Abstract
Tirofiban hydrochloride [l-tyrosine-N-(butylsulfonyl)-O-[4-(4-piperidinebutyl)] monohydrochloride, is a potent and specific fibrinogen receptor antagonist. Radiolabeled tirofiban was synthesized with either3H-label incorporated into the phenyl ring of the tyrosinyl residue or 14C-label in the butane sulfonyl moiety. Neither human liver microsomes nor liver slices metabolized [14C]tirofiban. However, male rat liver microsomes converted a limited amount of the substrate to a more polar metabolite (I) and a relatively less polar metabolite (II). The formation of I was sex dependent and resulted from an O-dealkylation reaction catalyzed by CYP3A2. Metabolite II was identified as a 2-piperidone analog of tirofiban. There was no evidence for Phase II biotransformation of tirofiban by microsomes fortified with uridine-5′-diphospho-α-d-glucuronic acid. After a 1 mg/kg i.v. dose of [14C]tirofiban, recoveries of radioactivity in rat urine and bile were 23 and 73%, respectively. Metabolite I and unchanged tirofiban represented 70 and 30% of the urinary radioactivity, respectively. Tirofiban represented >90% of the biliary radioactivity. At least three minor biliary metabolites represented the remainder of the radioactivity. One of them was identified as I. Another was identified as II. When dogs received 1 mg/kg i.v. of [3H]tirofiban, most of the radioactivity was recovered in the feces as unchanged tirofiban. The plasma half-life of tirofiban was short in both rats and dogs, and tirofiban was not concentrated in tissues other than those of the vasculature and excretory organs.
Vascular injury initiates the activation of platelets by local agonists, such as epinephrine, collagen, and thrombin followed by the adhesion and subsequent aggregation of the platelets. This is a consequence of a conformational change in the glycoprotein IIb/IIIa (GP IIb/IIIa) receptors on the platelet surface to allow the binding of fibrinogen and hence the crosslinking of the platelets (Leung et al., 1986;Phillips et al., 1988; Coller, 1990; Smyth et al., 1993). Compounds capable of blocking this binding should prevent thrombus formation and reduce the incidence of ischemic complications after angioplasty and coronary restenosis. A GP IIb/IIIa monoclonal antibody was shown to prevent a platelet agonist from causing platelet aggregation in experimental animals and humans (EPIC Investigators, 1994; Topol, 1995). The GP IIb/IIIa receptor has a high affinity for the tripeptide sequence arginine-glycine-aspartic acid, which is found in several adhesive proteins (Phillips et al., 1988; Frishman et al., 1995). Consequently, analogs and nonpeptide mimetics of this sequence were synthesized as inhibitors of the platelet-fibrinogen interaction (Samanen et al., 1991; Alig et al., 1992; Hartman et al., 1992;Zablocki et al., 1995). Several compounds have undergone development for the treatment of vascular occlusive disorders (Harrington et al., 1994; Kereiakes et al., 1994; Theroux et al., 1994; Anders et al., 1995; Muller et al., 1995).
Tirofiban hydrochloride,l-tyrosine-N-(butylsulfonyl)-O-[4-(4-piperidinylbutyl)] monohydrochloride, also known as L-700,462, and MK-0383, is a potent and specific fibrinogen receptor antagonist, and it has recently been approved (under the name Aggrastat) for use in the prevention of the progression of unstable angina to myocardial infarction. This paper describes experiments with rat, dog, and human tissues on the in vitro metabolism of tirofiban. In addition, experiments are described that pertain to the metabolic disposition of tirofiban, after i.v. dosing, in rats and dogs (the species used in toxicology studies).
Materials and Methods
Chemicals and Reagents.
Radiolabeled tirofiban hydrochloride was synthesized with either3H incorporated into the phenyl ring of the tyrosinyl residue or 14C in the butane sulfonyl moiety (Fig. 1). The radioactive preparations were at least 97 to 98% pure based on HPLC2 and were diluted with carrier drug when necessary. NADPH and uridine-5′-diphospho-α-d-glucuronic acid (UDPGA) were obtained from Sigma (St. Louis, MO). Solvents used for HPLC analysis were of analytical or HPLC grade. Liver microsomes from dexamethasone-induced rats and polyclonal antibodies to rat CYP3A2 and CYP2C11 were obtained from Xenotech (Kansas City, KS) and Gentest (Woburn, MA).
Animals and Treatments.
Male Sprague-Dawley rats and male pure bred beagle dogs were used. Doses were calculated on the basis of the free base. Food was provided to dogs at the end of each day during the studies. Rats were fed ad libitum, and water was available to the animals during the course of all studies.
Rats.
Tirofiban hydrochloride was dissolved in sterile saline. The i.v. doses were given through tail vein injection.
Plasma concentrations.
Rats that were implanted with jugular vein cannulas received doses (5 mg/kg) of tirofiban hydrochloride. Blood samples were taken in the presence of heparin at selected times, and plasma was harvested and stored at −20°C before analysis by radioimmunoassay.
Mass balance.
Urine and feces were collected (0–139 h) from a group of six rats that received 1 mg/kg i.v. [3H]tirofiban hydrochloride. The rats were housed in stainless steel metabolism cages.
Biliary excretion.
Rats were anesthetized with nembutal before undergoing aseptic surgery for bile duct cannulation. Bile ducts were cannulated with polyethylene tubing (PE-10). One group of three rats received 1 mg/kg i.v. [3H]tirofiban hydrochloride. Both bile and urine were collected. Rats, in groups of five, received [3H]tirofiban hydrochloride at 1, 5, and 20 mg/kg i.v. Bile and urine were collected (0–24 h) in containers cooled by dry ice and subsequently analyzed for radioactivity before metabolite isolation or HPLC profiling.
Tissue distribution of radioactivity.
Twelve rats each received a single dose (1 mg/kg i.v.) of [14C]tirofiban hydrochloride. Three animals/time point were sacrificed at 5 and 30 min, and at 2 and 24 h post dose. Blood and specified tissues were collected, and plasma was prepared by centrifugation of the blood. The tissues were rinsed, blotted, and weighed, and either combusted directly or after homogenization. All samples were analyzed for total radioactivity by liquid scintillation counting.
Dogs.
Formulations for i.v. dosing were prepared in sterile saline or 33% dimethyl sulfoxide in sterile saline.
Plasma concentrations.
Dogs in groups of three or four received 0.12, 1.2, and 3.6 mg/kg tirofiban hydrochloride as 120-min i.v. infusions. Blood was collected from the femoral vein, and plasma was harvested and stored at −20°C before analysis of parent tirofiban by radioimmunoassay.
Mass balance.
Four dogs, that were housed individually in stainless steel metabolism cages, received a 1 mg/kg i.v. dose of [3H]tirofiban hydrochloride. Urine and feces were collected (0–72 h).
Instrumental Methods.
Radioactivity was measured in a Tricarb 2500 TR liquid scintillation counter (Packard Instruments, Downers Grove, IL). Samples of urine and HPLC eluates were added directly to polyethylene vials containing 5 ml of Ready Solv (Beckman Instruments, Fullerton, CA). Fecal homogenates were combusted to 3H2O or14CO2 in a Packard Sample Oxidizer 306, and the radioactivity was measured with Monophase or a combination of Permafluor and Carbo-Sorb (Beckman Instruments). The combustion efficiency of the sample oxidizer was determined by comparing the radioactivity recovered from the combustion of samples spiked with a radioactive standard to that obtained by spiking the trapping solution with the same amount of standard. Radioactivity counting time was usually 10 min. An external standard (133Ba) was used to determine efficiency. Mass spectra were recorded on a ZAB HF instrument (VG Scientific, Grinstead, West Sussex, UK) for samples that were prepared in a glycerol or thioglycerol matrix. Combined liquid chromatography tandem mass spectometry was performed with a Finnegan TSQ 7000 mass spectrometer (San Jose, CA) that was operated under positive ion mode conditions. The samples were infused directly into the electrospray ionization source of the mass spectrometer. NMR analyses of metabolites were performed with a Model Unity 400 MHz spectrometer (Varian Associates, Inc., Lexington, MA). Chemical shifts were reported in parts per million relative to tetramethylsilane in CD3OD.
Chromatography of Urinary, Biliary, and Fecal Radioactivity.
Aliquots of urine, rat bile, and acetonitrile extracts of dog fecal homogenates were analyzed by HPLC on a μBondapak C-18 (30 cm × 4.6 mm) column (Waters Corp., Milford, MA) fitted with a Bio-Rad (Philadelphia, PA) or ODS10 guard column (Rainin Instrument 6 Inc., Woburn, MA). In the dose proportionality study in rats, the percentage of radioactivity present as tirofiban in bile was calculated from the ratio of the area representing [3H]tirofiban to that of the total area of the radiochromatogram. The amount of tirofiban excreted was then calculated from the fraction of the dose present in the bile samples. The acetonitrile extracts of dog feces were prepared by mixing aqueous fecal homogenates with five volumes of acetonitrile, centrifuging, and then concentrating the supernatant before injection. Metabolites were eluted with a mixture of acetonitrile and 0.05 M ammonium acetate at pH 4.5. By running a 30-min linear gradient, the acetonitrile fraction was changed from 10 to 40%. An in-line Ramona 5-LS detector (IN/US Systems, Inc., Tampa, FL) was used to monitor the eluates for radioactivity.
In Vitro Metabolism.
Microsomal incubations
[14C]Tirofiban was incubated with liver microsomes from male and female rats (pooled and either noninduced or dexamethasone-induced) and humans (three individuals). The standard incubation mixture contained 0.6 to 0.8 ml of a cocktail (prepared from 10 ml of 0.25 M sodium phosphate, pH 7.4, 2.5 ml of 0.1 M MgCl2, and 7.5 ml of water), 0.1 ml of 1 mM NADPH, 1 mg of microsomal protein, and 20.8 nmol of [14C]tirofiban in 5 μl of 3% methanol in water or 5.2 nmol. Final volume was adjusted to 1 ml with the appropriate amount of buffer so that the final concentration was either 20.8 or 5.2 μM. Incubations were performed at 37°C and 120 rpm for 30 min. At the end of the incubation period, 5 ml of acetone was added and the mixture was vigorously shaken for 6 min. After centrifugation, the supernatants were transferred to a second set of tubes and concentrated under a stream of nitrogen. The residues were dissolved in 150 μl of the HPLC mobile phase. Incubations also were performed with the addition of 50 μl of 3 mM UDPGA and 50 μl of 0.2% Triton X-100 in 0.1 M sodium phosphate buffer but with water replacing the NADPH. Blanks consisted of 30-min incubations with boiled microsomes or incubations minus NADPH. Experiments in which the incubation times were either 20 or 60 min were done with microsomal preparations from nine human livers. Additional experiments were performed with microsomes from noninduced male rats in the presence of either rat CYP3A2 or CYP2C11 antibodies.
Human liver slice incubation.
Samples of liver slices from two individuals were obtained from a tissue bank. [14C]Tirofiban (1 mg) was initially dissolved in 200 μl of 20% dimethyl sulfoxide in water, and 100 μl was added to precision-cut human liver slices (0.6 g) in Williams Medium E in a final volume of 10 ml to give a 113-μM solution of tirofiban. Incubations were carried out in Erlenmeyer flasks at 37°C for 4 to 5 h in an atmosphere of 5% CO2 in air. The mixture was sequentially extracted with three 5-ml aliquots of acetone. The combined extracts were concentrated under N2, and the residues were extracted with three 2-ml aliquots of acetonitrile. The combined extracts were concentrated under N2 and analyzed by HPLC.
Isolation of Metabolites.
Metabolite I
One volume of 0.1 M HClO4 was added to pooled 0- to 24-h rat urine or bile. The acidified samples were extracted with 2.5 volumes of ethyl acetate, followed by evaporation of the ethyl acetate extract to dryness in vacuo. The residue was reconstituted in methanol and subjected to semipreparative HPLC with a Waters uBondapak C-18 column. The charge was eluted with a 31-min linear gradient from 1 to 50% acetonitrile in 0.1% ammonium bicarbonate at a flow rate of 3 ml/min. Metabolite I and unchanged tirofiban were collected and further purified by HPLC on a β-cyclodextrin column (10 mm x 250 mm). The composition of the mobile phase under linear gradient conditions was changed from 10 to 50% acetonitrile in 0.05 M ammonium acetate, pH 4.5, in 20 min at a flow of 3 ml/min. The samples containing I and unchanged tirofiban were mixed with a thioglycerol matrix before high-resolution, fast atom bombardment analysis.
Metabolite II.
Bile was applied in two portions to a 10-g extraction cartridge of XAD-2 resin, followed by a water wash. [14C]Tirofiban and its metabolites were eluted with methanol. The eluates were concentrated and reconstituted with water (1 ml). The aqueous phase was mixed with an equal volume of 0.1 M perchloric acid and then extracted with CHCl3 (5 ml). The CHCl3 phase was removed and evaporated under nitrogen to give a residue, which was dissolved in aqueous methanol (0.25 ml) and subjected to HPLC. The mobile phase contained 0.05 M ammonium acetate, pH 4.5 (A), and acetonitrile (B). The linear solvent gradient was 10% to 40% B in A over 30 min. The radioactive fraction of the eluate that contained II was concentrated to dryness and stored at −20°C before analysis by combined liquid chromatography tandem mass spectrometry. Metabolite II was reconstituted in 300 μl of mobile phase (20% B in A) and 10 μl was injected. Samples for NMR analysis were prepared in a similar manner, except that an extra HPLC purification step was included in which mobile phase 0.1% trifluoroacetic acid in water [A] and acetonitrile [B] and a linear gradient 10 to 60% B in A over 30 min was used.
Hydrolysis and methylation of II.
A sample of II was treated with 0.5 ml of 6 N HCl (containing 0.2% phenol) and heated for 2 h at 110°C under partial vacuum in a digestion tube. The tube contents were cooled and concentrated under N2. The residue was dissolved in a mixture of acetonitrile and water (1/1, v/v) and fractionated by HPLC. The fraction that contained hydrolyzedIIwas methylated by reaction with anhydrous methanolic HCl at 40°C for 2 h. A sample of nonhydrolyzed II was also methylated.
Dog urine and feces.
Dog urine and aqueous fecal homogenates were mixed with acetonitrile (5 volumes) and centrifuged. The supernatant was concentrated and the residue was reconstituted in methanol before purification as described above for rat urine and bile.
Protein Binding.
A Centrifree Micropartition System (Amicon, Inc., Beverly, MA) was used to conduct protein binding studies. [3H]Tirofiban (0.01–25 μg/ml) was incubated in plasma (from rat, dog, monkey, and human) at 37°C for 30 min. Plasma samples were centrifuged at 2500 rpm for 15 min, and the radioactivity in the filtrate was measured. A control experiment was done to confirm the absence of nonspecific binding of radiolabeled material.
Plasma-Red Blood Cell Partition Studies.
Rat, dog, or human blood that had been treated with heparin was spiked with [3H]tirofiban (14–1794 ng/ml). The samples were incubated in a reciprocating water bath for 10, 30, or 60 min. After incubation, aliquots were removed for hematocrit and radiometric assays, while the remainder was centrifuged at 1150g (or 200g for human plasma that contained 20 ng of [3H]tirofiban/ml) so as to separate red blood cell and plasma fractions. Concentrations of [3H]tirofiban in plasma (0.025 ml) were obtained by measuring radioactivity in plasma, whereas tirofiban in the red blood cell fraction (Crbc) was calculated from Crbc = [Cwb −Cp(1 − Hct)]/Hct, whereCwb and Cp are the concentrations of tirofiban in whole blood and plasma, respectively.
Plasma Concentrations of Tirofiban.
A specific and validated radioimmunoassay was used for the determination of tirofiban in plasma (Hand et al., 1994). Plasma concentration data for tirofiban were analyzed by noncompartmental methods (Gibaldi and Perrier, 1982). Area under the curve (AUCs) values were calculated by the trapezoidal method. AUC values from 0 h to infinity were calculated from [AUC]0-∞ = [AUC]0-t+Ct /Ke, whereCt is the last detectable plasma concentration of the drug (at time t) andKe is the terminal slope. Plasma half-lives of tirofiban were determined from T1/2 = 0.693/Ke. The CL of tirofiban was calculated as CL = D/[AUC]0-∞, where D is the dose in mg/kg, and the AUC is expressed as ng · min · ml−1.Vdss was calculated asD[AUMC]/[AUC]2, where AUMC is the area under the first moment curve.
Results
Excretion.
Rats
After a bolus i.v. dose of [3H]tirofiban recoveries of excreted label in feces exceeded those in urine (Table1). The majority of the radioactivity was recovered in the first 24 h after the dose. Mean recoveries of radioactivity in 0- to 24-h bile and urine from bile duct-cannulated rats were 74 and 23%, respectively (Table 1). The amounts of total radioactivity equivalents and unchanged tirofiban excreted in rat bile were approximately proportional to the dose (Table2).
Dogs.
The recoveries of radioactivity (0–72 h) after an i.v. bolus dose are shown in Table 1; the radioactivity was primarily recovered in the feces.
Chromatography of Urine, Bile, and Feces.
HPLC radiochromatograms of rat urine showed the presence of two radioactive species: a polar metabolite (I) (vide infra) and unchanged tirofiban (Fig. 2A). MetaboliteI and tirofiban represented 16 and 6% of the dose, respectively. There was only one radioactive peak in the radiochromatograms of dog urine, and this was identified as unchanged tirofiban (data not shown). HPLC chromatography of bile from rats showed that the majority of the radioactivity was unchanged tirofiban and only 10% represented metabolites. The most polar of the metabolites was metabolite I, which represented 2 to 3% of the dose; the least polar was metabolite II (vide infra), which represented a further 2% (Fig. 2B). Unchanged [3H]tirofiban was represented by the only peak of radioactivity found in the radiochromatogram of dog fecal extract.
Identification of Metabolites.
The structures of the metabolites are shown in Fig.3. In rats, a major polar urinary metabolite was formed by O-dealkylation of tirofiban, whereas the metabolism of tirofiban was negligible in dogs.
Urinary metabolites.
The major rat urinary metabolite was isolated and identified asN-butylsulfonyl-l-tyrosine (I). The identification was made possible by a comparison of the HPLC retention times, and high resolution mass spectra of the metabolite I and the reference compound. Unchanged drug was identified in both rat and dog urine by comparative HPLC, and by high-resolution mass spectrometry (Table3). There was no evidence for tirofiban metabolites in dog urine.
Biliary metabolites.
The bile samples from rats contained one major radioactive compound that was identified as unchanged parent drug by HPLC and high-resolution mass spectrometry (Table 3). At least three minor biliary metabolites accounted for the remainder of the radioactivity. The most polar of the biliary metabolites was identified by HPLC and high-resolution mass spectrometry (Table 3) as I. A biliary metabolite (II) that was less polar than tirofiban was isolated. The NMR proton signals of the piperidone ring ofII were obscured, but those of the alkyl chains and the aromatic ring were identified and were unchanged relative to those of tirofiban. Product ion mass spectra of tirofiban and II were compared. The [M+H]+ ion forII was 455, i.e., 14 amu greater than tirofiban. A fragment group ion of m/z 140 in the spectrum of tirofiban was rationalized as being formed by cleavage at the ether linkage so as to include the piperidine ring together with the attendant alkyl chain. This ion was absent in the mass spectrum of II but instead there was an ion at m/z 154 (probable formula C9H16NO), which indicated that a methylene group (either on the piperidine ring or the alkyl chain linker) had undergone oxidation to a carbonyl group (Fig.4). There was a pronounced forward shift of 5.5 min in the HPLC retention time when II was converted to III by acid hydrolysis. Both II andIII underwent methylation to form IV (Fig.5). The product IV was analyzed by combined liquid chromatography tandem mass spectrometry. It had an [M+H]+ ion of 469 and major fragment ions of m/z 288 and m/z 154 that were consistent with the proposed structure. It was concluded thatII was an acid-labile lactam.
Fecal metabolites.
Radiochromatographic analyses of dog fecal extracts demonstrated that only unchanged tirofiban was present.
In vitro metabolites.
Chromatographic analysis of dried extracts of incubates revealed that neither human liver microsomes nor liver slices metabolized [14C]tirofiban. Male rat liver microsomes produced a single major polar metabolite that was identified asI by comparative HPLC. Approximately 10% and 30% of the substrate was converted to I by microsomal preparations from noninduced and dexamethasone-induced rats (Fig.6), respectively. In the presence of rat CYP3A2 antibody, the formation of I was negligible. However, CYP2C11 antibody had no effect. Female rat liver microsomes did not metabolize [14C]tirofiban (data not shown). There was no evidence for Phase II biotransformation of tirofiban under conditions in which [14C]tirofiban was incubated with rat or human liver microsomes fortified with UDPGA.
Tissue Distribution and Protein Binding.
There was a rapid distribution of radioactivity into tissues and maximum tissue concentrations generally occurred at 5 min post dose. The samples with the highest mean concentrations of radioactivity at 5 min included the liver, kidneys, and plasma. Tissue concentrations decreased rapidly after reaching their maximum concentrations, and at 24 h post dose radioactivity concentrations in all tissues examined were below 0.05 μg equivalents/g (Table4). [3H]Tirofiban was not highly bound to plasma proteins over the concentration range 0.01 to 25 μg/ml. Binding was species dependent and percent bound values in rat, dog, and human plasma were 85, 45, and 64, respectively.
Distribution of [14C]Tirofiban in Blood.
Under in vitro conditions, the ratio of [3H]tirofiban in whole blood to that in plasma for rat and dog ranged from 0.5 to 0.6 over a concentration range of 0.01 to 1.8 μg/ml. In human blood, the ratios were more dependent on concentration and were 0.6, 0.7, and 0.9 for concentration of [14C]tirofiban of 200, 50, and 20 ng/ml. A comparison was made of the label associated with platelet-rich and platelet-poor plasma when whole blood containing 20 ng of [14C]tirofiban was centrifuged either at 200 or 1150g. At the slower and higher centrifugation speeds the ratios of whole blood radioactivity to plasma radioactivity were 0.6 and 0.9, respectively.
Pharmacokinetics.
The value for plasma half-life of tirofiban in rats was 24 ± 7 min, and the mean value for volume of distribution at steady-state was 1.42 ± 0.74 liter/kg. This indicated that tirofiban was not highly concentrated in tissues other than those of the vasculature and excretory organs. Steady-state concentrations of tirofiban in dog plasma were achieved during a 120-min infusion of tirofiban and were proportional to the dose. After cessation of the infusion, plasma concentrations decreased in a biphasic manner (Fig.7). Pharmacokinetic parameters for tirofiban in dogs are shown in Table 5.
Discussion
When rats, with or without biliary cannulas, received [3H]tirofiban hydrochloride i.v., most of the label was eliminated in bile or feces. When dogs received [3H]tirofiban hydrochloride, radioactivity was recovered primarily in the feces. These results indicated that biliary excretion was the major route of elimination for tirofiban. The cumulative biliary excretion of tirofiban was measured in bile-cannulated rats that received i.v. doses of 1, 5, or 20 mg/kg [3H]tirofiban. Both total radioactive tirofiban equivalents and unchanged tirofiban concentrations in bile were approximately proportional to the dose; i.e., there was no evidence for saturation kinetics over the dose range.
The volume of distribution was small in dogs and the plasma half-life was approximately 35 min over a dose range of 0.12 to 3.6 mg/kg. Steady-state plasma concentrations of tirofiban were achieved in dogs during 120-min i.v. infusions of tirofiban. This observation was consistent with the relative short plasma half-life of tirofiban. Tirofiban exhibited linear kinetics in dogs as evidenced by the observations that the plasma concentrations were dose-proportional and that plasma clearance was relatively constant. Although the value forVdSS was larger in rats than in dogs, the short plasma half-life of tirofiban in rats was consistent with the rapid excretion of both tirofiban and its metabolites in bile.
In male rats, two routes of metabolism were identified for tirofiban. One involved O-dealkylation of the ether group (to generate the major urinary metabolite I) and the other resulted in metabolism of the piperidine ring (to form metabolite II, which was excreted in bile). These biotransformations were not observed either in the dog or human liver tissue.
The formation of I in male rats was sex-dependent. Although CYP2C11 is a major male-specific androgen hydroxylase in adult rat liver, rat CYP2C11 antibodies did not inhibit the formation of I in rat liver microsomal preparations. Another adult male-specific liver P-450 enzyme is CYP3A2. The participation of CYP3A2 in the formation ofI was indicated by the fact that the amount of Igenerated in microsomal preparations from rats pretreated with dexamethasone (a known CYP3A2 inducer) was substantially increased. Furthermore, rat CYP3A2 antibodies inhibited the formation ofI (Fig. 6).
The first step in the biosynthesis of II is probably similar to that which occurs in the cytochrome P-450 catalyzedN-dealkylation of alkylamines; i.e., the formation of an iminium ion (Murphy, 1973; Ortiz de Montellano, 1995). Either a cytosolic aldehyde oxidase or a microsomal oxidase may be required for the conversion of an iminium ion to a lactam. (Brandange and Lindblom, 1979; Obach and Vunakis, 1990; Williams et al., 1990). Human CYP2A6 has been implicated in the formation of cotinine (a lactam) from nicotine, and rat CYP2A1 antibodies inhibited this conversion (Nakajima et al., 1996).
For human blood that was incubated with [14C]tirofiban, the values for the ratios of whole blood to plasma radioactivity of 0.9, 0.7, and 0.6 were related inversely to the respective [14C]tirofiban concentrations of 20, 50, and 200 ng/ml. These results suggested that at the lower concentrations there was relatively less [14C]tirofiban in the plasma (resulting in higher ratios). When human plasma was separated by centrifuging blood containing 20 ng of [14C]tirofiban at low speeds, the ratio of label in whole blood to plasma fell to 0.6, i.e., there was an increase in the amount of label in plasma obtained following centrifugation at low speed relative to that obtained at higher speed. Tirofiban, unlike fibrinogen, can bind to fibrinogen receptors that are either in the active or inactive form (Bednar et al., 1995). Thus, it was concluded that this increase in plasma concentration of [14C]tirofiban at low centrifugation speed resulted from the presence of [14C]tirofiban-bearing platelets in the plasma fraction.
Tirofiban was not highly bound to plasma proteins of rats, dogs, and humans. There was rapid distribution of radioactivity in rat tissues minutes after an i.v. dose of [14C]tirofiban hydrochloride. Tissue concentrations decreased rapidly after reaching their maxima. Overall, the pattern of distribution indicated rapid and extensive biliary excretion following the i.v. dose. In general, tissue concentrations were much lower than plasma concentrations with the exception of the liver and kidney, which are the major excretory organs for tirofiban.
The low lipophilicity of the zwitterion of tirofiban limits its transcellular transport and results in a small volume of distribution whereby tirofiban is mainly confined to the vasculature and the organs of excretion. These properties result in a short half-life and are appropriate for a compound that acts on the glycoprotein IIb/IIIa receptor on platelets and which is to be used as an i.v. formulation. Because of the species and sex differences noted in this study, the biotransformation of tirofiban in male rats may not be predictive of the metabolic fate of tirofiban in humans.3
Acknowledgments
We thank S. White, J. Brunner, and K. Michel for their assistance in performing the dog studies. Technical assistance was also provided by L. R. Kauffman, and J. King. Dr. E. Aboaqye, and M. Peterson of Hazleton Wisconsin, Inc. are acknowledged for providing the rat tissue distribution data. Radioimmunoassays were performed by A. Barrish. Some of the mass spectral analyses were provided by Dr. M. R. Davis. We thank Y. Jakubowski, H. Jenkins, T. Marks, R. Chen, Dr. A. Rosegay, and Dr. A. Jones for providing the radiolabeled compounds. The metabolite standard (I) was synthesized by Dr. M. Egbertson. L. Hehn is thanked for providing secretarial assistance. L. Anderson (deceased) is also acknowledged for her contributions to the preparation of the manuscript.
Footnotes
-
Send reprint requests to: Stanley Vickers, Ph.D., WP 75A-203, Drug Metabolism, Merck Research Laboratories, West Point, PA 19486-0004. E-mail: stanley_vickers{at}merck.com
-
↵1 Deceased.
-
A preliminary report was presented at the International Society for the Study of Xenobiotics 7th International Meeting (ISSX), October 20–24, 1996, San Diego, California.
-
↵3 Metabolism of administered tirofiban in human subjects appears to be limited, and unchanged tirofiban is found in urine and feces (Merck and Co. Aggrastat prescribing information).
- Abbreviations used are::
- HPLC
- high-pressure liquid chromatography
- AUC
- area under the curve
- CYP
- cytochrome P-450
- Vdss
- volume of distribution at steady state
- T1/2
- plasma half-life
- UDPGA
- uridine-5′-diphospho-α-d-glucuronic acid
- Received May 7, 1999.
- Accepted August 10, 1999.
- The American Society for Pharmacology and Experimental Therapeutics