Randomized, Double-blind Comparison of Hirulog Versus Heparin in Patients Receiving Streptokinase and Aspirin for Acute Myocardial Infarction (HERO)

The administration of aspirin and thrombolytic therapy has been a major advance in the management of acute myocardial infarction, with several large studies demonstrating improved survival and preservation of left ventricular function.¹²³⁴⁵⁶ Early, complete reperfusion is associated with lower mortality. In the angiographic substudy of the GUSTO-I trial,⁷ mortality at 30 days was 4.4% among patients with TIMI grade 3 flow⁸ (complete perfusion) in the infarct-related artery at 90 minutes, 7.4% among patients with TIMI 2 flow (partial perfusion), and 8.9% among patients with TIMI 1 or 0 flow (occlusion). TIMI 3 flow was also associated with significant improvement in left ventricular function.⁹ Despite the clinical benefits of all the fibrinolytic regimens studied in the GUSTO-I trial, TIMI 3 flow at 90 minutes in the infarct-related artery was achieved in only 33% of patients treated with streptokinase, aspirin, and intravenous heparin and 54% of those treated with accelerated TPA, aspirin, and intravenous heparin.⁹ Reocclusion may occur in the first week in 5% to 15% of reperfused infarct-related arteries, with approximately half occurring in the first 24 hours,⁹¹⁰ and is associated with a marked increase in morbidity and mortality.¹¹

Thrombolysis and thrombosis are simultaneously occurring processes,¹² and a number of factors affect thrombosis during and after thrombolytic therapy. A residual mural thrombus containing fibrin-bound thrombin can promote local fibrin formation and is the most potent stimulus for platelet aggregation,¹³ and this potentiation is not inhibited by aspirin. Thrombolytic therapy generates plasmin, which is procoagulant, by activating factor V to Va and accelerating thrombin generation via the prothrombinase complex.¹⁴ Both streptokinase and TPA may activate platelets,¹⁵¹⁶¹⁷ and antibodies to streptokinase may also promote platelet aggregation.¹⁸ During effective thrombolytic therapy, the dynamic balance of lysis and rethrombosis is pivoted in favor of lysis. Adjuvant heparin therapy has not been shown to improve 90-minute patency with streptokinase⁷ or TPA¹⁹ and has several limitations apart from the small risk of bleeding. It has only a small effect on fibrin-bound thrombin,²⁰ and factor Xa in the prothrombinase complex is protected from heparin–antithrombin III.²¹ Furthermore, platelet factor 4 neutralizes heparin, and fibrin monomer II inhibits heparin–antithrombin III.²¹

Hirulog is a 20–amino acid synthetic peptide that directly inhibits free and clot-bound thrombin²² and, when used in appropriate regimens as an adjuvant during thrombolytic therapy, may prevent clot formation and extension and facilitate clot lysis. The present report is of the double-blind HERO trial, in which the safety and efficacy of two hirulog regimens for achieving early and complete flow of the infarct-related artery were compared with heparin among patients with acute myocardial infarction receiving streptokinase and aspirin.

Methods

Study Population

Patients were eligible if they presented within 12 hours of the onset of symptoms of infarction with ≥1 mm of ST-segment elevation in ≥2 limb leads and/or leads V₄ through V₆ of a 12-lead ECG or ≥2 mm of ST-segment elevation in ≥2 contiguous precordial V₁ through V₃ leads. Exclusion criteria included previous administration of streptokinase; history of systemic, gastrointestinal, or genitourinary bleeding within 3 months; history of cerebrovascular disease including stroke or transient ischemic attack within 6 months; history of intracranial neoplasm, arteriovenous malformation, or aneurysm; severe trauma or major surgery within the previous 3 months; puncture of noncompressible vessels within the previous 10 days; traumatic cardiopulmonary resuscitation within the previous month; severe and uncontrolled hypertension; and cardiogenic shock.

The study design was approved by the ethics committees at the participating centers, and all patients gave written, informed consent.

Study Design

Aspirin 150 to 325 mg was administered as soon as possible after admission and continued daily. Eligible patients were then randomized by telephoning a 24-hour service at the Clinical Trial Service Unit, Oxford, England. All patients were to receive 1.5 million units of streptokinase over a period of 30 to 60 minutes and were randomly allocated, in a double-blind manner, to receive either active heparin and placebo hirulog or one of two different blinded active hirulog regimens and placebo heparin (ie, a “double dummy” procedure):

1. Heparin regimen—5000 U intravenous bolus followed by an infusion of 1000 U/h in patients weighing <80 kg and 1200 U/h in patients weighing ≥80 kg for up to 60 hours. At 11 and at 24 hours, the heparin/placebo infusion could be adjusted upward according to a nomogram.

2. Low-dose hirulog regimen—0.125 mg/kg intravenous bolus followed by an infusion of 0.25 mg · kg⁻¹ · h⁻¹ for 12 hours and then 0.125 mg · kg⁻¹ · h⁻¹ for up to 60 hours.

3. High-dose hirulog regimen—0.25 mg/kg intravenous bolus followed by an infusion of 0.5 mg · kg⁻¹ · h⁻¹ for 12 hours and then 0.25 mg · kg⁻¹ · h⁻¹ for up to 60 hours.

The infusions were not to be adjusted downward during the first 24 hours unless there was bleeding. If clinically indicated (eg, aPTT >120 seconds or excessive oozing) after the first 24 hours of dosing, the rate of both infusions could be reduced by one third and maintained for 36 to 60 hours until the second angiogram (see below).

It was recommended that oral β-blocker therapy be commenced in patients without contraindications and that aspirin and oral β-blocker therapy be continued at discharge. Other medications could be given according to local practice.

Evaluation of Efficacy

The primary outcome was patency at 90 to 120 minutes as measured by TIMI 3 flow. Secondary outcomes were patency at 2 to 3 days; left ventricular function, global ejection fraction, end-systolic volume, and regional wall motion at 2 to 3 days in patients without a prior history of myocardial infarction; and incidence of death, stroke, reinfarction and cardiogenic shock at 35 days after treatment.

Cardiac Catheterization

Angiograms were performed 90 to 120 minutes after the commencement of the thrombolytic infusion. Infusion of study drugs was continued during the procedure. The infarct-related artery was imaged first, with the initial imaging aimed at optimizing assessment of TIMI flow by visualizing the entire vessel. Three LAO and two RAO views were taken of the left coronary artery, and at least one LAO and one RAO view were taken of the right coronary artery. Additional views were also taken to clearly image the infarct-related stenosis without overlapping vessels. A second angiogram was to be performed 36 to 60 hours after the start of the study infusion, unless rescue angioplasty was performed. The same views used in the first angiogram were duplicated, and left ventriculography in the 40° RAO projection with a grid or sphere for volume measurements was also performed.

All angiograms were analyzed at the Green Lane Hospital quantitative angiography laboratory using the Cardiovascular Measurement System. The film sequence and identification were blinded by an independent technician who cut the dates off the films. Patency was assessed by use of the TIMI flow scoring system⁸ on the first injection of contrast by two independent, experienced core laboratory reviewers. A third cardiologist became involved in the review process if agreement was not reached by the first two reviewers, and consensus was then reached by all three reviewers. Ventriculograms were analyzed separately by use of the Cardiovascular Measurement System for regional wall-motion analysis.

Bleeding

Patients were carefully monitored for bleeding. Bleeding was defined as major if it was clinically overt, with either a fall in the hemoglobin level of ≥3 g/dL or requirement for transfusion of two or more units of blood, or if it was retroperitoneal or intracranial. Bleeding was defined as minor if it was clinically overt but did not meet these criteria.

Reinfarction

Reinfarction within 18 hours after commencement of thrombolytic therapy was defined as recurrent, severe ischemic chest discomfort lasting >30 minutes accompanied by new or recurrent ST-segment elevation of ≥1 mm (≥0.1 mV) in at least two contiguous leads. Reinfarction after 18 hours was defined as development of new Q waves in two or more leads or reelevation of creatine kinase by ≥50% to above the upper limit of normal.

Statistical Analysis

Sample-size calculations were based on having ≥80% power at P<.05 to detect improvements in TIMI 3 flow at 90 to 120 minutes from 30% with heparin²³²⁴ to 46% with the combination of the two hirulog regimens (1:2 comparison) or to 48% with one of the hirulog regimens (1:1 comparison). It was estimated that 375 patients would be required (125 in each of the three groups). The independent data safety monitoring committee was to review interim results on the first 75 and 150 patients. All patients were to be included in intention-to-treat analyses. ANOVA and tests of trend were used to compare outcome, with probability values <.05 considered significant for primary outcome measures. Data are presented as mean±SD.

Results

Four hundred twelve patients were randomized from 26 hospitals in five countries (see “Appendix”). The baseline characteristics of the three groups of patients were similar, as shown in Table 1. The mean age was 62±12 years, and 75% of the patients were men. All patients received aspirin. The time from symptom onset to streptokinase initiation was similar in the three treatment groups, as was the duration of the streptokinase infusion and the time to initiation of the study drugs.

Angiographic Results

Four hundred patients (97%) had an early angiogram. The reasons for patients not having an angiogram included death in one, technical or equipment failure in two, withdrawal of consent in one, hemodynamic stability in four, and miscellaneous reasons in four patients. TIMI flow was assessable in 393 patients.

The Figure and Table 2 show early TIMI 3 flow for the three treatment groups. Patients allocated hirulog had TIMI 3 flow more frequently than those allocated heparin (P=.024). TIMI 3 patency was achieved in 35% of patients allocated heparin (95% CI, 28% to 44%) versus 46% with low-dose hirulog (95% CI, 38% to 55%) and 48% with high-dose hirulog (95% CI, 40% to 57%). Patients treated with high-dose hirulog showed a significant increase in early TIMI 3 flow patency compared with heparin that represented a 37% relative improvement (P=.03). No significant difference was observed in the combination of TIMI 2 and 3 flow (69% heparin versus 76% low-dose hirulog versus 68% high-dose hirulog).

The differences in TIMI 3 flow between high-dose hirulog, low-dose hirulog, and heparin appeared to be greater among patients treated earlier after the onset of symptoms (Table 2). Among patients treated within 3 hours after symptom onset, the rate of TIMI 3 patency was 52% in those treated with heparin versus 56% in those treated with low-dose hirulog and 70% in those treated with high-dose hirulog. The effect observed with hirulog compared with heparin among patients treated within 3 hours was not statistically significantly different from that observed among those treated later. For patients treated within 6 hours, the rate of TIMI 3 patency was 53% with high-dose hirulog.

Two- to Three-Day Angiogram

A second angiogram was performed in 323 patients (81%). Because of a higher death rate, a greater need for rescue angioplasty (13 heparin, 4 low-dose hirulog, and 6 high-dose hirulog patients), and a higher incidence of bleeding, fewer patients in the heparin group underwent a second angiogram. Sustained patency (maintenance of TIMI 3 flow, excluding patients undergoing angioplasty) was similar in the three treatment groups: heparin 70%, low-dose hirulog 78%, and high-dose hirulog 77% (probability value for trend, .18). Reocclusion (TIMI 2 or 3 flow reducing to TIMI 0 or 1 flow) occurred in 7% of patients treated with heparin, 4.6% of patients treated with low-dose hirulog, and 1.3% of patients treated with high-dose hirulog (probability value for trend, .09) (Table 3).

There were 262 patients with first infarctions who underwent assessment of left ventricular function at the angiogram on the second or third day (Table 4). No differences in ejection fraction, end-systolic volume, or regional wall motion were observed between the treatment groups.

Laboratory Results

Patients treated with low-dose or high-dose hirulog had less variability and lower aPTTs at both 11 hours (136±66 seconds for heparin versus 96±30 seconds for low-dose hirulog and 117±35 seconds for high-dose hirulog; P<.001 for aPTT levels in heparin versus hirulog patients) and 24 hours after commencement of streptokinase (95±51 seconds for heparin versus 67±21 seconds for low-dose hirulog and 83±23 seconds for high-dose hirulog; P<.001). The peak creatine kinase levels were similar: heparin 1944±140 U/L, low-dose hirulog 2220±237 U/L, and high-dose hirulog 1934±156 U/L.

Clinical Events

Table 5 shows the frequency of major clinical events. There were slightly fewer deaths and cases of cardiogenic shock and/or recurrent myocardial infarction among patients treated with high-dose hirulog, but these differences were not significant. For the combined end point of death or nonfatal myocardial infarction at 35 days, the rates were 15% with heparin, 11% with low-dose hirulog, and 8.8% with high-dose hirulog (P=.26).

Safety

There were two strokes in the heparin group (neither considered to be hemorrhagic), none in the low-dose hirulog group, and two (one of which was confirmed to be hemorrhagic) in the high-dose hirulog group (Table 6). Major bleeding and the need for transfusion occurred less frequently in patients treated with hirulog than in those treated with heparin (P<.05). Among patients receiving the high-dose hirulog regimen, there were nonsignificantly fewer major bleeds than with heparin. The main cause of bleeding was the catheterization site (42% of bleeds among heparin-allocated patients versus 40% of low-dose hirulog bleeds and 38% of high-dose hirulog bleeds).

Discussion

This study demonstrates that hirulog is more effective at achieving early TIMI 3 flow than heparin as an adjunct to streptokinase and aspirin in acute myocardial infarction and that the effect of hirulog may be dose dependent. The improved efficacy of hirulog was achieved at lower aPTT levels without increased bleeding.

Although hirulog and other agents have been shown to improve patency in animal models,²⁵²⁶ this is the first time that improved TIMI 3 flow has been shown with these agents in humans. For patients treated within 3 hours of symptom onset, the difference in TIMI 3 flow between hirulog and heparin appeared to be greater than among those treated later, but this was a post hoc analysis, and the effects of early and later treatment were not significantly different from each other. For patients randomized within 6 hours, the TIMI 3 flow rate with streptokinase and high-dose hirulog was similar to that observed with accelerated TPA and heparin in the GUSTO-I trial,²³ and there was a trend toward fewer major bleeds with high-dose hirulog than with heparin. As expected, because of the effects of hypercontractility of the noninfarct zones and myocardial stunning affecting contractility in the first few days after myocardial infarction,²⁷ there was no effect on global or regional left ventricular function.

Timing of Administration of Hirulog

Early administration of a direct thrombin inhibitor, either before administration of a thrombolytic agent or soon after, might be expected to inhibit more effectively the procoagulant effects of thrombolytic therapy. This is difficult to achieve in a trial setting because investigators want to provide thrombolytic therapy as soon as possible, and delays occur in obtaining consent, in randomization, and in preparation of blinded infusions.

In the recent TIMI-9B²⁸ and GUSTO-IIB²⁹ trials, which compared hirudin with heparin, the delays after initiation of thrombolytic therapy were 50 minutes and 35 minutes, respectively, and the combined results for the comparison of hirudin versus heparin did not indicate any difference in the 30-day rates of death and nonfatal myocardial infarction. These results in patients receiving thrombolytic therapy might have been importantly affected by the delay between administration of thrombolytic therapy and commencement of heparin or hirudin. In the HERO study, hirulog or heparin was administered at a mean time of 24 minutes after the infusion of streptokinase was begun, and consequently the effects of earlier initiation of antithrombin therapy may also have been underestimated.

Choice of Dose

The choice of an appropriate dose of drug that maximizes efficacy without inducing unacceptable bleeding is critical for the clinical development of antithrombotic drugs. In the current study, the hirulog doses chosen were based on prior experience with hirulog in various indications (including prevention of deep vein thrombosis, angioplasty, unstable angina, and acute myocardial infarction).^{2430313233343536} A three-part dosing scheme was used: a bolus infusion, lytic-phase dosing, and reduced-rate maintenance dosing. Pharmacokinetic data have demonstrated that bolus dosing is necessary to rapidly attain target steady-state plasma levels, and the dose was chosen to avoid peak blood levels that might be associated with bleeding. The lytic-phase dosing used either a dose of 0.5 mg · kg⁻¹ · h⁻¹, which is similar to that used previously in acute myocardial infarction,²⁴ or a lower dose, 0.25 mg · kg⁻¹ · h⁻¹. Experience in unstable angina and prevention of deep vein thrombosis has demonstrated that doses between 0.12 and 1 mg · kg⁻¹ · h⁻¹ actively suppress thrombotic disease activity,^30313233 and so maintenance doses in this range were used. A previous trial of patients with acute myocardial infarction who were given streptokinase reported a higher rate of TIMI 3 flow with a 0.5 mg · kg⁻¹ · h⁻¹ dose of hirulog than with a higher dose of 1 mg · kg⁻¹ · h⁻¹.²⁴ Only 55 patients were included in that comparison, and unlike the current trial, bolus doses of hirulog were not administered. The high and low doses used in the current trial were half those used in the previous trial, ie, our high dose was the same as the lower, more effective dose used in the previous trial.

Bleeding

Even at the higher dose, hirulog was associated with more stable and lower aPTTs than heparin. The heparin infusion was adjusted with the use of a nomogram, which may have facilitated achievement of the target aPTT. The incidence of major bleeding was not increased with either of the hirulog regimens. Although hirudin and hirulog should have similar efficacy because they act through similar mechanisms,²⁶ they may have different risk-benefit ratios because of differences in elimination (the half-life of hirudin is 2 to 3 hours whereas that of hirulog is 36 minutes).²⁶ Excess bleeding occurred with high-dose hirudin in the TIMI-9A³⁷ and GUSTO-IIA³⁸ trials. In the GUSTO-IIB²⁹ and TIMI-9B²⁸ trials with a lower dose of hirudin (bolus reduced from 0.6 to 0.1 mg/kg and infusion reduced from 0.2 to 0.1 mg · kg⁻¹ · h⁻¹) and in the absence of protocol-mandated early cardiac catheterization, the incidence of major bleeding in thrombolytic-treated patients was similar with hirudin and heparin.³⁹

Conclusions

Streptokinase continues to be used widely, not only because it is cheaper but also, at least in part, because it is associated with a lower incidence of intracerebral bleeding than TPA.²³⁴⁰⁴¹ Indeed, even when thrombolytic therapy was provided free of charge in the GUSTO-II trial, investigators chose streptokinase as the drug of choice for 10% of their patients in the United States and 42% elsewhere. The impact of streptokinase on early patency has greater potential for improvement than that seen with TPA, and some investigators have found streptokinase to be more procoagulant than TPA,⁴² so that patients treated with streptokinase may be particularly likely to benefit from adjunctive direct thrombin inhibition therapy. If these beneficial effects can be achieved without substantially increasing the risk of intracranial or other major hemorrhage, this might translate into worthwhile improvement in clinical outcome and, potentially, lower costs.

The present study shows that early patency achieved with streptokinase can be improved by adjunctive administration of the direct-acting thrombin inhibitor hirulog. The improved antithrombotic effect and the gain in patency were achieved at lower aPTT levels and were not associated with an increased bleeding risk, possibly because of the favorable pharmacokinetic profile of hirulog. Early successful reperfusion of the infarct-related artery is associated with improved survival.⁷ The present trial was not, however, large enough to reliably assess the effects of hirulog treatment on clinical events, and the trend toward fewer deaths and cases of cardiogenic shock or recurrent infarction with high-dose hirulog was not significant. Also, differences in early angiographic patency and trends for lower rates of reocclusion may not necessarily translate into reduced mortality. Large-scale, randomized trials comparing aspirin plus an appropriate dose of hirulog versus aspirin plus heparin (and/or aspirin alone) among patients receiving thrombolytic therapy are now needed to determine whether these beneficial effects on early patency translate into worthwhile reductions in mortality and other major clinical events that outweigh hemorrhagic risks.

Selected Abbreviations and Acronyms

Appendix A1

HERO Trial Personnel

Liaison committee: H. White (chairman), R. Collins, P. Théroux, B. Adelman, K. Findlen, S. Tremlett, and P. Butera.

Data and safety monitoring committee: C. Hennekens (chairman), R. Peto, D. Julian, and D. Hunt.

Electrocardiographic core laboratory: D. Goodman.

Angiographic core laboratory: H. White, J. French, B. Webber, and B. Williams (Green Lane Hospital, Auckland, New Zealand).

Trial monitoring and data analysis: Corning Besselaar.

Trial investigators (in order of number of patients randomized): P. Aylward (Flinders Medical Centre, Adelaide, Australia), M. Frey (Heart Center of Sarasota, Fla), H. White (Green Lane Hospital, Auckland, New Zealand), J. Adgey (Royal Victoria Hospital, Belfast, Northern Ireland), R. Nair (University Hospital of Cleveland, Ohio), S. Hillis (Western Infirmary, Glasgow, Scotland), Y. Shalev (Sinai Samaritan Medical Center, Milwaukee, Wis), M. Brown (Royal Adelaide Hospital, Adelaide, Australia), H. Ikram (Christchurch Hospital, New Zealand), P. Thompson (Sir Charles Gairdner Hospital, Perth, Australia), J. Machecourt (Centre Hospitalier et Universitaire de Grenoble, France), D. Reid (Freeman Hospital, Newcastle Upon Tyne, England), M. Antoine Slama (Hôpital Béclère, Paris, France), R. Foale (St Mary’s Hospital, London, England), G. Nelson (Royal North Shore Hospital, Sydney, Australia), A. Thomson (Royal Hobart Hospital, Australia), H. Charleson (Waikato Hospital, Hamilton, New Zealand), J. Horowitz (Queen Elizabeth Hospital, Adelaide, Australia), J. Puel (Place du Docteur Baylac, Paris, France), G. Timmis (William Beaumont Hospital, Royal Oak, Mich), I. Sarembock (UVA Health Sciences Center, Charlottesville, Va), R. Stomel (Botsford Hospital, Farmington Hills, Mich), M. Schwartz (Cardiology Associates, Annapolis, Md), J. Kaski (St Georges Medical School, London, England), M. Been (Walsgrave Hospital, England), and C. Guerot (Hôpital Boucicaut, Paris, France).

1A full list of the HERO trial investigators and participating centers appears in the Appendix.

• Download figure
• Download PowerPoint

Footnotes