Arterial Thrombin Activity After Angioplasty in an Atherosclerotic Rabbit Model
Time Course and Effect of Hirudin
Background A 2-hour infusion of the direct thrombin inhibitor hirudin at the time of balloon angioplasty limits restenosis in the focally atherosclerotic rabbit. Although short-term administration of hirudin may have a prolonged biological effect, the effect of hirudin on vessel thrombin activity has not been previously studied in an animal model of angioplasty. We hypothesized that a short intravenous infusion of hirudin would result in prolonged inhibition of arterial wall–associated thrombin activity (ATA) after angioplasty.
Methods and Results Sixty-one rabbits received recombinant hirudin (r-hirudin) (1 mg/kg bolus plus 1 mg · kg−1 · h−1×2 hours) or bolus heparin (controls, 150 U/kg) intravenously at the time of femoral balloon angioplasty. ATA was measured through exposure of arterial segments ex vivo to fibrinogen and conducting an assay for fibrinopeptide A (FPA). ATA was low in nonballooned, atherosclerotic vessels (FPA=0.5±0.3 ng · mL−1 · mg−1) but increased significantly at 24 hours after angioplasty in the heparin group (3.7±0.9 ng · mL−1 · mg−1, P<.01 versus baseline, n=9) but not in the hirudin group (FPA=1.4±0.3; P=NS versus baseline, P<.02 versus heparin controls, n=8). The time course of ATA after angioplasty was assessed in 44 rabbits. Thrombin activity peaked at 48 hours and declined to baseline at 72 hours and 7 days. FPA values between the heparin and r-hirudin groups were similar at these later time points.
Conclusions A 2-hour intravenous infusion of r-hirudin suppressed ATA measured 24 hours after angioplasty in the focally atherosclerotic rabbit. This prolonged biological effect may account, in part, for the reduction in restenosis seen in this model.
Experimental studies in animals have suggested that thrombin may play a key role in restenosis after balloon angioplasty, a phenomenon reported to occur in 30% to 40% of patients undergoing this coronary intervention. In a rabbit model of focal femoral atherosclerosis, a 2-hour infusion of the direct thrombin inhibitor recombinant hirudin (r-hirudin) reduced angiographic restenosis and resulted in a 50% reduction in cross-sectional area narrowing by plaque 4 weeks after angioplasty compared with heparin-treated control animals.1 The mechanism of the beneficial effect of direct thrombin inhibition on restenosis is not known, but thrombin has been shown to mediate a number of processes that may be involved in the restenosis response. Thrombin acts via specific cellular receptors to activate platelets, enzymatically converts fibrinogen to fibrin, and has been shown to be a major mediator of arterial thrombosis in vivo.2 3 Thrombin is a potent mitogen for smooth muscle cells in culture and stimulates the release of cellular chemoattractants and growth factors.4 5 6 Mural thrombus after angioplasty may serve as both a reservoir of catalytically active thrombin and scaffolding for the ingrowth of multiple cell types producing mitogenic, chemotactic, and vasoactive substances.7 8 The attenuation of any or all of the above processes may contribute to the beneficial effect of thrombin inhibition on restenosis.
Although a short-term infusion of hirudin is known to reduce restenosis in our model, the extent and duration of thrombin inhibition in the vessel wall resulting from short-term hirudin administration are unknown. Hirudin is rapidly cleared from the plasma, but previous results have suggested that its biological effects outlast its plasma half-life.9 In a rabbit model of preexisting jugular venous thrombi, both hirudin and heparin were effective in limiting in vivo clot extension at early time points after drug administration. However, only hirudin was effective in continuing to limit fibrin deposition after it was cleared from the plasma.9 The prolonged effect of hirudin is thought to be due, at least in part, to the ability of hirudin to bind and inhibit clot- and matrix-bound thrombin, which are inaccessible to the heparin/antithrombin III complex.10 Although these observations were important in determining the effects of hirudin on preexisting clot in vitro and in vivo, no study has specifically examined the time course of the effects on vessel wall–associated thrombin activity after arterial injury.
The present study was undertaken to specifically test the hypothesis that a short-term infusion of hirudin would result in prolonged inhibition of arterial wall–associated thrombin activity after angioplasty. To test this hypothesis, we developed a method for determining arterial wall–associated thrombin activity at different time points after angioplasty. We examined the duration and extent of the inhibition by hirudin of thrombin activity in the vessel wall after a 2-hour infusion compared with heparin-treated control animals. We chose 24 hours after angioplasty as our primary end point for determining whether there was sustained inhibition of arterial wall–associated thrombin activity because it is well beyond the 4 hours required for complete plasma clearance of hirudin after intravenous administration in rabbits.
The study design is presented in summary in Fig 1⇓. The rabbits were housed according to Animal Welfare Act specifications, and all surgical procedures were performed with a sterile technique and general anesthesia. Briefly, bilateral femoral artery atherosclerosis was induced in 65 male New Zealand White rabbits through air desiccation injury and a 28-day high cholesterol diet, as described previously.1 Before angioplasty, blood samples were drawn for measurement of activated partial thromboplastin time (aPTT), and baseline angiography was performed as previously described. Animals were then randomly assigned to receive bolus heparin (150 U/kg; porcine intestinal mucosa, 1000 USP U/mL, Elkins-Sinn Inc) or intravenous r-hirudin (1-mg/kg bolus plus a 2-hour infusion [1 mg · kg−1 · h−1]; CGP 39393, CIBA-GEIGY Ltd). After administration of the anticoagulant bolus, femoral artery angioplasty was performed under fluoroscopic guidance at the site of greatest luminal narrowing. Three 1-minute, 10-atm inflations were performed 1 minute apart with a 2.5-mm ACS balloon angioplasty catheter (Advanced Cardiovascular Systems, Inc). After angioplasty, intra-arterial lidocaine was administered at a dose of 2 mg/kg to limit arterial vasospasm, and angiography was repeated 15 minutes later. Catheters were removed, and rabbits were then fed standard rabbit chow until the time of death. The initial group of animals were killed after endothelial injury and 4 weeks of cholesterol feeding (time 0; n=4) or at 24 hours after angioplasty (n=17). Additional animals were killed at 48 hours (n=20), 72 hours (n=11), or 7 days (n=13) after angioplasty. Before death, a Berman catheter was advanced via a left carotid arteriotomy under fluoroscopic guidance into the descending aorta; blood samples were withdrawn for repeat aPTT measurements; and angiography was repeated. Bilateral femoral dissection was then performed to isolate the femoral arteries, and 3-0 silk sutures were placed under each artery at both the proximal and distal ends of the angioplasty site with minimal trauma to the vessel. The distal arterial tree was then perfused with 0.9% saline at 100 mm Hg pressure for 10 minutes. At the start of the infusion, the rabbits were administered an overdose of pentobarbital sodium. Just before the end of the saline infusion, the proximal and distal sutures were tied, leaving a blood-free lumen. The vessels were excised and placed into fresh saline. The sutured segments were trimmed away, leaving the vessel open at its proximal and distal ends. The vessels were then divided in cross section, and both halves were weighed. Vessel segments were then immediately assayed for arterial wall–associated thrombin activity.
Measurement of Vessel Wall–Associated Thrombin Activity
Vessel wall–associated thrombin activity was quantified by measuring fibrinopeptide A (FPA) generation on exposure of each vessel segment ex vivo to fibrinogen. For each assay, 40 mg of human fibrinogen (Chromogenix) was reconstituted in 20 mL PBS (0.01 mol/L PO4, 0.15 mol/L NaCl, pH 7.4) and rocked for 5 minutes at room temperature. Any residual plasminogen was removed from the fibrinogen solution by two additions of cyanogen bromide–activated l-lysine agarose (Sigma Chemical Co) and centrifugation at 2000g for 10 minutes at 4°C.
To determine the specificity of FPA generation as a measure of vessel wall–associated thrombin activity, one half of the first 49 vessels were incubated with hirudin, 10 μg/mL in PBS, for 5 minutes before exposure to fibrinogen, and the other vessel half was incubated in PBS. Because r-hirudin is a specific thrombin inhibitor, suppression of FPA generation by r-hirudin would confirm that FPA production in this model was specifically due to thrombin activity. Both vessel halves were then separately placed into 1 mL of a 2-mg/mL human fibrinogen solution at 37°C for 1 hour. A thrombin standard curve was created at the time of each assay by exposing the fibrinogen to known quantities of thrombin (bovine thrombin, Organon Teknica) ranging from 1×10−2 to 1×10−6 NIH U/mL. After the 1-hour incubation, the vessels were removed from the fibrinogen, and 100 mL of a citrate solution (Anticoagulant Solution, Diagnostica Stago) containing heparin and specific protease inhibitors was added to inhibit any residual thrombin activity. The remaining fibrinogen was precipitated from the solution with bentonite according to the manufacturer’s instructions. After centrifugation, the supernatant was stored at −70°C for measurement of FPA content. FPA levels were measured by a blinded observer using a commercially available FPA competitive ELISA assay (Asserachrom, Diagnostica Stago). FPA values for each vessel were determined with the use of serial dilutions to ensure that the final measured value fell within the linear range of the FPA assay. Background FPA levels were determined for each plasminogen-free fibrinogen solution used in the experiments. Vessel wall–associated thrombin activity was calculated by subtracting the background FPA value from each vessel segment FPA value and then normalizing by the vessel segment weight. After the first 49 vessels were assayed in this manner, the hirudin preincubation step was no longer used because FPA values in hirudin-preincubated vessels were uniformly close to zero and lower than values obtained in baseline vessels.
One-way ANOVA was used to determine differences between treatment groups. For nonnormally distributed series, nonparametric tests of significance were used. In the case of multiple comparisons, the Newman-Keuls method was used to assess for statistical significance. A value of P<.05 was considered significant. Unless otherwise specified, results are expressed as mean±SEM.
FPA Assay Variability and Thrombin Standard Curve Generation
Multiple aliquots of the supernatants from each assay underwent FPA determination as outlined above. The intra-assay variability of our FPA measurement was <1% (correlation coefficient R2=.99; Fig 2⇓). Thrombin standard curves created for each assay demonstrated that the linear range of our assay extended from 1×10−3 to 1×10−6 NIH U/mL. Least squares linear regression on thrombin standard curves demonstrated an average R2=.96.
Coagulation Parameters and Angiography
Arterial blood for measurement of aPTT was drawn before heparin or hirudin treatment at the time of angioplasty (baseline), immediately after angioplasty, and at the time of death. There were no significant differences in coagulation parameters between the hirudin and heparin groups either at baseline, after angioplasty, or at the time of death. At baseline, the average value for aPTT was 123±13 seconds. Immediately after angioplasty, the average value for aPTT was >200 seconds in both treatment groups. In the group killed at 24 hours after angioplasty, average aPTT in the hirudin versus heparin groups was 112±13 versus 108±7 (P=NS), which did not differ significantly from the baseline value.
Quantitative analysis performed on the postangioplasty angiogram in all animals killed at 24 hours after angioplasty revealed no significant differences in minimal luminal diameter between animals in the heparin (1.2±0.1 mm) and hirudin (1.3±0.2 mm) groups. Four of 91 vessels (two vessels at 48 hours and one vessel each at 72 hours and 7 days after angioplasty) that were patent immediately after angioplasty were occluded at the time of death and were excluded from analysis.
Vessel Wall–Associated Thrombin Activity: Effects of Balloon Angioplasty and Hirudin Treatment
The prospectively defined primary end point of our study was to measure vessel wall–associated thrombin activity 24 hours after angioplasty in the two treatment groups. Baseline arterial wall–associated thrombin activity was low in nonballooned, atherosclerotic vessels (FPA=0.5±0.3 ng · mL−1 · mg−1; 3.1±2.3×10−7 NIH thrombin U/mg). At 24 hours after angioplasty, vessel wall–associated thrombin activity was significantly elevated in the heparin-treated group (3.7±0.9 ng FPA · mL−1 · mg−1; 44±12×10−7 NIH thrombin U/mg; P<.01 versus baseline) but was not significantly above baseline in the hirudin-treated group (1.4±0.3 ng FPA · mL−1 · mg−1; 15±4.3×10−7 NIH thrombin U/mg; P=NS versus baseline, P<.02 versus heparin). Therefore, a short-term infusion of hirudin at the time of angioplasty resulted in significantly lower arterial wall–associated thrombin activity at 24 hours compared with bolus heparin (Fig 3⇓).
Vessel Wall–Associated Thrombin Activity: Time Course After Balloon Angioplasty
After documenting vessel wall–associated thrombin activity at baseline and 24 hours after angioplasty, we determined arterial wall–associated thrombin activity at three later time points after angioplasty. In the heparin control animals, vessel thrombin activity peaked at 48 hours after balloon angioplasty (9.0±2.0 ng FPA · mL−1 · mg−1; 110±25×10−7 NIH thrombin U/mg; P<.0001 versus baseline). By 72 hours, vessel thrombin activity returned toward baseline levels (FPA=1.6±0.7 ng · mL−1 · mg−1; 16±9.4×10−7 NIH thrombin U/mg; P=NS versus baseline) and remained low at 7 days after angioplasty (FPA=3.0±1.3 ng · mL−1 · mg−1; 34±16×10−7 NIH thrombin U/mg; P=NS versus baseline). Animals treated with hirudin had FPA values of 7.5±2.2, 2.1±0.7, and 2.2±1.0 ng · mL−1 · mg−1 at 48 hours, 72 hours, and 7 days, respectively. These values are similar to those found in the heparin group. Thus, the difference between the hirudin- and heparin-treated animals in arterial thrombin activity 24 hours after balloon angioplasty was no longer present at the later time points studied (Fig 4⇓).
Effect of Ex Vivo Hirudin Preincubation
Exposure of vessel segments to excess hirudin (10 μg/mL) ex vivo, before incubation with fibrinogen, suppressed >95% of FPA generation compared with non–hirudin-preincubated segments in both treatment groups at all time points studied (0.09±0.02 versus 3.9±0.5 ng FPA · mL−1 · mg−1; P<.00001). This indicates that the generation of FPA in the assay system was specifically thrombin mediated (Fig 5⇓).
In this study, we hypothesized that a 2-hour infusion of hirudin at the time of balloon angioplasty would result in prolonged suppression of arterial wall–associated thrombin activity. We prospectively defined, as our primary end point, measurement of arterial wall–associated thrombin activity at 24 hours after angioplasty because it is a time well beyond the 4 hours expected for total plasma clearance of hirudin in rabbits.11 Vessel wall–associated thrombin activity was low in nonballooned, atherosclerotic vessels and rose significantly at 24 hours after balloon angioplasty in animals treated with bolus heparin. In contrast, arterial wall–associated thrombin activity in the hirudin-treated group did not rise significantly at 24 hours after angioplasty compared with baseline and was significantly lower than in the heparin-treated group. This finding supports our original hypothesis. In addition, we assessed the time course of arterial wall–associated thrombin activity at three later time points after angioplasty. Arterial wall–associated thrombin activity peaked at 48 hours, declined to baseline levels by 72 hours, and remained low at 7 days.
Thrombin and Restenosis
We have previously shown that this same 2-hour infusion of hirudin reduces restenosis at 28 days after balloon angioplasty in the rabbit model compared with heparin-treated control animals.1 Because hirudin is known to be a potent and specific direct inhibitor of thrombin, we concluded that inhibition of thrombin is a key factor leading to a reduction in restenosis in this model. Thrombin has multiple cellular effects that may promote the development of restenosis. Thrombin acts by cleaving the terminal portion of its cellular receptor, unmasking a new amino terminus, which acts as a tethered ligand to activate the receptor.12 This receptor mediates thrombin-induced platelet activation as well as smooth muscle cell mitogenesis.5 12 Thrombin is also a chemoattractant for multiple cell types producing growth factors such as platelet-derived growth factor, basic fibroblast growth factor, and transforming growth factor-β, all of which have been implicated in the restenosis response.6 Finally, thrombin is a major mediator of arterial thrombosis after deep arterial injury, and the resulting thrombus can be incorporated into the plaque or serve as a scaffolding for cell types involved in the restenosis cascade.8 13
Direct Thrombin Inhibitors: Evidence for Prolonged Effects
The ability of hirudin to bind and inhibit clot- and matrix-bound thrombin may result in biological effects in tissues that outlast the plasma half-life of the drug. Agnelli et al9 demonstrated that 3-hour infusions of heparin and hirudin were both effective in inhibiting fibrin deposition into preformed thrombi in the jugular veins of rabbits. However, venous thrombi of the hirudin-treated animals continued to resist fibrin accretion 9 hours after discontinuation of the infusion (long after the 4 hours necessary for plasma clearance), whereas clots in heparin-treated animals continued to extend during that time period.9 The present study is the first description in the literature of a prolonged inhibitory effect of hirudin on vessel wall–associated thrombin activity in an animal model of balloon angioplasty. Although it cannot be proved from the current observations that the prolonged inhibition of vessel wall–associated thrombin activity limits long-term restenosis, there are several reasons why a prolonged inhibition may be necessary to reduce restenosis. Once formed on an injured vessel wall, residual thrombus may be even more thrombogenic than deep injury14 and can serve as a reservoir of catalytically active thrombin. This may have significant implications for the development of the restenosis lesion, as the mitogenic effects of thrombin on cultured vascular smooth muscle cells require prolonged exposure to thrombin.15 Studies in an animal model of vascular injury have confirmed that thrombin receptor expression remains elevated for a prolonged period after injury and is highest in areas of actively proliferating cells.16 This supports the hypothesis that thrombin acts (possibly with other known mitogens, such as basic fibroblast growth factor and platelet-derived growth factor) to initiate and maintain cell proliferation after vascular injury.
A second finding in our study was that vessel thrombin activity at 48 hours after angioplasty was no different in the hirudin- and heparin-treated animals and was, in fact, higher than thrombin activity measured at 24 hours after angioplasty in both groups. Thus, the main difference between hirudin and heparin treatment in terms of suppression of vessel wall–associated thrombin activity occurred early (ie, <48 hours after angioplasty). As hirudin has been shown to effectively limit restenosis and heparin has a minimal effect in this model, the above finding suggests that an early time window of effective thrombin inhibition after angioplasty is crucial for limiting restenosis.
Return of Vessel Anticoagulant Activity
The abrupt decrease in vessel wall thrombin activity from 48 to 72 hours after angioplasty suggests that the endogenous anticoagulant activity of the vessel wall begins to return by 3 days after intervention. Pasche et al17 reported that thrombin activity of the rabbit aorta increased substantially early after balloon injury and returned to baseline by 7 days. Many processes during the healing phase may be responsible for this, including increased production of endogenous anticoagulants such as tissue-type plasminogen activator, antithrombin III, and heparan sulfates. Endothelial regrowth, which is a key factor for maintenance of long-term anticoagulant activity of the vessel wall, may begin as early as 3 days after angioplasty but may not be complete until 7 to 14 days after deep arterial injury.18 It should be noted, however, that vessel thrombin activity in the present study showed a trend toward persistent elevation above baseline levels between 3 and 7 days after balloon injury. Previous studies in a rabbit aortic injury model have indicated that elevations in thrombin activity may be detected for up to 2 weeks after injury.19 It is currently not known whether a chronic low-level elevation in thrombin activity is important for lesion development after angioplasty in our model.
The ex vivo measurement of thrombin activity in the present study is unique in that the FPA generation reflects thrombin activity associated with the entire vessel wall, not only the luminal surface, as measured in a previous study.17 We could not determine the precise location within the vessel of the thrombin activity measured in our assay because the assay itself permanently alters vessel histology. Although previous studies with light microscopy have not demonstrated differences in the degrees of mural thrombus between hirudin- and heparin-treated animals at 24 hours after angioplasty,1 we cannot exclude the possibility that mural thrombus was a source of FPA generation because the biochemical assay used in the present study measures functionally active rather than histologically apparent thrombin. Nevertheless, our assay likely reflects thrombin activity from additional sources within the vessel wall, as balloon angioplasty is known to cause deep arterial injury with thrombosis potentially occurring at the luminal surface, within plaque, at sites of medial tears, or in the adventitia. Finally, the relevance of animal models to the investigation of thrombin activation in human coronary disease is unknown. However, a recent report20 indicates that markers of thrombin generation measured in blood from human coronary arteries early after balloon angioplasty significantly increase in a substantial proportion of patients undergoing the procedure, despite adequate heparinization. Furthermore, the direct antithrombin agents hirudin and hirulog have been shown to decrease the incidence of acute complications occurring after percutaneous transluminal coronary angioplasty in human trials.21 22
The present study demonstrated that a 2-hour infusion of hirudin in rabbits at the time of angioplasty, in a dose that has been shown to reduce restenosis, results in a significant decrease in arterial wall–associated thrombin activity 24 hours after injury in comparison with bolus heparin. We have further demonstrated that arterial wall–associated thrombin activity remains significantly elevated above baseline for at least 48 hours after angioplasty and returns toward baseline at 72 hours and 7 days after injury. The decrease in vessel thrombin activity in hirudin-treated compared with heparin-treated animals, however, was no longer present at 48 hours after angioplasty. This suggests that the early time period before 48 hours after balloon angioplasty may be a crucial window for thrombin inhibition to limit restenosis in this model. These observations may also provide insights into the time course of vessel wall “passivation” after deep arterial injury.
This study was supported in part by the Department of Health and Human Services Research Grant (NIH) R01-HL-47849 (Dr Sarembock), NIH Physician Scientist Award HL-02592 (Dr Humphries), and American Heart Association, Virginia Affiliate, Grant-in-Aid VA-94-G-18 (Dr Gimple).
Reprint requests to Ian J. Sarembock, MB, ChB, MD, Cardiovascular Division, Department of Medicine, Box 158, University of Virginia Health Sciences Center, Charlottesville, VA 22908.
- Received October 16, 1995.
- Revision received December 13, 1995.
- Accepted December 19, 1995.
- Copyright © 1996 by American Heart Association
Sarembock IJ, Gertz SD, Gimple LW, Owen RM, Powers ER, Roberts WC. Effectiveness of recombinant desulfatohirudin in reducing restenosis after balloon angioplasty of atherosclerotic femoral arteries in rabbits. Circulation. 1991;84:232-243.
Heras M, Chesebro JH, Penny WJ, Bailey KR, Badimon L, Fuster V. Effects of thrombin inhibition on the development of acute platelet-thrombus deposition during angioplasty in pigs. Circulation. 1989;79:658-665.
Lam JYT, Chesebro JH, Steele PM, Heras M, Webster MWI, Badimon L, Fuster V. Antithrombotic therapy for deep arterial injury by angioplasty: efficacy of common platelet inhibition compared with thrombin inhibition in pigs. Circulation. 1991;84:814-820.
McNamara CA, Sarembock IJ, Gimple LW, Fenton IJ, Coughlin SR, Owens GK. Thrombin stimulates proliferation of cultured rat aortic smooth muscle cells by a proteolytically activated receptor. J Clin Invest. 1993;91:94-98.
Agnelli G, Renga C, Weitz JI, Nenci GG, Hirsh J. Sustained antithrombotic activity of hirudin after its plasma clearance: comparison with heparin. Blood. 1992;80:960-965.
Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin III-independent inhibitors. J Clin Invest. 1990;86:385-391.
Heras M, Chesbro JH, Webster MWI, Mruk JS, Grill DE, Penny WJ, Bowie EJW, Badimon L, Fuster V. Hirudin, heparin, and placebo during deep arterial injury in the pig: the in vivo role of thrombin in platelet-mediated thrombosis. Circulation. 1990;82:1476-1484.
Badimon L, Lassila BJ, Vallabhajosula S, Chesebro JH, Fuster V. Residual thrombus is more thrombogenic than severely damaged vessel wall. Circulation. 1988;78(suppl II):II-119. Abstract.
Bachhuber BG, Sarembock IJ, Gimple LW, McNamara CA, Owens GK. Thrombin-induced mitogenesis in cultured rat aortic smooth muscle cells requires prolonged thrombin exposure. Am J Physiol. 1995;268:C1141-C1147.
Wilcox JN, Rodriguez J, Subramanian R, Ollerenshaw J, Zhong C, Hayzer DJ, Horaist C, Hanson SR, Lumsden A, Salam TA, Kelly AB, Harker LA, Runge M. Characterization of thrombin receptor expression during vascular lesion formation. Circ Res. 1994;75:1029-1038.
Topol EJ, Bonan R, Jewitt D, Sigwart U, Kakkar VV, Rothman M, de Bono D, Ferguson J, Willerson JT, Strony J, Ganz P, Cohen MD, Raymond R, Fox I, Maraganore J, Adelman B. Use of a direct antithrombin, hirulog, in place of heparin during coronary angioplasty. Circulation. 1993;87:1622-1629.
Cannon CP, Braunwald E. Hirudin: initial results in acute myocardial infarction, unstable angina, and angioplasty. J Am Coll Cardiol. 1995;25(suppl):30S-37S.