Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Platelet activation is increased in patients undergoing coronary thrombolysis and may delay reperfusion and induce acute reocclusion (Fitzgerald et al., 1991). The increase in platelet activation has a number of consequences. There is an enhanced formation of TXA₂ (Fitzgerald et al., 1988; Kerins et al., 1989), the major cyclooxygenase product of arachidonic acid in platelets (Hamberg et al., 1975). TXA₂ is a potent platelet activator, acting to amplify the response to weak agonists such as ADP and low concentrations of thrombin and epinephrine (Hamberg et al., 1975). That TXA₂ is functionally important is demonstrated by the response to TXA₂ receptor antagonists in experimental models of thrombolysis, where they accelerate reperfusion and inhibit reocclusion (Fitzgerald et al., 1989; Golino et al., 1988). Moreover, aspirin reduces mortality in patients with acute myocardial infarction treated with streptokinase (ISIS-2, 1988).

In addition to generating TXA₂, the activation of platelets results in their aggregation. Platelet aggregation reflects the activation of an adhesion receptor, GP IIb/IIIa, to bind fibrinogen (Coller, 1990). Antagonists of GP IIb/IIIa have profound effects in experimental models of thrombolysis, enhancing the response to plasminogen activators and abolishing reocclusion (Fitzgerald and FitzGerald, 1989; Nicolini et al., 1994; Gold et al., 1988; Mickelson et al., 1990).

A possible mediator of the increased platelet activity and aggregation during coronary thrombolysis is thrombin. There is strong biochemical evidence, based on measurement of thrombin-antithrombin complexes, for the de novo generation of thrombin during coronary thrombolysis (Merlini et al., 1995; Galvani et al., 1994). There is also evidence of increased thrombin activity during coronary thrombolysis (Fitzgerald and FitzGerald, 1989; Merlini et al., 1995; Galvani et al., 1994; Eisenberg et al., 1987; Jang et al., 1990; Yao et al., 1992). It is unclear, however, whether thrombin is responsible for the increased platelet activity and thromboxane formation that occurs during coronary thrombolysis. In this study, we examine the response to a novel and specific thrombin inhibitor, Ro 46-6240 (Rudd et al., 1992), in a model of coronary thrombolysis that is associated with marked platelet activation and wherein the functional response is highly platelet-dependent (Fitzgerald et al., 1989; Fitzgerald and FitzGerald, 1989).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. ADP and thrombin were obtained from Sigma Chemical Co. (St. Louis, MO). Human recombinant t-PA, Ro 46-6240 (napsagatran, N-[N4-[[(S)-1-amidino-3-piperidinyl]methyl]-N2-(2-naphthalenesulfonyl)-L-asparaginyl]-N-cyclopropylglycine; Carteaux et al., 1995) and Ro 44-9883 (lamifiban, [[1-(N-(p-amidinobenzoyl)-L-tyrosyl]-4-piperidinyl]oxy]acetic acid; Carteaux et al., 1993) were kind gifts from Dr. Sebastien Roux (F. Hoffmann-La Roche Ltd., Basel, Switzerland).

Animal studies. All animal studies were reviewed and approved by the Animal Committee at University College Dublin, Ireland. Mongrel dogs, weighing 13 to 30 kg, were studied using the previously described protocol (Fitzgerald et al., 1989). Briefly, after anesthesia with pentobarbitone (30 mg/kg), the animal was intubated and ventilated with a Harvard respirator (Harvard Apparatus Co., Natick, MA). Through a thoracotomy, the circumflex coronary artery was dissected free, and all branches prior to the first obtuse marginal branch of the circumflex were ligated. A 4 to 5-mm needle electrode was inserted into the lumen, and a Doppler flow probe (Crystal Biotech, Holliston, MA) was placed proximal to the electrode. The wires from the electrode and the flow probe were brought to the surface and placed in a subcutaneous pouch. The dog was allowed to recover for a period of 5 to 7 days, by which time biochemical indices of platelet activation had returned to base line (Fitzgerald et al., 1989). Postoperatively, heparin (200 IU/kg s.c.) was administered every 8 hr for the first 48 hr.

Experimental procedures. On the day of the study, the dogs were sedated with acepromazine (91 mg/kg) and morphine sulphate (1-2 mg/kg), and the wires were retrieved via a small skin incision. The flow probe was connected to a directional pulsed Doppler flowmeter (545C-4 Bioengineering, University of Iowa City, IA). Coronary blood flow velocity was recorded throughout the experiment with an Apple MacLab multichannel recorder. A 200-µA current was passed through the needle electrode to induce endothelial injury. This resulted in coronary thrombosis and subsequent occlusion, detected as abolition of the signal from the Doppler flow probe. The thrombotic occlusion occurred in 1 to 2 hr, and the current was discontinued 30 min later. Two hours after complete coronary occlusion, t-PA 10 µg/kg/min was administered as an infusion through a peripheral vein; the infusion was continued until 10 min after reperfusion. This results, in about 60 min, in reperfusion that is always complicated by acute reocclusion in control experiments (Fitzgerald et al., 1989; Fitzgerald and FitzGerald, 1989). Ro 46-6240 and Ro 44-9883 were administered either alone or in combination as a bolus 30 min before t-PA. This was immediately followed by a continuous infusion that was continued until 2 hr after reperfusion. Dogs were randomly assigned to receive one of the following drug regimens:

a)	Vehicle only		(n = 10)
b)	Ro 46-6240	25 µg/kg  + 5 µg/kg/min	(n = 5)
c)	Ro 46-6240	50 µg/kg  + 10 µg/kg/min	(n = 4)
d)	Ro 46-6240	100 µg/kg  + 20 µg/kg/min	(n = 3)
e)	Ro 46-6240	200 µg/kg  + 40 µg/kg/min	(n = 5)
f)	Ro 44-9883	312 µg/kg  + 2 µg/kg/min	(n = 4)
g)	Ro 46-6240	25 µg/kg  + 5 µg/kg/min
		+
	Ro 44-9883	312 µg/kg  + 2 µg/kg/min	(n = 7)

Thromboxane biosynthesis. Urine was collected throughout in hourly aliquots by catheter for measurement of 2,3-dinor-TXB₂, the major enzymatic metabolite of TXA₂ in vivo. 2,3-Dinor-TXB₂ was determined by isotope dilution using a deuterated internal standard with quantitation by negative ion-chemical ionization, gas chromatography-mass spectrometry, as previously described (Fitzgerald et al., 1989; Fitzgerald and FitzGerald, 1989).

Coagulation assays. Plasma fibrinogen concentration was determined by the Clauss method as thrombin clottable protein using a fibrometer (BBL Fibrosystem, Becton Dickinson, Cockeysville, MD). The coefficient of variation between measurements on the same sample was 3%, and each sample was run in triplicate. The aPTT was determined in citrated plasma as the clotting time after the addition of Actin FS and CaCl₂ 25 mM. The PT was determined in citrated plasma as the clotting time after the addition of Dade thromboplastin C (Baxter, Miami, FL).

Platelet aggregation studies. Ex vivo platelet aggregation was analyzed by light transmission (Bio Data PAP-4, Horsham, PA) before, 15 min after and 1 hr after the administration of drugs (Fitzgerald et al., 1989; Fitzgerald and FitzGerald, 1989). Blood samples were collected into plastic tubes containing 3.8% solution of sodium citrate (9 vol blood to 1 vol sodium citrate). Platelet-rich plasma was obtained by centrifuging blood samples at 900 × g for 45 sec, and platelet-poor plasma was obtained by centrifuging the residual blood at 900 × g for 10 min. The platelet agonists used were ADP at a concentration of 1 to 5 µM and thrombin at a final concentration of 0.1 to 0.5 U/ml. Agonists were added in a volume of 50 µl or less to 500-µl aliquots of platelet-rich plasma. The degree of platelet aggregation was reported as a percentage of maximal increase in light transmission in platelet-rich plasma as compared with platelet-poor plasma.

Bleeding rate analysis. The bleeding rate was assessed from the incision on the chest wall used to retrieve the flow probe and the electrode terminal. The wound was standardized so that it was half an inch in length and involved the full thickness of the subcutaneous tissues. Bleeding rate was determined by packing the incision with gauze and determining the increase in weight of the gauze hourly throughout the experiment (Fitzgerald et al., 1991).

Plasma drug measurements. Plasma concentrations of Ro 46-6240 were determined functionally using a bioassay (Tschopp TB, F. Hoffmann-la Roche, Basel, Switzerland). Two hundred microliters of citrated plasma was mixed with 400 µl of acetone and incubated for 5 min at room temperature. After centrifugation at 12,000 × g for 2 min at room temperature, 500 µl of the supernatant was evaporated under nitrogen and resuspended in 167 µl of 0.05 M Tris, 0.1 M NaCl, 0.1% (w/v) PEG 6000 and 0.02% (v/v) Tween 80, pH 7.8. The thrombin inhibitory activity was measured with a chromogenic substrate kinetic assay run on a Cobas Bio spectrophotometric centrifugal analyzer (F. Hoffmann-la Roche). The final concentration of human thrombin was 12 nM, and that of the substrate (methoxysulfonyl-D-Leu-Gly-Arg-paranitroanilide) was 100 µM. The concentration of Ro 46-6240 in the samples was calculated from a standard curve constructed with canine plasma containing known amounts of drug. The detection limit of the assay was 20 ng/ml.

Statistical analysis. The data were analyzed by Kruskal-Wallis one-way analysis of variance with a subsequent Mann-Whitney test for comparison between groups. These analyses make no assumptions as to the distribution of the data. Chi-square analysis was also used where appropriate. Values are expressed as mean ± S.E.M. P < .05 was considered statistically significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Occlusive thrombi developed in all dogs within 82 ± 10 min after application of the current. After the administration of t-PA, reperfusion occurred in all control animals, with a mean time to reperfusion of 52 ± 5 min. Complete reocclusion occurred in every case at 38 ± 8 min (n = 10), and no further reperfusion was detected. Ro 46-6240, the selective thrombin inhibitor, induced a dose-dependent prolongation of the aPTT and PT that was evident 15 min after administration of the bolus of drug (table 1). There was no effect on plasma fibrinogen before or after the infusion of t-PA in control animals or in those receiving Ro 46-6240 (table 2). Ro 46-6240 also reduced the time to reperfusion in a dose-dependent manner (P < .01) and delayed or prevented reocclusion (table 3). At the lowest dose of Ro 46-6240 (5 µg/kg/min), complete reocclusion was not prevented, but cyclical episodes of reocclusion/reperfusion were seen for up to 2 hr after reperfusion. At 20 µg/kg/min and 40 µg/kg/min, reocclusion was prevented in all but one experiment, where reocclusion occurred at 115 min despite prolongation of the aPTT and high plasma drug levels.

Ro 44-9883, the antagonist of the platelet GP IIb/IIIa, at 2 µg/kg/min did not alter the time to reperfusion or the rate of or time to reocclusion when it was administered alone. However, when Ro 44-9883 was combined with the lowest dose of Ro 46-6240 (5 µg/kg/min), which alone had little effect, the time to reperfusion was reduced, and reocclusion was prevented in every case (table 3). It is worth noting that the cycles of reocclusion seen after treatment with Ro 46-6240 5 µg/kg/min were abolished by Ro 44-9883, which indicates that they reflected activation of the platelet GP IIb/IIIa.

Platelet aggregation studies. At base line, the platelets of all animals responded to ADP with an EC₅₀ of 2 ± 1 µM and to thrombin with an EC₅₀ of 0.3 ± 0.1 U/ml. Ro 46-6240 abolished thrombin-induced platelet aggregation at even the lowest dose of 5 µg/kg/min. Suppression of aggregation was similar at 15 min after administration of Ro 46-6240 and 1 hr after reperfusion. In contrast, ADP-induced platelet aggregation was not influenced by Ro 46-6240 at any dose used, a result that confirms its specific antithrombin activity. When the GP IIb/IIIa antagonist Ro 44-9883 was added, ADP-induced platelet aggregation was also prevented (fig. 1).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Platelet aggregation to thrombin 0.1 U/ml and ADP 2 µM before and after treatment with the thrombin inhibitor Ro 46-6240 alone (n = 4) and combined with Ro 44-9883 (n = 4), a GP IIb/IIIa antagonist.

Urinary 2,3-dinor TXB₂. Urinary excretion of 2,3-dinor TXB₂, the major metabolite of TXA₂ in dogs and an index of in vivo TXA₂ formation, increased during the induction of thrombolysis and after reperfusion (fig. 2). In control animals, urinary excretion of 2,3-dinor TXB₂ increased from 3 ± 1 to 33 ± 5 and 37 ± 9 ng/mg creatinine at 30 min and 2 hr after reperfusion, respectively (n = 7). Excretion of the metabolite was suppressed by Ro 46-6240 in a dose-dependent manner (P < 0.01). However, at the lowest dose of Ro 46-6240 (5 µg/kg/min), urinary 2,3-dinor-TXB₂ was unaltered (6 ± 1 ng/mg creatinine at base line rose to 45 ± 4 ng/mg creatinine at 2 hr of reperfusion; n = 5). Similarly, Ro 44-9883 at a dose of 2 µg/kg/min did not alter excretion of 2,3-dinor-TXB₂ (3 ± 3 ng/mg creatinine at base line rose to 51 ± 6 ng/mg creatinine at 2 hr of reperfusion; n = 3). However, combining Ro 46-6240 and Ro 44-9883 abolished the increase in this metabalite (4 ± 1 ng/mg creatinine at base line rose to 5 ± 1 ng/mg creatinine at 2 hr of reperfusion, n = 4).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Urinary excretion of 2,3-dinor-TXB2 after coronary thrombosis but before t-PA (base), immediately at reperfusion and at 2 hr after reperfusion. Ro 46-6240 dose-dependently suppressed the excretion of 2,3-dinor-TXB2. At the highest dose, urinary 2,3-dinor-TXB2 was 5.6 ± 1 ng/mg creatinine vs. 37 ± 9 ng/mg creatinine in controls. A similar effect was seen combining the lowest dose of Ro 46-6240 and a subthreshold dose of Ro 44-9883.

Bleeding rate. Bleeding from the skin incision was minimal before administration of thrombolytic therapy. The infusion of t-PA resulted in a small increase in bleeding in control animals. At the lowest concentrations of Ro46-6240 (5 and 10 µg/kg/min), there was no increase in bleeding compared with control. However, there was a 3 ± 2.4-fold increase in bleeding at 20 µg/kg/min, and a 6.4 ± 4-fold increase in bleeding at 40 µg/kg/min, compared with control (fig. 3). The increase in bleeding persisted over the period of observation after reperfusion (2 hr) despite the discontinuation of t-PA. The combination of the lowest dose of Ro 46-6240 with Ro 44-9883 increased the bleeding rate by 5.4 ± 2-fold compared with control (fig. 4).

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Bleeding rate after reperfusion in animals treated with increasing doses of the thrombin inhibitor Ro 46-6240. * P < 0.05 vs. control, ** P < 0.01 vs. control.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Bleeding rate after reperfusion in animals treated with the GP IIb/IIIa antagonist Ro 44-9883 2 µg/kg/min alone and combined with the thrombin inhibitor Ro 46-6240 5 µg/kg/min.

Plasma drug levels. The mean plasma t-PA level at reperfusion was 660 ± 100 ng/ml; it returned to base-line levels at 2 hr of reperfusion in control animals (fig. 5). Plasma t-PA was unaltered by coinfusion of the thrombin inhibitor Ro 46-6240 alone or in combination with the GP IIb/IIIa antagonist Ro 44-9883. Plasma concentrations of Ro 46-6240 reached steady state 15 min after the bolus and infusion of drug and increased in a dose-dependent manner (fig. 6). The plasma concentration of Ro 46-6240 at 5 µg/kg/min (294 ± 49 nM, n = 5) was not altered when it was combined with the GP IIb/IIIa antagonist Ro 44-9883 (380 ± 60 nM, n = 7). Likewise, plasma concentrations of Ro 44-9883 were similar when it was infused alone and when it was combined with Ro 46-6240 (fig. 7).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Plasma t-PA levels at steady state during infusion of the plasminogen activator. There were no significant differences between groups.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Plasma concentrations of Ro 46-6240 at steady-state infusion.

View larger version (38K): [in this window] [in a new window]   Fig. 7.   Plasma concentrations of Ro 44-9883 and Ro 46-6240 when infused alone or in combination.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

There is compelling evidence of increased platelet activation during coronary thrombolysis (Fitzgerald et al., 1991). Thus the formation of TXA₂, determined as excretion or plasma levels of its stable enzymatic metabolites, is increased markedly in patients treated with either t-PA or streptokinase. In addition, antiplatelet agents greatly enhance the response to plasminogen activators in the human (ISIS-2, 1988) and in experimental models of coronary thrombosis (Fitzgerald et al., 1989; Fitzgerald and FitzGerald, 1989; Nicolini et al., 1994; Gold et al., 1988; Mickelson et al., 1990). These agents include antagonists of the platelet GP IIb/IIIa, which suggests that platelet aggregation is a major consequence of the enhanced platelet activation. What triggers this intense platelet activity during thrombolytic therapy is unclear. Several agonists have been implicated, including serotonin and TXA₂ (Fitzgerald et al., 1989; Golino et al., 1988). However, their role may be secondary, because they are released upon platelet activation and serve to amplify the response to primary agonists, such as thrombin, collagen and epinephrine. It is consistent with this hypothesis that inhibition of serotonin or of TXA₂ alone does not prevent reocclusion, which suggests that some other agonist is involved.

In this study, we examined the relationship between thrombin and platelet activation in a well-characterized model of coronary thrombolysis (Fitzgerald et al., 1989; Fitgerald and FitzGerald, 1989). Several lines of evidence implicate thrombin as a platelet agonist in this setting. First, clinical and experimental studies demonstrate increased thrombin activity during coronary thrombolysis in that plasma levels of fibrinopeptide A, a peptide cleaved by thrombin from the A chain of fibrinogen, are elevated (Merlini et al., 1995; Galvani et al., 1994; Eisenberg et al., 1987). In addition, inhibitors of factor Xa enhance the response to thrombolytic therapy, which implies de novo thrombin generation (Sitko et al., 1992). Second, in experimental models of coronary thrombolysis, thrombin inhibition evokes a response similar to that seen with platelet inhibitors (Fitgerald and FitzGerald, 1989; Jang et al., 1990; Yao et al., 1992; Rudd et al., 1992). In this model of coronary thrombolysis, the coronary thrombus formed is platelet-rich, and there is marked platelet activation during and after thrombolysis, detected as an increase in TX formation (Fitzgerald et al., 1989).

We examined the response to Ro 46-6240, a competitive inhibitor of thrombin with an apparent K_i of 34 nM. Moreover, Ro 46-6240 is a potent inhibitor of clot-bound thrombin (Gast et al., 1994). The thrombin-inhibitory activity is highly specific, the K_i for other serine proteases being several orders of magnitude higher (Carteaux et al., 1995). In our experiments, Ro-46-6240 inhibited reocclusion in a dose-dependent manner and, at the highest doses, accelerated reperfusion. As an index of platelet activation in vivo, we measured the formation of TXA₂. TXA₂ is the major cyclooxygenase product of arachidonic acid in platelets and is formed in response to most platelet activators, including thrombin (Offermans et al., 1994, Brune and Ullrich, 1991). It is rapidly inactivated in aqueous medium by hydrolysis to the inactive metabolite TXB₂. Measurement of TXB₂ is fraught with difficulties, because it is generated during blood sampling and its excretion in urine reflects renal production (FitzGerald et al., 1983). TXB₂ is further metabolized enzymatically to several products, including 2,3-dinor-TXB₂. Consequently, we monitored platelet activation by determining the excretion of this stable enzymatic metabolite.

Previous studies in this model of thrombolysis have shown a marked increase in TXA₂ formation coincident with coronary reperfusion (Fitzgerald et al., 1983). The increase in TXA₂ biosynthesis persists unabated for at least 4 hr after reperfusion, long after the discontinuation of t-PA. In this study, the specific thrombin inhibitor Ro 46-6240 suppressed urinary 2,3-dinor-TXB₂ in a dose-dependent manner, and little increase in the thromboxane metabolite was detected at the highest doses. The suppression of 2,3-dinor-TXB₂ coincided with an acceleration of reperfusion and inhibition of reocclusion, which have been shown previously to be platelet-dependent events (Fitzgerald et al., 1989; Fitgerald and FitzGerald, 1989). Thus Ro 46-6240 markedly suppressed functional and biochemical evidence of platelet activation in vivo.

Similar effects were seen when a subthreshold dose of Ro 46-6240 was administered in combination with Ro 44-9883. Ro 44-9883 is a potent peptidomimetic antagonist of the platelet GP IIb/IIIa (Carteaux et al., 1993). At high dose, Ro 44-9883 abolishes coronary reocclusion (Murphy, Pratico and Fitzgerald, unpublished observations), as has been reported with other antagonists of the platelet GP IIb/IIIa. At the dose used, Ro 44-9883 inhibited platelet aggregation in response to ADP ex vivo. Alone, it had no discernible effect on the response to t-PA or on biochemical evidence of platelet activation in vivo. However, when Ro 44-9883 was combined with a subthreshold dose of the thrombin inhibitor, reocclusion was prevented. Similar findings have been reported with a structurally distinct GP IIb/IIIa antagonist, integrelin, combined with the thrombin inhibitor hirudin (Nicolini et al., 1994). Moreover, the combination of Ro 44-9883 and Ro 46-6240 abolished the increase in urinary 2,3-dinor-TXB₂. It is worth emphasizing that this synergistic interaction did not result from an alteration in the plasma concentration of either drug.

These data suggest that one of the consequences of thrombin generation during coronary thrombolysis is activation of the platelet GP IIb/IIIa. The findings of this study also suggest that the increase in TXA₂ formation in this model is due largely to platelet aggregation. TXA₂ is formed upon platelet activation by agonists through G protein-linked activation of phospholipases (Brass et al., 1993). However, the largest component of TXA₂ formation upon platelet stimulation occurs coincident with platelet aggregation (Shattil et al., 1994). This reflects occupancy of the platelet GP IIb/IIIa by fibrinogen, which triggers "inside-out" transmembrane signaling and further TXA₂-dependent cell activation.

In conclusion, thrombin is the primary agonist mediating reocclusion in a model of coronary thrombolysis. Thrombin induces platelet activation and aggregation, which in turn result in a marked increase in TXA₂ formation. The increase in TXA₂ biosynthesis largely reflects platelet aggregation (indeed, this may be a useful marker of aggregation in vivo). Coadministration of a platelet GP IIb/IIIa antagonist and a thrombin inhibitor reduces the dose required of either drug alone, and the combination may prove useful in enhancing the response to thrombolytic therapy of acute myocardial infarction.