OUP user menu

★ fast track ★

Adjunctive antithrombotic therapy with primary percutaneous coronary intervention in ST elevation myocardial infarction: ATOLL in perspective

Harvey D. White
DOI: http://dx.doi.org/10.1093/eurheartj/ehq317 ehq317 First published online: 11 October 2011

This commentary refers to ‘Intravenous enoxaparin or unfractionated heparin in primary percutaneous coronary intervention for ST-elevation myocardial infarction: the international randomised open-label ATOLL trial’, by G. Montalescot et al., doi:10.1016/S0140-6736(11)60876-3, published in The Lancet 2011;378:693–703.

An ATOLL is a circular-shaped coral reef surrounding a lagoon providing protection to the marine life. The Tuvalu Atoll is 3–5 m above sea level and is the country most threatened by climate change.

Embedded Image

Tuvalu Atoll

The prevalence of myocardial infarction (MI) is falling, related to better primary prevention and pharmacological control of risk factors.1,2 Mortality is also falling as reperfusion therapies are instituted in a timely manner3 and, in some countries, in-hospital mortality following primary percutaneous coronary intervention (PCI) is ∼4%.4 These achievements are a credit to the investigators who developed the evidence base and to those who have implemented the knowledge and developed systems of care. However, there are further improvements that need to be made.

For patients with ST elevation myocardial infarction (STEMI), timely reperfusion is of the utmost importance, and it remains bewildering and concerning as to why ∼40% of eligible patients still do not receive any form of reperfusion therapy and, when they do, they often receive it late.5

Adjunctive antiplatelet and anticoagulant therapy have improved outcomes with fibrinolysis5 and, when fibrinolysis is coupled with PCI, enoxaparin reduces death or MI at 30 days compared with unfractionated heparin (UFH); 13.0% vs. 16.7% UFH, relative risk (RR) 0.77, P = 0.013.

Antithrombotic therapy with primary PCI is necessary to decrease thrombotic complications of death/MI and stent thrombosis, but for adjunctive antiplatelet and anticoagulant therapy there is a smaller evidence base than with fibrinolysis (Figure 1).

Figure 1

(A) Efficacy of anticoagulants and antiplatelet agents in primary percutaneous coronary intervention. According to prespecified ischaemic endpoints except for HORIZONS and ATOLL. Data are shown for 30 days for the endpoint of cardiovascular death, MI, and stroke. *300 mg clopidogrel loading followed by 75 mg/day. **600 mg clopidogrel loading and 150 mg for 1 week followed by 75 mg/day for 30 days. ***In PLATO, 35.6% received a 600 mg loading dose pre-randomization to 24 h followed by 75 mg/day.6,7,8,11,15 (B) Bleeding rates with anticoagulants and antiplatelet agents in primary percutaneous coronary intervention. Different definitions for bleeding were used in the different trials. Here TIMI non-CABG major bleeding rates have been used for comparative purposes for all trials in the percutaneous coronary intervention populations. Biv, bivalirudin; Clop, clopidogrel; ENOX, enoxaparin; Pras, prasugrel; Ticag, tricagrelor; UFH, unfractionated heparin; TVR, target vessel revascularization.

Antiplatelet therapy

In CURRENT (Clopidogrel optimal loading dose Usage to Reduce Recurrent EveNTs) in a two by two randomization, 300–325 mg of aspirin was compared with 75–100 mg of aspirin and a 600 mg loading dose of clopidogrel followed by 150 mg/day for 7 days and then 75 mg/day until 30 days compared with a loading dose of 300 mg followed by 75 mg/day for 30 days. In the overall patient group, which included 29.2% with STEMI, there was no difference in the primary endpoint of cardiovascular death, MI, or stroke at 30 days for the aspirin or clopidogrel randomizations. For the post-randomization group of patients undergoing primary PCI (n = 6364), the primary endpoint was similar in patients receiving standard clopidogrel and patients receiving double-dose clopidogrel; 5.0% vs 4.2%, hazard ratio (HR) 0.84, confidence interval (CI) (0.66–1.05), P for interaction = 0.817 (with NSTEMI/unstable angina group). Major bleeding (CURRENT definition) was increased in the overall group of patients undergoing PCI with double dosing clopidogrel; 1.6% vs. 1.1%, P = 0.009.6

Prasugrel, a non-reversible P2Y12 inhibitor, was compared with clopidogrel in 2438 patients undergoing primary PCI in the TRITON (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel) trial. The composite of cardiovascular death, MI, and stroke was similar at 15 months in both groups. In contrast to the overall patient group, where bleeding was significantly increased there was no significant difference in thrombolysis in myocardial infarction (TIMI) non-coronary artery bypass grafting (CABG) major bleeding at 30 days 1.9% clopidogrel vs 3.1% prasugrel, P = 0.11.7

Ticagrelor is a reversible P2Y12 inhibitor with functional recovery of platelets occurring ∼48 h after drug cessation. In the PLATO (Platelet Inhibition and patient Outcomes) trial 7544 patients undergoing primary PCI within 24 h were randomized to receive ticagrelor or clopidogrel [300 mg loading with option to give 600 mg (35.6%)]. Ticagrelor reduced cardiovascular death, MI, or stroke at 1 year by 13%, from 10.8% to 9.4%; HR 0.87, 95% confidence interval (CI) 0.75–1.01, P = 0.07.8 Total mortality was reduced with ticagrelor from 6.1% to 5.0%, P = 0.05. Definite stent thrombosis was reduced from 2.4% to 1.6% with ticagrelor, P = 0.03. Major bleeding as defined by the PLATO and TIMI non-CABG bleeding criteria was similar. However, in the overall trial, TIMI non-CABG major bleeding was significantly increased with ticagrelor, as was fatal intracranial haemorrhage.9


There are no trials of UFH with primary PCI. Fondaparinux has been shown to have no benefit and to be associated with increased catheter thrombosis as compared with UFH.10 Bivalirudin has the advantage over UFH of inhibiting clot-bound thrombin. The HORIZONS (Harmonizing Outcomes With Revascularization and Stents) trial compared primary PCI with UFH [60 IU/kg intravenous bolus and boluses with target activated clotting time (ACT) 200–250 s] plus a glycoprotein IIb/IIIa inhibitor with bivalirudin alone (0.75 mg/kg intravenous bolus followed by 1.75 mg/kg/h infusion).11 In the 3602 patients undergoing primary PCI, bivalirudin, as compared with UFH plus glycoprotein IIb/IIIa inhibitors, reduced the 30-day composite of cardiovascular death, MI, and target vessel revascularization or stroke and major bleeding from 12.1% to 9.2%, P = 0.005, largely due to less major bleeding; 8% vs. 4.9%, P <0.001. There was increased stent thrombosis within 24 h in the bivalirudin group; 1.3% vs. 0.3%, P <0.001; but not at 30 days, 2.5% vs 1.9%, P = 0.30.

Bivalirudin resulted in lower 30-day rates of cardiac death; 2.9% vs. 1.8%, P = 0.03 and death from all causes; 3.1% vs. 2.1%, P = 0.047. Bivalirudin also reduced the non-prespecified rate of CV death, MI, and stroke from 4.7 to 4.1%; P = 0.04 (Gregg Stone, personal communication).

The low molecular weight heparin (LMWH) enoxaparin does not need monitoring and does not induce von Willebrand factor release and platelet activation. In addition enoxaparin inhibits thrombin more than UFH due to a higher anti-Xa to anti-IIa ratio, and increases release of tissue factor pathway inhibitors.

In a meta-analysis of 13 trials comparing bolus administration of LMWHs with UFH in patients undergoing PCI there was no difference in death and MI but there was a 43% reduction in major bleeding, P = 0.002.12 This is in contrast to findings with enoxaparin in non-ST elevation acute coronary syndromes where bleeding was increased in patients undergoing PCI when multiple subcutaneous injections of enoxaparin were given before PCI.13

In FINESSE (Facilitated Intervention with Enhanced Reperfusion Speed to Stop Events), trial centres pre-specified the use of either enoxaparin or UFH in a trial comparing facilitated PCI with abciximab alone or in combination with reteplase with primary PCI with UFH.14 Enoxaparin was administered in a bolus of 0.5 mg/kg to 759 patients and UFH to 1693 patients. In this non-randomized comparison there was lower non-intracranial TIMI major bleeding with enoxaparin; 2.6% vs. UFH 4.4%, P = 0.045. Enoxaparin was associated with lower death, MI, urgent revascularization, or refractory ischaemia at 30 days (5.3% vs. 8.0% UFH, P = 0.0005). All-cause 90-day mortality was also reduced15 with enoxaparin; 3.8% vs. 5.6%, P = 0.046.

The ATOLL (Acute STEMI Treated with primary angioplasty and intravenous enoxaparin Or UFH to Lower ischemic and bleeding events at short- and Long-term follow-up) trial is the first randomized study evaluating enoxaparin in primary PCI. In this trial 910 patients were randomized to either UFH 50–70 IU with a target ACT of 300–350 s and, if IIb/IIIa inhibitors were planned to be given, an ACT of 200–300 s or to intravenous enoxaparin 0.5 mg/kg regardless of whether IIb/IIIa inhibitors were to be given or not. Subcutaneous enoxaparin was recommended to be given to discharge. Of note 70% were randomized pre-hospital.

The trial was statistically underpowered for a realistic 20% reduction and the primary endpoint was negative. At 30 days the primary endpoint of death, MI, procedural failure or non-CABG major bleeding occurred in 28% of patients randomized to enoxaparin and 33.7% of patients randomized to UFH RR 0.83; 95% CI 0.68–1.01; P = 0.006. The main prespecified ischaemia endpoint was death, recurrent MI, acute coronary syndromes (ACS) or urgent revascularization which was reduced from 11.3 to 6.7%; P = 0.01 with enoxaparin. The non-prespecified endpoint of death, MI and revascularization was also decreased; P = 0.04 (Figure 1A). This endpoint is not comparable to the primary endpoint of CV death MI and stroke in other contemporary trials and no statistical adjustment was made for multiple analyses of secondary endpoints. There was no effect on TIMI 3 flow or ST resolution.

Major bleeding according to TIMI non-CABG criteria was similar in both groups; 2.9% enoxaparin vs. 2.4% UFH (Figure 1B). The multiple subcutaneous injections of enoxaparin following PCI may explain the increased bleeding compared with single bolus administration12 and over anticoagulation may have occurred if renal impairment was not recognized, whereas a single bolus regardless of renal function is safe.16 On the other hand the radial approach (67.5%) would be expected to have reduced bleeding by 70%.17

There was a trend for the relatively high 30 day mortality to be reduced; 6% UFH vs. 4% enoxaparin; P = 0.08. This compares to the 3.1% 30 day mortality in patients randomized to UFH plus IIb/IIIa antagonists and 2.1% in patients randomized to bivalirudin in HORIZONS. Enoxaparin did not reduce infarct size or improve TIMI flow and bleeding was not reduced which may have reduced mortality. In the absence of a plausible mechanism, it is likely that the trend for lower mortality with enoxaparin occurred by chance. The endpoint of death, MI or major bleeding was reduced from 15.0% with UFH to 10.0% with enoxaparin; P = 0.03.


Patients undergoing primary PCI require effective antiplatelet and anticoagulant agents that act quickly and reduces ischaemic complications of MI, CV death and stent thrombosis. As bleeding is also related to long-term mortality17 it is desirable to have agents that do not increase bleeding. The ability to give an agent that does not require monitoring is an advantage and cost is an important consideration. The ATOLL trial investigators have shown that enoxaparin is safe in patients undergoing primary PCI and may have a clinically relevant effect in reducing ischaemic complications compared with UFH. They have moved us closer to the goal of further improving the outcomes of patients suffering STEMI.


I would like to thank Charlene Nell, Team Support Administrator, Green Lane Cardiovascular Research Unit, for excellent secretarial assistance.

Conflict of interest: H.W. has received research grants from Sanofi-aventis; Eli Lilly; Medicines Company; NIH; Pfizer; Roche; Johnson & Johnson; Schering Plough; Merck Sharpe & Dohme; AstraZeneca; GlaxoSmithKline; Daiichi-Sankyo Pharma Development; and Bristol-Myers Squibb; and he has received a consultancy fee from Regado Biosciences.


  • The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.


View Abstract