MATRIX Antithrombin

From Wiki Journal Club
Jump to: navigation, search
Valgimigli M, et al. "Bivalirudin or unfractionated heparin in acute coronary syndomes". New England Journal of Medicine. 2015. 373(11):997-1009.
PubMedFull text

Clinical Question

Among patients undergoing PCI for an acute coronary syndrome, is the use of the direct thrombin inhibitor bivalirudin superior to unfractionated heparin with discretionary use of glycoprotein IIb/IIIa inhibitors with respect to ischemic or bleeding outcomes?

Bottom Line

In patients presenting with either STEMI or high risk NSTE-ACS undergoing PCI, the use of bivalirudin was not superior to unfractionated heparin within respect to a net ischemic endpoint of death from any cause, MI, or stroke. There was also no difference in a net clinical benefit endpoint of death from any cause, MI, stroke, target-vessel revascularization, stent thrombosis, or BARC 3/5 bleeding. Bivalirudin was associated with a 2.5% absolute decrease in bleeding endpoints at the cost of a 0.5% absolute increase in stent thrombosis.

Major Points

The optimal agent for anticoagulation in the peri-procedural period for patients with high risk ACS undergoing PCI has been long-debated. The HORIZONS-AMI trial [1] demonstrated a 40% reduction in risk for major bleeding with bivalirudin with a 1% absolute increase in acute stent thrombosis. Similarly, the EUROMAX[2] trial also showed decreased bleeding with bivalirudin at the cost of a slightly higher rate of stent thrombosis. Unfortuantely, these conclusions were limited by trial design which mandated use of IV glycoprotein IIb/IIIa inhibitors in the heparin group only making it unclear whether these differences were driven solely by IIb/IIIa inhibitor effect.

The recent HEAT-PPCI trial sought to clarify this issue in patients presenting with STEMI by reserving IIb/IIIa inhibitor use solely for ischemic complications. This trial demonstrated a 50% relative increase in ischemic complications with bivalirudin when compared to heparin with similar bleeding rates.

The similarly designed 2015 MATRIX Antithrombin study was performed to confirm the validity of these findings in a broader population of patients including those with high risk NSTE-ACS. MATRIX Antithrombin randomized 7,200 patients presenting with either STEMI or high risk NSTE-ACS with plan for urgent PCI to peri-procedural anticoagulation with either unfractionated heparin or bivalirudin. IIb/IIIa use was left to physician discretion in the heparin group and was asymmetric in this trial, with 25.9% use in the heparin group and 4.6% use in the bivalirudin group. Overall MATRIX Antithrombin demonstrated no difference with regard to a primary ischemic endpoint of death from any cause, MI, or stroke at 30 days with bivalirudin when compared to heparin. Bivalirudin was associated with 1% absolute decrease in major bleeding but also a 0.5% absolute increase in stent thrombosis. There was no difference in a net clinical endpoint including both ischemic and bleeding outcomes. Interestingly, there was 30% reduction in all-cause mortality in the bivalirudin group; however, given borderline significance, no difference in net clinical outcomes, and the use of multiple comparisons without statistical adjustment, this finding is possibly attributable to chance and should be interpreted with caution.

MATRIX Antithrombin adds to the growing body of evidence that in the contemporary period when glycoprotein IIb/IIIa use is relatively uncommon, bivalirudin is associated with a modest reduction in bleeding at the risk of higher stent thrombosis rates. This appears to apply in both the STEMI and NSTE-ACS settings. Unfortunately, the conclusions of MATRIX Antithrombin are limited by asymmetric use of IIb/IIIa inhibitors, a limitation observed in several previous trials. However, given an inability to demonstrate superiority in any net clinical endpoint and the potential gravity of increased stent thrombosis (an effect observed more consistently in bivalirudin trials), these findings suggest it may be reasonable to reserve the use of bivalirudin solely for patients intolerant of heparin products. For reasons stated above, the mortality difference observed with bivalirudin is potentially spurious and requires further study before affecting clinical management.


No guidelines have been published reflecting the results of this study.


  • Multicenter, open-label, nested randomized, controlled trial
  • N=7213
    • UFH (n=3603)
    • Bivalirudin(n=3610)
  • Setting: 60 centers in Italy
  • Mean follow-up: 30 days
  • Analysis: Intention-to-treat
  • Corimary outcomes: Death from any cause, MI, or stroke; Death from any cause, MI, stroke, or BARC 3/5 bleeding; Death from any cause, MI, stroke, urgent target-vessel revascularization, definite stent thrombosis, or BARC 3/5 bleeding


Inclusion Criteria

  • STEMI presenting within 12 hours after the onset of symptoms or 12-24 hours after symptom onset with evidence of continuing ischemia or previous fibrinolytic treatment
  • Non-ST-segment elevation ACS with
    • History of new or worsening angina within 7 days
    • High Risk Criteria (2/3 required)
      • Age 60 or older
      • Cardiac biomarker elevation
      • ECG changes compatible with iscehmia
    • Candidate for PCI after completion of coronary angiography

Exclusion Criteria

  • Life expectancy less than 30 days
  • Allergy or intolerance to heparin or bivalirudin
  • Treament with low-molecular-weight heparin within the past 6 hours
  • Treatment with glycoprotein inhibitor within the last 3 days
  • Allergy or contraindication to any study drug or IV contrast
  • Contraindication to coronary angiography
  • CrCl < 30 mL/min or HD dependency
  • Previous enrollment in the study
  • PCI in the previous 30 days

Baseline Characteristics

From the bivalirudin group.

  • Demographics: Age 65.4 years, 75.7% male
  • Comorbidities: BMI 27.2, DM 22.6%, smoker 56%, hyperlipidemia 44.2%, HTN 62.7%, family history of CAD 27.5%, previous MI 14.7%, previous PCI 14.8%, previous CABG 3.5%, previous TIA/stroke 5.0%, PAD 8.2%, COPD 6.0%, renal failure 1.3%, dialysis 0.1%


  • Patients randomly assigned 1:1 to receive bivalirudin or unfractionated heparin
    • Patients assigned to bivalirudin subsequently randomized 1:1 to a post-PCI bivalirudin infusion or no post-PCI infusion as part of nested MATRIX Treatment Duration trial
    • Randomization stratified according to the intended new or ongoing use of P2Y12 inhibitor and type of ACS (STEMI vs. NSTEMI vs. UA)
  • All interventions administered in open label fashion
  • Bivalirudin dosed with 0.75mg/kg bolus followed by 1.75mg/kg infusion
  • Heparin administered at 70-100U/kg in patients not receiving IIb/IIIa inhibitiors and 50-70U/kg in patients receiving IIb/IIIa inhibitors
  • IIb/IIIa inhibitors could be provided before PCI in the heparin group based on treating physician's judgement. IIb/IIIa inhibitors were reserved for bailout ischemic therapy only (no reflow or giant thrombus) in the bivalirudin group.
  • Clinical follow up at 30 days


Comparisons are bivalirudin vs. unfractionated heparin

Coprimary Outcomes

Death from any cause, MI, or stroke
10.3% vs. 10.9%; RR 0.94 95% CI [0.81-1.09]; p=0.44
Death from any cuase, MI, stroke, or BARC 3/5 bleeding
11.2% vs. 12.4%; RR 0.89 95% CI [0.78-1.03]; p=0.12
Death from any cuase, MI, stroke, urgent target-vessel revascularization, definite stent thrombosis, or BARC 3/5 bleeding
11.5% vs. 12.6%; RR 0.90 95% CI [0.79-1.04]; p=0.15

Secondary Outcomes

Death from any cause
1.7% vs. 2.3%; RR 0.71 95% CI [0.51-0.99]; p=0.04
Cardiovascular death
1.6% vs. 2.3%; RR 0.70 95% CI [0.49-0.98]; p=0.04
Urgent target-vessel revascularization
1.5% vs. 1.0%; RR 1.48 95% CI [0.97-2.28]; p=0.07
Definite stent thrombosis
1.0% vs. 0.6%; RR 1.71 95% CI [1.00-2.93]; p=0.048

Subgroup Analyses

  • The effect of bivalirudin versus heparin on the rates of the primary outcomes, death from any cause, and major bleeding was consistent across subgroups.

Adverse Events

Any bleeding
11.0% vs. 13.6%; RR 0.79 95% CI [0.69-0.91]; p=0.001
BARC 3/5 bleeding
1.4% vs. 2.5%; RR 0.55 95% CI [0.39-0.78]; p<0.001


  • Heparin and bivalirudin were administered in open-label fashion, increasing risk of bias in outcomes assessment
  • IIb/IIIa inhibitor use was left to the discretion of the treating physician with marked asymmetry in use. 25.9% of patients randomized to heparin arm and 4.6% of patients randomized to bivalirudin received IIb/IIIa inhibitors. This likely results in overestimation of the bleeding risk observed with heparin and possibly underestimation of the ischemic protection afforded by bivalirudin.
  • Use of multiple comparisons without adjustment
  • Different heparin bolus doses depending on whether a IIb/IIIa inhibitor was administered, introducing confounding by dose in the heparin arm
  • Wide range of acceptable bolus dosing in heparin arm (70-100U/kg if not receiving IIb/IIIa, otherwise 50-70U/kg), raising question of whether increased bleeding observed may be due to inappropriately high dosing rather than medication effect
  • Nested randomized trial design necessitates the use of prolonged post-PCI infusion of bivalirudin in half of the bivalirudin arm, the effects of which are unknown. Given the null result of MATRIX Treatment Duration, this is unlikely to have affected the results


  • Study sponsored by the Italian Society of Invasive Cardiology (GISE), a non-profit organization
  • Grant support from the Medicines company and Terumo Medical (Medicines company provided bivalirudin for the trial)
  • Limited editing for style performed by MedLink Healthcare Communicatinnsos (funded by GISE)

Further Reading

  1. Bivalirudin during primary PCI in acute myocardial infarction. N Engl J Med 2008;358:2218-30.
  2. Bivalirudin started during emergency transport for primary PCI. N Engl J Med 2013;369:2207-17