Pharmacokinetics and pharmacodynamics of Ro 44–3888 after single ascending oral doses of sibrafiban, an oral platelet aggregation inhibitor, in healthy male volunteers

Abstract

Aims

This study constituted the first administration of the oral platelet inhibitor, sibrafiban, to humans. The aim was to investigate the pharmacokinetics and pharmacodynamics of Ro 44–3888, the active principle of sibrafiban, after single ascending oral doses of sibrafiban. Particular emphasis was placed on intersubject variability of the pharmacokinetic and pharmacodynamic parameters of Ro 44–3888.

Methods

The study consisted of three parts. Part I was an open ascending-dose study to determine target effect ranges of sibrafiban. Part II, a double-blind, placebo-controlled, parallel-group study, addressed the intersubject variability of pharmacokinetic and pharmacodynamic parameters of the active principle at a sibrafiban dose achieving an intermediate effect. Part III was a double-blind, placebo-controlled, ascending-dose design covering the complete plasma concentration vs pharmacodynamic response curve of sibrafiban.

Results

At sibrafiban doses between 5 mg and 12 mg, the pharmacokinetics of free Ro 44–3888 in plasma were linear whereas those of total Ro 44–3888 were non-linear because of the saturable binding to the glycoprotein IIb-IIIa receptor. Saturation of the GP IIb–IIIa receptor was reached at plasma concentrations of 15.9 ng ml⁻¹. At sibrafiban doses up to 2 mg, ADP-induced platelet aggregation was inhibited by 50%, whereas the inhibition of TRAP-induced platelet aggregation was about 20–30%. At the higher doses, ADP-induced platelet aggregation was almost completely inhibited while a clear dose-response could be observed with TRAP-induced inhibition of platelet aggregation at sibrafiban doses of 5 to 12 mg. Ivy bleeding time increased very steeply with dose with a significant prolongation observed at doses of 5 to 7 mg of sibrafiban (5–7 min, >30 min in one case). At a sibrafiban dose of 12 mg, the stopping criterion for dose escalation (prolongation of the Ivy bleeding time >30 min in three out of four subjects per dose group) was reached. The interindividual coefficients of variation of the integrated pharmacokinetic and pharmacodynamic parameters (AUC and AUE) were below 20%, thus lying well within the pre-set level of acceptance.

Conclusions

With a low intersubject variability of its pharmacokinetic and pharmacodynamic parameters, linear pharmacokinetics and pharmacodynamic effects closely related to its plasma concentrations, Ro 44–3888 has good pharmacological prerequisites for a well controllable therapy of secondary prevention of arterial thrombosis in patients with acute coronary syndrome.

Introduction

Thrombotic occlusion occurs frequently as the terminal event in the manifestation of a vascular problem. At high shear rates, platelets play a major role in thrombus formation. Coronary angiography, angioscopy, and pathological studies have confirmed the presence of platelet-rich occlusive coronary artery thrombi in several clinical syndromes, such as unstable angina or myocardial infarction.

Antiplatelet drugs are used in a wide range of disorders, either as sole agents or as adjuncts to other therapies. Aspirin has been shown to be clinically effective in a number of ischaemic conditions and has been in use for many years. The newer agents, ticlopidine and clopidogrel, are also effective and may prove to be superior to aspirin in certain indications. However, their different spectrum of side effects, mainly that of ticlopidine, may eventually limit their widespread use [1].

In the treatment and prevention of coronary artery disease, the efficacy of both aspirin and ticlopidine is limited since their inhibition of platelet activation is incomplete [2, 3]. The final common pathway leading to the formation of the platelet plug is platelet aggregation involving the platelet integrin receptor, glycoprotein (GP) IIb-IIIa, which is able to recognise an arginine-glycine-aspartic acid (RGD) sequence within adhesive molecules, such as fibrinogen and von Willebrand factor [4].

Among the various antiplatelet drugs, the direct GP IIb-IIIa antagonists have been most extensively studied in the acute coronary syndromes. After percutaneous transluminal coronary angioplasty (PTCA), the chimeric Fab fragment of the anti-GP IIb-IIIa monoclonal antibody 7E3 (ReoPro®) administered together with aspirin and heparin resulted in a 35% reduction of clinical endpoints (death, myocardial infarction, early re-occlusion, repeated PTCA) as compared with a combination of aspirin and heparin [5].

Sibrafiban (Z-S-{[1(2{[4-(aminohydroxyiminomethyl) -benzoyl]amino}-1-oxo-propyl)-2-piperidinyl]oxy}-acetic acid ethyl ester) is a double prodrug of the potent, specific, selective, and reversible fibrinogen receptor antagonist Ro 44–3888. After oral administration, the double prodrug sibrafiban (amidoxime ethyl ester) is converted to the pharmacologically inactive prodrug, Ro 48–3656 (amidoxime carboxylic acid), and subsequently to the active principle, Ro 44–3888 (amidine carboxylic acid), as shown in Figure 1.

Figure 1.

Chemical structures of sibrafiban and its metabolites.

Sibrafiban is in development for secondary prevention of arterial thrombosis in patients with acute coronary syndrome (ACS) comprising myocardial infarction and unstable angina. Since intravenous GP IIb-IIIa antagonists cannot be used for secondary prevention of arterial thrombosis, an oral compound offering this therapeutic use is of considerable clinical value.

Methods

Subjects

In total, 49 healthy male volunteers (aged 21 to 41 years) participated in the study. Each subject participated in only one part of the study and received only one dose of sibrafiban. No concomitant medications other than those required to treat potential adverse events were permitted. Written informed consent was obtained from each subject before screening. The study was conducted in full accordance with the principles of the ‘Declaration of Helsinki’ (as amended in Tokyo, Venice and Hong Kong). The study protocol was approved by the Ethics Review Committee (Institutional Review Board) of the study centre prior to study start.

Design

The study consisted of three parts. Part I was an open evaluation of ascending oral doses of sibrafiban. Two subjects each received single oral sibrafiban doses of 1 mg, 2 mg, or 4 mg until ex vivo platelet aggregation induced by adenosine 5′-diphosphate (ADP) was inhibited by more than 50% but less than 100% in both subjects at any given dose. Tablets of sibrafiban or placebo were taken with 200 ml of water immediately after a standard breakfast.

Part II was a double-blind, parallel-group comparison of sibrafiban (5 mg; n=20) and placebo (n=5) to assess intersubject variability of the pharmacokinetic parameters and tolerability of Ro 44–3888. The level of the interindividual coefficient of variation (CV) of AUC for free drug was prospectively set at ≤35%. A higher level of pharmacokinetic variability in healthy male volunteers was considered unacceptable for a compound with a potentially narrow therapeutic window. Part III was a double-blind, placebo-controlled, ascending-dose study. Four subjects each received oral sibrafiban at doses of 7 mg, 10 mg, or 12 mg to cover the full range of the pharmacodynamic effect of sibrafiban. Two subjects per dose group received matching placebo. Dose escalation was stopped once the modified Ivy bleeding time at peak effect exceeded 30 min in three of the four subjects at any given dose.

Eligible subjects were admitted to the study centre on the evening before receiving sibrafiban and were discharged after completing the 72 h assessments. Each subject underwent a full physical examination prior to receiving sibrafiban. In addition, the medical history, body mass, vital signs, and a 12-lead electrocardiogram (ECG) were recorded. The presence of occult blood in faeces was assessed, and routine serological and clinical chemistry tests were conducted. In addition, platelet aggregation and platelet count were determined. Moreover, urine was screened for drugs of abuse. Screening procedures (with the exception of screening for drugs of abuse, serology, medical history, physical examination, and evaluation of occult blood in faeces) were repeated at 72 h. A follow-up physical examination and test for occult blood in faeces were carried out 7 to 10 days after administering sibrafiban.

Pharmacokinetic assessments

Blood samples (4 ml) for the pharmacokinetic assessments were collected at baseline and at specific times up to 72 h after administration of sibrafiban. To determine the binding of Ro 44–3888 to the GP IIb-IIIa receptor, drug concentrations in plasma treated with either K-ethylene-diamine-tetra-acetic acid tripotassium salt (K3-EDTA) or trisodium citrate (total and free concentrations, respectively) were measured by high-performance liquid chromatography (h.p.l.c.).

For four subjects receiving 5 mg sibrafiban (Part II), total plasma concentrations of sibrafiban were analysed by liquid chromatography/tandem mass spectrometry (LC-MS/MS) since the h.p.l.c. method was not specific enough to detect sibrafiban. To prevent degradation of sibrafiban, 1 ml of the plasma was immediately transferred to polypropylene tubes containing dichlorvos (25 μg) in tetrahydrofuran-water (10 μl; 1:4 v/v). All plasma samples were stored at −20° C until required for h.p.l.c. or LC-MS/MS analysis.

Ro 48–3656 and Ro 44–3888 in urine were determined for all subjects in Part II and for the subjects in the 12 mg dose group in Part III. Urine was collected over 72 h after dosing. The total volume and pH of each sample were recorded. An aliquot (approx. 10 ml) was transferred to a polypropylene tube and stored at −20° C until required for h.p.l.c. analysis.

H.p.l.c. method

Ro 44–3888 and Ro 48–3656 in urine and plasma (citrated or K3-EDTA-treated) were determined simultaneously using a sensitive and automated h.p.l.c. method involving column switching, gradient elution, and ultraviolet (u.v.) detection. The limits of quantification of both Ro 44–3888 and Ro 48–3656 in plasma and urine were 1 ng ml⁻¹ and 25 ng ml⁻¹, respectively [6].

LC-MS/MS method

Concentrations of sibrafiban, Ro 48–3656, and Ro 44–3888 in K3-EDTA-treated plasma were determined simultaneously by column switching LC-MS/MS. The limit of quantification for both sibrafiban and Ro 44–3888 was 0.2 ng ml⁻¹, while that for Ro 48–3656 was 0.5 ng ml⁻¹. The method involved protein precipitation (plasma), enrichment of the analytes on a standard-bore trapping column, separation on a narrow-bore analytical column, and detection by ion-spray tandem mass spectrometry [7]. Since the purpose of LC-MS/MS analysis in this study was to determine the plasma levels of sibrafiban, plasma samples were collected only for 12 h after dosing. The plasma concentrations of Ro 48–3656 and Ro 44–3888 measured by LC-MS/MS in these samples were not used for pharmacokinetic evaluations.

Pharmacodynamic assessments

Source of agonists for pharmacodynamic assessments

Ex vivo platelet aggregation was induced by either adenosine 5′-diphosphate (ADP) or thrombin receptor agonist peptide (TRAP). ADP (Grade III) was obtained from Sigma-Aldrich Co Ltd. TRAP (H-SFLLR-NH₂) was produced by F. Hoffmann-La Roche Ltd, Basel, Switzerland, using various coupling procedures and a combination of acid-labile protecting groups. The peptide was purified by h.p.l.c. using a LiChrosorb PR-18 column (Merck, Darmstadt, Germany). The purity of the peptide exceeded 95%, as assessed by thin layer chromatography, analytical h.p.l.c., and mass spectrometry.

Inhibition of ADP- and TRAP-induced platelet aggregation

To assess ex vivo platelet aggregation induced by either ADP or TRAP, blood samples were drawn at baseline and at various time points up to 72 h (Part I) or 48 h (Parts II and III) after dosing with sibrafiban. Blood was collected in Monovet® tubes containing 14.7 mm trisodium citrate (1 ml sodium citrate solution+9 ml blood by volume). The tubes were slowly tilted forwards and backwards (without shaking) and centrifuged at 180 g_max at room temperature for 10 min. The platelet-rich plasma (PRP) was carefully transferred to a labelled plastic tube. The remaining blood was spun at 2000 g_max for 15 min to produce platelet-poor plasma (PPP). The platelet count in the PRP sample was determined for each subject. PRP samples were diluted to a platelet count of approximately 200×10⁹ l⁻¹ using PPP as diluent. Platelet aggregation was assessed with a four-channel PAP-4 Aggregometer (Bio-Data Corporation, Horsham, PA 19040, USA) which was calibrated daily.

ADP_maxvalues were corrected for baseline as follows: ADP_max(post dose)/ADP_max(baseline)×100=relative ADP_max(% baseline). The results were expressed as the percentage inhibition (%) of relative ADP_max (ADP_maxI) as follows: ADP_maxI=100−relative ADP_max. The same equations were applied to TRAP_maxvalues.

Ivy bleeding time

Ivy bleeding time was measured at screen, at baseline, and at various time points up to 72 h (Part I) or 24 h (Parts II and III) after dosing with sibrafiban.

Safety assessments

Adverse events, laboratory variables, vital signs, and ECGs were recorded throughout the study.

Evaluation

Pharmacokinetics

Maximum plasma concentration (C_max), time to reach maximum plasma concentration (t_max), area under the plasma concentration vs time curve from zero to infinity (AUC), and apparent clearance (CL/F) were calculated for free and total Ro 44–3888 in citrate-treated and K3-EDTA-treated plasma, respectively.

Pharmacokinetic parameters were either read directly or derived from the plasma concentration vs time data using noncompartmental methods [8]. The dose of Ro 44–3888 used in all calculations was 0.8952 mg (equivalent to 1 mg sibrafiban). All calculations were conducted using SAS 6.08 [9].

Urinary excretion of Ro 48–3656 and Ro 44–3888 was given as the percentage of the administered dose (expressed as Ro 44–3888).

Estimation of binding parameters

Ro 44–3888 does not bind to proteins other than the GP IIb-IIIa receptor [10]. For the estimation of the binding parameters, the values for free and total plasma concentrations of Ro 44–3888 for all subjects participating in the study were pooled. The analysis was carried out with NONMEM IV [11] running on a micro VAX under VMS.

The binding parameters of Ro 44–3888 were estimated using the following equation: C_b=B_maxC_u/(K_d+C_u), where C_b=bound plasma concentration of Ro 44–3888, C_u=free plasma concentration of Ro 44–3888, B_max=concentration at which maximum binding occurred, and K_d=the dissociation constant.

Pharmacodynamics

The pharmacodynamic effect of Ro 44–3888 was assessed on the basis of inhibition of ex vivo ADP-or TRAP-induced platelet aggregation as well as the modified Ivy bleeding time. The area under the effect curve from zero to 24 h for ADP-induced platelet aggregation (AUE-ADP_0-24h), maximum inhibition of ADP-induced platelet aggregation (E_max), and the time at maximum inhibition of ADP-induced platelet aggregation (t_maxE) were determined.

Statistical analyses

For Part I, no statistical analysis was carried out because of the limited sample size (two subjects per dose group). For Parts II and III, descriptive statistics (mean, coefficient of variation, number of observations, minimum, maximum) were calculated for all pharmacokinetic parameters. Dose proportionality for AUC and C_max values (both dose-normalised) was assessed over a sibrafiban dose range of 5-12 mg using an exploratory Type III analysis of variance (ANOVA) with the classification variable dose group. All calculations were conducted using SAS 6.08 [9].

Results

Clinical events (safety)

Sibrafiban and its metabolites were generally well tolerated at all doses. No serious adverse events were reported, and there was no clinically significant effect on routine laboratory parameters or vital signs. All subjects completed the study.

The most frequently reported adverse event was bruising, but the overall incidence in subjects treated with sibrafiban and those receiving placebo was similar. Some adverse events related to bleeding (nosebleeds, blood in saliva) occurred but none of these was serious.

Pharmacokinetics

Part I

Since the plasma levels of Ro 44–3888 after 1 mg, 2 mg, and 4 mg sibrafiban were low in most cases and the results of two subjects had to be excluded for technical reasons, only a limited pharmacokinetic evaluation was possible.

Part II

The main objective of Part II of this study was to assess the intersubject variability of the pharmacokinetic parameters of Ro 44–3888. The intersubject variability of the AUC derived from the free and total concentrations of Ro 44–3888 was 17% and was thus well below the pre-set limit of 35%. The main pharmacokinetic parameters and corresponding CVs for free and total plasma concentrations of Ro 44–3888 are shown in Tables 1 and 2, respectively.

Table 1.

Pharmacokinetic parameters of free Ro 44–3888 in plasma after 5 mg, 7 mg, 10 mg, and 12 mg sibrafiban.

Table 2.

Pharmacokinetic parameters of total Ro 44–3888 in plasma after 5 mg, 7 mg, 10 mg, and 12 mg sibrafiban.

It was of interest to determine whether unchanged sibrafiban was present in human plasma. Sibrafiban concentrations were low (C_max about 3-4 ng ml⁻¹ after a 5 mg dose of sibrafiban), reaching the limit of quantification approximately 3 h after drug intake. The rapid cleavage of the double prodrug was paralleled by a steep rise in plasma concentrations of the prodrug and the active principle.

Part III

Figure 2 shows the plasma concentration vs time profiles for free and total Ro 48–3656 and Ro 44–3888 after administration of the highest dose of sibrafiban (12 mg). Tables 1 and 2 show the main pharmacokinetic parameters derived from free and total plasma concentrations of Ro 44–3888 after 5 mg, 7 mg, 10 mg, and 12 mg of sibrafiban. C_max and AUC of the free concentrations increased in a dose-proportional manner, indicating that saturation of the activating enzymes did not occur at the dose range studied. However, total concentrations of Ro 44–3888 (Table 2) did not increase in a dose-proportional manner as a result of the saturable binding of Ro 44–3888 to the GP IIb-IIIa receptor. This was also reflected in CL/F derived from the total plasma concentrations of Ro 44–3888 (Table 2). After 5 mg of sibrafiban, CL/F was markedly lower than after the 12 mg dose because most of the active principle after 5 mg of sibrafiban was bound to the receptor rendering less free drug available for elimination. In contrast, CL/F of free Ro 44–3888 was not affected by the dose (Table 1). To confirm these findings, dose proportionality and the binding to the GP IIb-IIIa receptor were studied.

Figure 2.

Plasma concentrations vs time profiles for free and total Ro 48–3656 and Ro 44–3888 after 12 mg sibrafiban. Data are mean plasma concentrations (ng ml⁻¹). ▪ total Ro 44–3888, in K3-EDTA-treated plasma, □ free Ro 44–3888, in citrate-treated plasma, • total Ro 48–3656, in K3-EDTA-treated plasma, ○ free Ro 48–3656, in citrate-treated plasma.

Dose proportionality and binding to the GP IIb-IIIa receptor

Figure 3 shows the dose-normalised AUC vs the dose for free and total plasma concentrations of Ro 44–3888. While there was a proportional relationship between the dose-normalised AUC and the dose for free Ro 44–3888, the dose-normalised AUC of total Ro 44–3888 at the 5 mg dose was more than proportionally higher than the AUC associated with 7 mg, 10 mg, or 12 mg sibrafiban. Nevertheless, it is apparent that the kinetics of total Ro 44–3888 turned linear again once saturation of the GP IIb-IIIa receptor had been reached. These findings were confirmed by the statistical analysis using ANOVA where P values of 0.9405 and 0.0001 for free and total Ro 44–3888, respectively, were obtained.

Figure 3.

AUC/dose vs dose for free and total Ro 44–3888 in plasma. ▪ AUC/dose derived from concentrations of total Ro 44–3888, in K3-EDTA-treated plasma, □ AUC/dose derived from concentrations of free Ro 44–3888, in citrate-treated plasma.

Figure 4 shows the relationship between bound and free Ro 44–3888 in plasma. The E_max-like shape of the curve indicates that saturation of the GP IIb-IIIa receptor was reached. The binding parameters estimated by NONMEM were 3.61 ng ml⁻¹ for K_d and 15.9 ng ml⁻¹ for B_max.

Figure 4.

Bound vs free concentrations of Ro 44–3888 in plasma (all subjects). ♦ 1 mg, ◊ 2 mg, ▴ 4 mg, ▪ 5 mg, ○ 7 mg, • 10 mg, □ 12 mg sibrafiban.

Urinary excretion

The total average excretion of drug-related material (Ro 48–3656 and Ro 44–3888) in urine was 44.4% (CV=18%; range 18.3% to 57.8%). Overall, 29.3% (CV=19.6%; range 13.5% to 41.1%) of the administered dose was present in urine as Ro 44–3888, indicating that the bioavailability of the active principle after oral administration of sibrafiban was at least 14%.

Pharmacodynamics

Inhibition of ADP-induced platelet aggregation

At 4 h to 10 h after administration of 1 and 2 mg sibrafiban, ADP-induced platelet aggregation was inhibited by 50%. At a sibrafiban dose of 4 mg, ADP-induced aggregation was almost completely inhibited (90 to 100%).

At a sibrafiban dose of 5 mg, the intersubject variability of the pharmacodynamic parameters was as low as that of the pharmacokinetic parameters, with the CV of AUE(0,24h) amounting to 15.6% (Table 3). The time course of the pharmacodynamic effect correlated with that of the plasma concentrations of the active principle. The peak pharmacodynamic effect (85% inhibition of ADP-induced aggregation) occurred at approximately 6 h (range 4 to 10 h) and was barely apparent 24 h after dosing (Figure 5).

Table 3.

Descriptive statistics of ADP-induced platelet aggregation after 5 mg sibrafiban (n=20).

Figure 5.

Time course of inhibition of ADP–induced platelet aggregation after 5 mg, 7 mg, 10 mg, and 12 mg sibrafiban or placebo. ▪ 5 mg, ○ 7 mg, • 10 mg, □ 12 mg sibrafiban, ▵ placebo.

At higher doses of sibrafiban (7, 10 and 12 mg), ADP-induced platelet aggregation was completely inhibited (Figure 5). After 24 h, a residual inhibition of approximately 30% was observed but interindividual variability was relatively high at this time point. Based on the AUE values (data not shown), the effect of sibrafiban on platelet aggregation was more pronounced at 10 mg and 12 mg than at 7 mg. However, since even 7 mg sibrafiban achieved complete inhibition, ADP-induced aggregation proved unsuitable to assess the pharmacodynamic effect of sibrafiban in the higher dose range.

Inhibition of TRAP-induced platelet aggregation

After 1 and 2 mg sibrafiban, inhibition of TRAP-induced platelet aggregation was between 20% and 30% at about 6 h after dosing. Because there were only two subjects per dose group and the intersubject variability was fairly high at these low doses, it remains uncertain whether the measured inhibition of TRAP-induced platelet aggregation was a true drug effect or simply reflected the baseline variability of this parameter.

After doses of 4 and 5 mg sibrafiban, inhibition of TRAP-induced platelet aggregation followed a time course similar to that of ADP-induced platelet aggregation, although peak inhibition amounted to only about 35% (Figure 6). Thus, the effect of sibrafiban on platelet aggregation induced by the weak agonist, ADP, was greater than that induced by the strong agonist, TRAP.

Figure 6.

Time course of inhibition of TRAP–induced platelet aggregation after 5 mg, 7 mg, 10 mg, and 12 mg sibrafiban or placebo. ▪ 5 mg, ○ 7 mg, • 10 mg, □ 12 mg sibrafiban, ▵ placebo.

A clear dose-response relationship was observed with TRAP-induced platelet aggregation after 5 mg, 7 mg, and 10 mg sibrafiban, while the difference between the 10 and 12 mg doses was relatively small (Figure 6). The peak pharmacodynamic effect was observed approximately 6 h after dosing. At all doses, baseline levels of TRAP-induced platelet aggregation were restored 24 h after dosing. Clearly, the dose-response relationship obtained with TRAP-induced platelet aggregation rendered TRAP a more useful agonist to monitor the pharmacodynamic effect of sibrafiban in the higher dose range.

Ivy bleeding time

The time course of the prolongation of the Ivy bleeding time followed closely the plasma concentrations of Ro 44–3888 with the peak effect occurring approximately 6 h after dosing (Figure 7). Ivy bleeding time returned to baseline levels within 24 h after administration of all doses.

Figure 7.

Time course of median modified Ivy bleeding time after 5 mg sibrafiban or placebo. ▪ sibrafiban 5 mg, ▵ placebo.

Ivy bleeding time increased very steeply with the dose. At sibrafiban doses up to 4 mg, Ivy bleeding time was similar to that after placebo (approx. 3 min). At 5 mg sibrafiban, Ivy bleeding time began to increase, reaching prolongation times of 5 to 6 min. In one subject, Ivy bleeding time was prolonged by approximately 30 min.

After the 7 mg dose sibrafiban, Ivy bleeding time was similar to that after the 5 mg dose, but there was no maximum prolongation up to 30 min. Prolongations of Ivy bleeding time ≥30 min were treated as ‘prolongation to infinity’, and measurements were stopped thereafter.

At the higher doses of sibrafiban (10 and 12 mg), Ivy bleeding time was clearly prolonged over that measured after placebo. After the 10 mg dose, Ivy bleeding time exceeded 30 min in two of the four subjects, while the remaining two subjects had prolongations of 6 to 8 min. After 12 mg, the values at peak effect exceeded 30 min in three of the four subjects in this group. For the fourth subject in this group, a maximum bleeding time of 5 min was measured. Although interindividual variation of the bleeding time was considerable, the pre-established criterion for terminating dose escalation (bleeding time exceeding 30 min in three of the four subjects) was clearly met at 12 mg. For this reason, no higher doses of sibrafiban were given.

Discussion

Pharmacokinetics

A low intersubject variability and linear pharmacokinetics resulting in predictable plasma concentrations are important prerequisites for a compound with a potentially narrow therapeutic window. For a double prodrug, such as sibrafiban, the activating enzyme systems are another potential source of variability besides the absorption and elimination steps.

Activation of the double prodrug, sibrafiban (amidoxime ethyl ester), involves two steps, i.e., ester cleavage to the pharmacologically inactive prodrug, Ro 48–3656 (amidoxime carboxylic acid), followed by reduction of Ro 48–3656 to the active principle, Ro 44–3888 (amidine carboxylic acid). While the first activation step is catalysed by ubiquitous esterases, the second is catalysed by an enzyme system that is not yet fully characterised but is known not to involve P-450 isoenzymes. Studies in rat and human liver tissue indicated that the amidoxime reductase (AOR) activity involved in the reduction of Ro 48–3656 to Ro 44–8888 is located in the outer membrane of liver mitochondria. It is assumed that this mitochondrial AOR activity is similar to the microsomal N-hydroxyamine reductase system involved in the reduction of amidoximes, N-hydroxyguanidines, and N-hydroxylamines [12].

In this study in healthy volunteers, enzymatic conversion yielded consistent plasma levels of the active principle at all doses of sibrafiban. The pharmacokinetics of free Ro 44–3888 in plasma were linear in the dose range of sibrafiban studied. The conversion of sibrafiban to the active principle was not saturated at the doses used, and only very low concentrations of unchanged sibrafiban were found in human plasma. Bioavailability does not appear to be of concern since the full pharmacodynamic effect of Ro 44–3888 was achieved with 12 mg sibrafiban.

Moreover, intersubject variability of all pharmacokinetic parameters of Ro 44–3888 was very low at all doses, indicating that absorption, metabolism, and elimination of sibrafiban are predictable in healthy male volunteers. Nonetheless, it remains to be shown whether this predictability will also hold in the target population of post-ACS patients.

In excretion balance studies in dogs [10], approximately 82% of the total radioactivity was recovered in urine after intravenous administration of radiolabelled sibrafiban. If this finding can be confirmed in humans, renal function will have a large impact on the plasma levels of Ro 44–3888 and may have to be taken into account when establishing the dose regimens to reduce intersubject variability. In addition, the contribution of enzymatic activation of sibrafiban as well as other factors, e.g., gender and age, to the overall intersubject variability warrant further study by a covariate analysis in the target population.

Pharmacodynamics

At 5 mg sibrafiban, the low intersubject variability of pharmacokinetic parameters was paralleled by the data for inhibition of ADP-induced platelet aggregation (AUE-ADP(0,24h). The effect of sibrafiban on ex vivo ADP- or TRAP-induced platelet aggregation was clearly dose-related, although ADP as the aggregation agonist was useful only at lower doses of sibrafiban (1 to 5 mg). At higher doses (7 to 12 mg), the effect was best measured using TRAP since ADP-induced aggregation was completely inhibited at these doses.

Modelling of the plasma concentrations of Ro 44–3888 as well as the relationship of inhibition of ADP-induced platelet aggregation and plasma concentrations of Ro 44–3888 will be the subject of a future publication. At present, it is concluded that the pharmacodynamic effects were closely related to the plasma levels of Ro 44–3888. Since the inhibitory effect of Ro 44–3888 on platelet aggregation induced by either ADP and TRAP declined rapidly after 12 h and baseline levels were restored by 24 h after dosing, a twice-daily dosing regimen may be required to maintain the pharmacodynamic effect for 24 h. Moreover, the pharmacokinetics of Ro 44–3888 as well as the concentration-effect relationship may be different in the target patient population. Thus, the optimal dose for Phase III clinical trials will be more reliably determined on the basis of Phase II studies in patients.

Ivy bleeding time exhibited a steep dose-response curve, possibly indicating a ‘threshold concentration’ for the incidence of minor bleedings. Although it is reasonable to assume a correlation between the prolongation of Ivy bleeding time and mucosal bleedings, no clear relationship between bleeding time prolongation and frequency of minor bleedings was observed in this study. The incidences of adverse events related to bleeding in the active treatment groups and the placebo group were similar. Clearly, more studies, particularly in patients, are warranted to evaluate the risk of bleeding associated with prolonged Ivy bleeding time.