ANTICOAGULATION AND PCI 
UFH is better 
Jim Nolan 
University Hospital of North Staffordshire
CONFLICTS OF INTEREST
BV IS NOT NEW 
• 1990 Development by Biogen (as Hirulog) 
– Initial angioplasty trial comparing BV and UFH 
published in N Engl J Med 1995;333:764-9 
– Disappointing results 
– 1994 - Decision for no further investment 
• Product licensed to The Medicines Company 
– Re-examination of data
A SLOW START FOR BV 
• 1998 - new application for FDA approval 
– PCI in unstable angina 
– Application rejected 
“Bivalirudin was equivalent but not superior to 
heparin in preventing angioplasty complications” 
• Re-analysis of trial 
• New outcome measures - ‘new results’
A SLOW START FOR BV 
• Dr Clive Meanwell (CEO The Medicines Company) 
"It now turns out that the NEJM paper was plain wrong." 
(Heartwire news 21 Dec 2000) 
• 2000 FDA approval granted 
• 2010 Extension of patent
LAST 14 YEARS 
• Series of RCTs comparing BV v UFH 
• Interpreted to indicate reduced bleeding and 
superior clinical outcomes 
• BV is the antithrombotic agent of choice in PCI 
• But……
ISSUES WITH THE TRIALS 
• Mix efficacy and safety outcome measures 
(composite end point of net clinical benefit) 
• Complex multiarm trial designs 
• Unusual loose bleeding definition invented for 
use in trials/end point modification 
• Trials do not compare BV v UFH 
• Compare BV v TOO MUCH ANTITHROMBOTIC 
THERAPY ( differential GPI or excess heparin 
dosing) leading to mandatory excess of 
bleeding in comparator arm
BLEEDING AFTER PCI AND MORTALITY 
( Mamas et al,HEART 2014,n=533,333)
GIVEN THE CONCERNS RE THE BV 
LITERATURE WHAT DO WE WANT 
• Comparison of BV v UFH in contemporary 
practice 
• Symmetrical use of bailout GPI in both arms 
• Appropriate dosing of UFH ( 70 not 100u/kg) 
• HEAT-PPCI
Primary Safety Outcome Measure 
•Major bleeding - 
• Type 3-5 bleeding as per BARC definitions
1917 patients scheduled for emergency angiography 
29 (1.5%) already randomised in the trial 
59 (3.0%) met one or more other exclusion criteria 
1829 eligible for recruitment 
1829 Randomised 
Representative ‘Real-World’ Population
Characteristic Bivalirudin Heparin 
Median age (years) 62.9 63.6 
Female sex (%) 28.5 26.9 
Caucasian race (%) 95.8 95.9 
Diabetes mellitus (%) 12.6 15.1 
Previous MI (%) 13.5 10.3 
eGFR (ml/min/1.73m2) 80.0 80.0 
Haemoglobin (g/dl) 13.6 13.7
Characteristic Bivalirudin (%) Heparin (%) 
P2Y12 use - Any 99.6 99.5 
- Clopidogrel 11.8 10.0 
- Prasugrel 27.3 27.6 
- Ticagrelor 61.2 62.7 
GPI use 13.5 15.5 
Radial arterial access 80.3 82.0 
PCI performed 83.0 81.6
Characteristic Bivalirudin (%) Heparin (%) 
Thrombectomy 59.1 57.6 
Single vessel Tx 93.2 90.3 
Any stent implant 92.8 92.2 
DES implantation 79.8 79.9 
TIMI III flow - post PCI 93.3 92.7
Event curve shows first event experienced
Bivalirudin Heparin 
n % % n 
MACE 79 8.7 % v 5.7 % 52 
Absolute risk increase = 3.0% (95% CI 0.6, 5.4) 
Relative risk = 1.52 (95% CI 1.1 – 2.1) P=0.01
ARC definite or probable stent thrombosis events 
Bivalirudin Heparin 
n % % n 
All Events 24 3.4 % v 0.9 % 6 
Relative risk = 3.91 (95% CI 1.6 - 9.5) P=0.001
Major Bleed BARC grade 3-5 Minor Bleed BARC grade 2 
Bivalirudin Heparin 
n % % n 
Minor Bleed 83 9.2 % v 10.8 % 98 
Major or Minor 113 12.5 % v 13.5 % 122 
Minor Bleed P=0.25 Major or Minor P=0.54
Subgroup Relative Risk (95% CI) 
P Value for 
interaction 
All patients 1.52 (1.09, 2.13) 
Arterial access site 0.87 
Radial 1.58 (1.01, 2.48) 
Femoral 1.45 (0.70, 2.98) 
Diabetes 0.35 
Yes 2.22 (1.04, 4.76) 
No 1.54 (1.04, 2.28) 
Age 0.11 
≥75 1.09 (0.68, 1.77) 
<75 1.97 (1.23, 3.16) 
Favours Bivalirudin Favours Heparin 
1
Subgroup Relative Risk (95% CI) 
P Value for 
interaction 
P2Y12 agent used 0.78 
Clopidogrel 1.34 (0.54, 3.31) 
Prasugrel 1.91 (0.87, 4.21) 
Ticagrelor 1.41 (0.93, 2.14) 
Left Ventricular Function Impaired 0.67 
Yes 1.28 (0.84, 1.95) 
No 1.63 (0.64, 4.16) 
PCI attempted 0.88 
Yes 1.55 (1.06, 2.28) 
No 1.45 (0.71, 2.96) 
Favours Bivalirudin Favours Heparin 
2
HEAT INDICATES THAT WHEN 
UFH IS COMPARED WITH BV 
USING AN APPROPRIATE STUDY 
DESIGN UFH IS BETTER
WHAT ABOUT THE REST OF THE 
LITERATURE?
LANCET META-ANALYSIS 2014 
• Cavender, Sabatine Lancet 2014; 384: 599-606 
• 16 RCTs 33 958 patients ( Includes HEAT and 
BRIGHT ) 
• 2422 patients with MACE 1406 with major 
bleed
REDUCED MACE WITH UFH (ALL CLINICAL SYNDROMES) 
MACE 
Death 
MI 
I-D Revasc 
Stent Thrombosis 
RR = 1.09 (1.01-1.17) p = 0.02 
Acute 
Sub Acute 
0.4 0.8 1 1.25 2.5 5 
Favours Bivalirudin Favours Heparin
BLEEDING EVENTS
Major Bleeding with Equivalent Provisional GPI use 
ISAR REACT 3 
ARMYDA-7 
BRIGHT (UFH) 
HEAT-PPCI 
NAPLES III 
Overall 
0.78 (0.51-1.19) 
Favours Bivalirudin Favours Heparin
Bleeding Risk is Affected by 
Heparin Dose
Major Bleeding in Trials with Provisional GPI use 
ISAR REACT 3 
ARMYDA-7 
BRIGHT (UFH) 
HEAT-PPCI 
NAPLES III 
Overall 
0.78 (0.51-1.19) 
Favours Bivalirudin Favours Heparin
Conclusions 
• UFH (50 - 70 u/Kg) is the drug of choice 
• BV has no advantage in MACE reduction or 
reducing bleeding risk 
• Potential for annual global savings 
– US $ 700 million 
• The ‘bivalirudin story’ is interesting 
? An example of “bad pharma” 
• ESC Revascularisation Guidelines 2014 
recommend UFH in preference to BV for PCI

Nolan J - AIMRADIAL 2014 - Heparin is better

  • 1.
    ANTICOAGULATION AND PCI UFH is better Jim Nolan University Hospital of North Staffordshire
  • 2.
  • 3.
    BV IS NOTNEW • 1990 Development by Biogen (as Hirulog) – Initial angioplasty trial comparing BV and UFH published in N Engl J Med 1995;333:764-9 – Disappointing results – 1994 - Decision for no further investment • Product licensed to The Medicines Company – Re-examination of data
  • 4.
    A SLOW STARTFOR BV • 1998 - new application for FDA approval – PCI in unstable angina – Application rejected “Bivalirudin was equivalent but not superior to heparin in preventing angioplasty complications” • Re-analysis of trial • New outcome measures - ‘new results’
  • 5.
    A SLOW STARTFOR BV • Dr Clive Meanwell (CEO The Medicines Company) "It now turns out that the NEJM paper was plain wrong." (Heartwire news 21 Dec 2000) • 2000 FDA approval granted • 2010 Extension of patent
  • 6.
    LAST 14 YEARS • Series of RCTs comparing BV v UFH • Interpreted to indicate reduced bleeding and superior clinical outcomes • BV is the antithrombotic agent of choice in PCI • But……
  • 7.
    ISSUES WITH THETRIALS • Mix efficacy and safety outcome measures (composite end point of net clinical benefit) • Complex multiarm trial designs • Unusual loose bleeding definition invented for use in trials/end point modification • Trials do not compare BV v UFH • Compare BV v TOO MUCH ANTITHROMBOTIC THERAPY ( differential GPI or excess heparin dosing) leading to mandatory excess of bleeding in comparator arm
  • 8.
    BLEEDING AFTER PCIAND MORTALITY ( Mamas et al,HEART 2014,n=533,333)
  • 9.
    GIVEN THE CONCERNSRE THE BV LITERATURE WHAT DO WE WANT • Comparison of BV v UFH in contemporary practice • Symmetrical use of bailout GPI in both arms • Appropriate dosing of UFH ( 70 not 100u/kg) • HEAT-PPCI
  • 12.
    Primary Safety OutcomeMeasure •Major bleeding - • Type 3-5 bleeding as per BARC definitions
  • 13.
    1917 patients scheduledfor emergency angiography 29 (1.5%) already randomised in the trial 59 (3.0%) met one or more other exclusion criteria 1829 eligible for recruitment 1829 Randomised Representative ‘Real-World’ Population
  • 14.
    Characteristic Bivalirudin Heparin Median age (years) 62.9 63.6 Female sex (%) 28.5 26.9 Caucasian race (%) 95.8 95.9 Diabetes mellitus (%) 12.6 15.1 Previous MI (%) 13.5 10.3 eGFR (ml/min/1.73m2) 80.0 80.0 Haemoglobin (g/dl) 13.6 13.7
  • 15.
    Characteristic Bivalirudin (%)Heparin (%) P2Y12 use - Any 99.6 99.5 - Clopidogrel 11.8 10.0 - Prasugrel 27.3 27.6 - Ticagrelor 61.2 62.7 GPI use 13.5 15.5 Radial arterial access 80.3 82.0 PCI performed 83.0 81.6
  • 16.
    Characteristic Bivalirudin (%)Heparin (%) Thrombectomy 59.1 57.6 Single vessel Tx 93.2 90.3 Any stent implant 92.8 92.2 DES implantation 79.8 79.9 TIMI III flow - post PCI 93.3 92.7
  • 17.
    Event curve showsfirst event experienced
  • 18.
    Bivalirudin Heparin n% % n MACE 79 8.7 % v 5.7 % 52 Absolute risk increase = 3.0% (95% CI 0.6, 5.4) Relative risk = 1.52 (95% CI 1.1 – 2.1) P=0.01
  • 19.
    ARC definite orprobable stent thrombosis events Bivalirudin Heparin n % % n All Events 24 3.4 % v 0.9 % 6 Relative risk = 3.91 (95% CI 1.6 - 9.5) P=0.001
  • 20.
    Major Bleed BARCgrade 3-5 Minor Bleed BARC grade 2 Bivalirudin Heparin n % % n Minor Bleed 83 9.2 % v 10.8 % 98 Major or Minor 113 12.5 % v 13.5 % 122 Minor Bleed P=0.25 Major or Minor P=0.54
  • 21.
    Subgroup Relative Risk(95% CI) P Value for interaction All patients 1.52 (1.09, 2.13) Arterial access site 0.87 Radial 1.58 (1.01, 2.48) Femoral 1.45 (0.70, 2.98) Diabetes 0.35 Yes 2.22 (1.04, 4.76) No 1.54 (1.04, 2.28) Age 0.11 ≥75 1.09 (0.68, 1.77) <75 1.97 (1.23, 3.16) Favours Bivalirudin Favours Heparin 1
  • 22.
    Subgroup Relative Risk(95% CI) P Value for interaction P2Y12 agent used 0.78 Clopidogrel 1.34 (0.54, 3.31) Prasugrel 1.91 (0.87, 4.21) Ticagrelor 1.41 (0.93, 2.14) Left Ventricular Function Impaired 0.67 Yes 1.28 (0.84, 1.95) No 1.63 (0.64, 4.16) PCI attempted 0.88 Yes 1.55 (1.06, 2.28) No 1.45 (0.71, 2.96) Favours Bivalirudin Favours Heparin 2
  • 23.
    HEAT INDICATES THATWHEN UFH IS COMPARED WITH BV USING AN APPROPRIATE STUDY DESIGN UFH IS BETTER
  • 24.
    WHAT ABOUT THEREST OF THE LITERATURE?
  • 25.
    LANCET META-ANALYSIS 2014 • Cavender, Sabatine Lancet 2014; 384: 599-606 • 16 RCTs 33 958 patients ( Includes HEAT and BRIGHT ) • 2422 patients with MACE 1406 with major bleed
  • 26.
    REDUCED MACE WITHUFH (ALL CLINICAL SYNDROMES) MACE Death MI I-D Revasc Stent Thrombosis RR = 1.09 (1.01-1.17) p = 0.02 Acute Sub Acute 0.4 0.8 1 1.25 2.5 5 Favours Bivalirudin Favours Heparin
  • 27.
  • 28.
    Major Bleeding withEquivalent Provisional GPI use ISAR REACT 3 ARMYDA-7 BRIGHT (UFH) HEAT-PPCI NAPLES III Overall 0.78 (0.51-1.19) Favours Bivalirudin Favours Heparin
  • 29.
    Bleeding Risk isAffected by Heparin Dose
  • 30.
    Major Bleeding inTrials with Provisional GPI use ISAR REACT 3 ARMYDA-7 BRIGHT (UFH) HEAT-PPCI NAPLES III Overall 0.78 (0.51-1.19) Favours Bivalirudin Favours Heparin
  • 31.
    Conclusions • UFH(50 - 70 u/Kg) is the drug of choice • BV has no advantage in MACE reduction or reducing bleeding risk • Potential for annual global savings – US $ 700 million • The ‘bivalirudin story’ is interesting ? An example of “bad pharma” • ESC Revascularisation Guidelines 2014 recommend UFH in preference to BV for PCI

Editor's Notes

  • #11 This is a single-centre randomised controlled trial which recruited patients for 22 months in 2012 and 2013. It compared bivalirudin and unfractionated heparin in STEMI patients, randomised at presentation – in their acute phase management with Primary PCI
  • #12 We used abciximab as our GPI agent. This was reserved for bailout use in both groups - in accordance with the indications described in the ESC guidelines
  • #13 The primary safety outcome measure was the rate of major bleeding, defined as BARC ‘Type 3 – 5’ events
  • #14 All of these patients were randomised – to create an unselected ‘Real-World’ population
  • #15 Baseline characteristics were well matched between the two groups. The population was predominantly Caucasian: The median age was 63 and about 14% were diabetics
  • #16 Dual antiplatelet therapy was essentially universal GPI use was similar with 13 – 15 % use in both groups Most procedures were performed with radial access and PCI was performed in over 80% of cases
  • #17 Of the Cases managed with PCI – the key aspects of procedure performance matched institutional and national norms and were well-matched between the groups.
  • #18 The event curves demonstrate an early separation reflecting the substantial early hazard in the PPCI setting.
  • #19 MACE events were significantly more common in the Bivalirudin arm at 8.7% compared to 5.7% in patients randomised to heparin - an absolute risk increase of 3% The relative risk of an event with bivalirudin was 1.52
  • #20 This difference is driven by an four-fold increase in the rate of stent thrombosis observed with bivalirudin therapy.
  • #21 Rates of minor bleeding – and rate of combined major or minor bleeding were also similar
  • #22 We had a number of pre-specified subgroups. This figure shows – for each group - the relative risk of a primary outcome MACE event and the associated 95% confidence intervals. Results to the right of the line of unity favour heparin. We see consistent heparin advantage – on this slide 0 demonstrated for the arterial access site (radial or femoral): For the presence or absence of treated diabetes and for age with a dichotomous cut off at 75 years.
  • #23 The effect is also independent of the oral P2Y12 inhibitor used and the subsequent left ventricular function. Interestingly, heparin advantage is also seen in patients who were not managed with immediate PCI.