtained using nonlinear regression by the Simplex algorithm with weight = 1 (Yeh and Remphrey, 1990). The area under the plasma concentration-time curve (AUC) and the area under the first moment curve (AUMC) (following i.v. administration) were calculated using the linear trapezoidal method up to the last measured concentration. The additional area was estimated from the last measured concentration, and the value of t_1/2 (<3% of the total AUC was extrapolated). Plasma clearance was calculated from i.v. data as the ratio of dose to AUC. For each individual, the actual i.v. dose administered was used (the weight of the dosing solution administered divided by the formulation density multiplied by the assayed potency of the dosing solution administered; i.v. potency = 0.746 mg/ml; and formulation density = 1.005 g/ml). The steady-state volume of distribution (V_dss) was calculated as
where MRT is the mean residence time and  is the infusion time.

Statistical Methods. Pharmacokinetic parameters (AUC for the i.v and oral solutions and C_max for the oral solution) for etoricoxib and total radioactivity following a single dose were analyzed using an analysis of variance (ANOVA) model containing factors for subject and treatment (equivalent to a paired t test). The natural log transformation was applied to AUC and C_max data. Prior to statistical analyses, the AUC and C_max data were adjusted for the actual dose administered.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Absorption and Excretion of [¹⁴C]Etoricoxib. The mean concentration-time profiles of etoricoxib and radioactivity in plasma following i.v. and oral administration are shown in Fig. 2, and the geometric mean AUC ratios of etoricoxib-to-total radioactivity (95% CI) were 0.73 (0.70, 0.76) and 0.75 (0.70, 0.82), respectively, indicating that etoricoxib accounts for the majority of the radioactivity in plasma following either route of administration. Following the i.v. dose, clearance was low (57 ml/min), and the harmonic mean t_1/2 was 24.8 h (Table 1). The absolute bioavailability of the oral solution was estimated to average 83%, with a T_max and C_max of 1 h and 1.36 µg/ml, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Mean concentrations of etoricoxib (nanograms per milliliter) and total radioactivity (nanograms · equivalents per milliliter) in plasma following administration of a single dose of [14C]etoricoxib to six healthy male volunteers: (A) i.v. dose (25 mg, 100 µCi), and (B) p.o. dose (100 mg, 100 µCi).

Metabolism of [¹⁴C]Etoricoxib. Radiochromatograms of 0- to 24-h pooled urine from the six subjects were qualitatively and quantitatively similar following i.v. and p.o. administration of [¹⁴C]etoricoxib (Fig. 3). Mass spectral analysis of concentrated urine samples, and of isolated metabolite fractions, demonstrated that the 6'-carboxylic acid metabolite and the O--D-glucuronide conjugate of 6'-hydroxymethyl-etoricoxib both contributed to the radioactivity of the major peak (retention time = 9 min). By changing from a Zorbax C₈ base-deactivated column to a Phenomenex Luna C₁₈(2) column, and eluting with an acidic mobile phase, the two components were separated (Fig. 3), permitting quantitation of the glucuronide conjugate in the samples.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Representative radiochromatograms of extracts from urine (0-24 h) of a human subject following i.v. (A) and oral (B) administration of [14C]etoricoxib obtained under neutral pH mobile phase conditions and an extract run under acidic mobile phase conditions (C). Metabolite peaks are identified as 1, 6'-carboxy-etoricoxib; 2, 6'-hydroxymethyl-etoricoxib O--D-glucuronide; 3, 6'-hydroxymethyl-etoricoxib-1'-N-oxide; 4, etoricoxib-1'-N-oxide; and 5, 6'-hydroxymethyl-etoricoxib.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Representative radiochromatograms of extracts of 0 to 24 h urine (A), day 2 feces (B), and pooled (0.25 to 24 h) plasma (C) of a human subject following i.v. administration of [14C]etoricoxib (25 mg, 100 µCi) obtained under neutral mobile phase conditions. Metabolite peaks are identified as 1, 6'-carboxy-etoricoxib; 2, 6'-hydroxymethyl-etoricoxib O--D-glucuronide; 3, 6'-hydroxymethyl-etoricoxib-1'-N-oxide; 4, etoricoxib-1'-N-oxide; and 5, 6'-hydroxymethyl-etoricoxib.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The purpose of this study was to investigate the absorption, disposition, and mass balance of [¹⁴C]etoricoxib in healthy male volunteers. Mass balance was reasonably achieved, with total recovery of radioactivity averaging 80 and 90% of dose following oral and i.v. administration of [¹⁴C]etoricoxib, respectively. Recovery of radioactivity averaged 60% of dose in urine following oral administration, 70% of dose in urine following i.v. administration, and 20% in feces following either route of administration. In addition, the plasma concentration profiles following administration of both i.v. and oral doses indicate that etoricoxib accounted for the majority of the radioactivity in plasma (~75%). The concentrations of etoricoxib and radioactivity declined in parallel, suggesting either reversible metabolism or formation rate-limited metabolic elimination. Since the metabolism of etoricoxib did not appear to be reversible, the parallel decline is likely a result of formation rate-limited metabolic elimination.

The results also showed that radiolabeled etoricoxib was well absorbed following administration of an oral solution in PEG, and the absolute bioavailability averaged 83% (range of 70-99%). This oral bioavailability is somewhat lower than that reported for tablet formulations (~100%), although the T_max and apparent t_1/2 of the two formulations were similar (Agrawal et al., 2001, 2002). However, this result is consistent with that from a previous study, in which the tablet formulation of etoricoxib was found to be slightly more bioavailable than an oral solution in PEG (N. G. B. Agrawal, unpublished). It is assumed that the hepatic extraction is low, because the observed mean plasma clearance following the i.v. dose was low (57 ml/min, blood-to-plasma ratio approaches unity; Kassahun et al., 2001) relative to hepatic blood flow (~1500 ml/min; Davies and Morris, 1993). Assuming that the i.v. dose is only metabolized by the liver, the hepatic extraction is calculated to be ~0.04. With negligible gut first pass metabolism, the oral bioavailability (fraction bioavailable ~0.83) can be readily calculated based on the product of the fraction absorbed (0.86) and the estimated fraction surviving hepatic first pass (0.96).

Analysis of the excreta for etoricoxib and metabolites demonstrated that etoricoxib is metabolized extensively. Less than 1% of an administered i.v. and oral dose was recovered intact in urine over the first 24 h postdose, indicating that renal excretion plays a minimal role in the elimination of etoricoxib. However, the metabolites of etoricoxib are largely excreted renally, because the recovery of radioactivity in urine averaged >60% of dose. In fact, the 6'-carboxylic acid metabolite accounted for the majority of the radioactivity in urine (~80%), over the first 24 h postdose, and small amounts of four other metabolites were found also, including the etoricoxib 1'-N-oxide (~4%), 6'-hydroxymethyl-etoricoxib (~3%), the 1'-N-oxide of 6'-hydroxymethyl-etoricoxib (~2%), and the glucuronide of 6'-hydroxymethyl-etoricoxib (~7%) (Fig. 1). Similarly, the majority of the radioactivity in feces was accounted for by the 6'-carboxylic acid metabolite (65 and 70% following i.v. and oral administration, respectively), and less than 2% was recovered as etoricoxib. Since the recovery of radioactivity in urine and feces was similar following i.v. and oral administration, the presence of radioactivity in feces (20% of dose) is likely attributable to biliary secretion of etoricoxib (minor) and its metabolites.

The observation that hydroxylation of the 6'-methyl moiety predominated over 1'-N-oxidation is in accord with the results of prospective in vitro studies (Chauret et al., 2001; Kassahun et al., 2001), and preclinical metabolism and disposition studies with rats and dogs (R. A. Halpin, unpublished). The in vitro data show that the oxidative metabolism is catalyzed by multiple P450s in the presence of NADPH-fortified human liver microsomes, with CYP3A4 playing an important role (~60%), and the remainder of the activity shared more or less equally among other P450s (e.g., CYP2C9, 2D6, 1A2, and 2C19) (Kassahun et al., 2001). In addition, the oxidation of the 6'-hydroxymethyl to the corresponding 6'-carboxylic acid requires human liver cytosol and NAD⁺ (D. E. Slaughter, unpublished). All of these metabolites have been tested for COX-1 and COX-2 inhibitory activity in vitro. None are active as inhibitors in the COX-1 whole blood assay, and only the 1'-N-oxide and the 6'-carboxylic acid show weak activity in the COX-2 assay (IC₅₀ values of approximately 20 µM versus approximately 1 µM for etoricoxib) (Chauret et al., 2001). Because of the low levels in plasma and weak COX-2 activity, the metabolites of etoricoxib are unlikely to contribute to the inhibition of COX-2 in vivo.

In some respects, the profile of etoricoxib is similar to that of other COX-2-selective agents such as rofecoxib, valdecoxib, and celecoxib, which are well absorbed and metabolized extensively. However, rofecoxib undergoes reductive as well as oxidative metabolism (Halpin et al., 2002), whereas valdecoxib is metabolized by P450s (CYP3A4 and CYP2C9) and undergoes direct glucuronidation (Karim et al., 2001; Yuan et al., 2002). Like etoricoxib, celecoxib is metabolized extensively via methyl hydroxylation, with further oxidation to the corresponding carboxylic acid (Paulson et al., 2000). However, the primary oxidative step is catalyzed largely by a single P450 (CYP2C9), and the drug interaction profile is different from that of etoricoxib (Garnett, 2001; Tang et al., 2000, 2001). In addition, unlike etoricoxib, a large fraction of the oral dose (~58%) is recovered in feces (Paulson et al., 2000).

Based upon the results of this study, it is concluded that etoricoxib is a low clearance compound (relative to hepatic blood flow), with a t_1/2 of 24.8 h. When administered orally, in solution form, etoricoxib is rapidly absorbed (T_max of 1 h), and the extent of absorption is good (~86% as total drug equivalents). The combination of low clearance and good absorption gives rise to an absolute oral bioavailability of ~83% for an oral solution in PEG. In addition, etoricoxib is metabolized extensively (>98%) via 6'-methyl hydroxylation (major) and 1'-N-oxidation. These metabolites are either excreted directly, or are metabolized further to secondary metabolites that are also excreted (urine > feces). Although etoricoxib is cleared by oxidative metabolism, this metabolism is a relatively high K_m process (K_m > 0.1 mM) in vitro (Kassahun et al., 2001) and, as expected, the pharmacokinetics of etoricoxib are linear over the clinically relevant dose range (Agrawal et al., 2001, 2002). Moreover, the involvement of multiple P450s, and the low first pass effect, effectively minimizes the potential for significant drug interactions with potent P450 inhibitors. In agreement with this hypothesis, ketoconazole, a potent inhibitor of CYP3A4, did not elicit a clinically significant effect on the pharmacokinetics of etoricoxib (N. G. B. Agrawal, unpublished).