Implementation of ESC/ACC Definition of Myocardial Infarction in Contemporary, Large

RCTs: A Systematic Review

Sergio Leonardi, L. Kristin Newby, E. Magnus Ohman, Paul W. Armstrong.

November 16th 2010

Chicago, IL – AHA Scientific Sessions

Disclosures Information

None of the authors have relevant financial disclosures

Background

Myocardial Infarction (MI) is a key endpoint in RCTs evaluating new therapies

However heterogeneity in MI definition may affect comparisons across RCTs as well as meta-analyses

The 2000 ESC/ACC MI definition 1 consensus recommendations were aimed at resolving this

1: Antman E, Bassand J-P, Klein W, et al. Myocardial infarction redefined -- A consensus document of The Joint European Society of Cardiology/American College of Cardiology committee for the redefinition of myocardial infarction: The Joint European Society of Cardiology/ American College of Cardiology Committee. J Am Coll Cardiol 2000;36:959-69.

Hence, we explored the extent to which they are applied in contemporary, large, cardiovascular RCTs

Methods – Search Criteria

We performed a systematic review of CV RCTs with > 500 patients where MI was part of the primary endpoint initiated after the 2000 ESC/ACC MI redefinition

publication

Search terms included: Acute Coronary Syndrome Myocardial Infarction Coronary Artery Disease Percutaneous Coronary Intervention Coronary Artery By-pass Grafting

Metrics of Guideline Recommendations Adherence

Adherence to 2000 ESC/ACC consensus document was captured using 3 of its key recommendations Use of troponin to define endpoint MI Separate reporting of spontaneous and procedural MI Enzymatic infarct size reporting (i.e., AUC or peak

biomarker value)

We evaluated: % RCTs referencing the 2000 ESC/ACC consensus

document & % of RCTs referencing any consensus document

endorsed by the ACC, AHA, or ESC

Flowchart for Study Screening Process

Time Period Explored :

Sep 1, 2000 to May 5, 2010

Exclusion if any of the following:

1. ≤ 500 pts enrolled

2. MI not part of the primary EP

3. Started before Sep 2000

Summary of RCTs Evaluated

2,729 RCTs screened 134 (5%) met inclusion criteria

Of these 55 (41%) RCTs had primary results including 297,467 pts, 13,526 end-point MIs and a median FU of 9 months (IQR: 1-15.6 months)

9 additional RCTs had design paper published but not primary results (from which MI def’n can be assessed)

MIs contributed a median 40.3% (IQR: 22.9, 61.2) of events in the primary composites, a % that decreased with increasing number of components

Relationship Between Proportion MI Events Within Primary Endpoint and Number of Components

2 Comp

(n=7 RCTs)

3 Comp

(n=28 RCTs)

4 Comp

(n=11 RCTs)

>4 Comp

(n=8 RCTs)

Pro

po

rtio

n o

f M

I eve

nts

wit

hin

th

e p

rim

ary

EP

Index Event At Enrollment into RCTs

STE-ACS (n=18 RCTs)

NSTE-ACS (n=34 RCTs)

Stable CAD (n=32 RCTs)

Other Cond (n=11 RCTs)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Referencing of Consensus Documents in RCTs

55 RCTs with primary results + 9 Only Design = 64 RCTs evaluable. Overall, 31.2% of RCTs (20/64) sourced a consensus document

ESC/ACC 2000 (n=5)Universal MI 2007 (n=5)Other Cons Doc (from AHA,ESC,AHA) (n=10)None (n=44)

Use of Troponin to Define Endpoint MI

12 RCTs (18.7%) had no MI definition published 52 residual RCTs evaluable for troponin use

38.5% (20/52) used Troponin to define MI [66.7% (12/18) among those that referenced a consensus document] Only 1 used troponin for procedural MI 2 used troponin only if CK-MB not available No RCT specified the 99th percentile as the MI

decision limit

Separate Reporting and Infarct Size

Only 1/55 RCT (1.8%) reported separately spontaneous and procedural MI in the primary results

NO RCTs reported infarct size, either by area under the curve of biomarker release or peak values

Conclusions

MI contributes substantially to primary outcome measures in contemporary large RCTs

However, there is surprisingly little implementation

of ESC/ACC recommendations for MI definition and reporting

Appropriate strategies for uniform implementation of the MI endpoint in cardiovascular RCTs appear urgently required

Contribution of MI to Primary Endpoint in RCTs by Revascularization Groups

Group 1: Interventional RCTs

All patients underwent a coronary revascularization (PCI/CABG) either as part of the randomized intervention or as inclusion criterion

Rate of coronary revascularization ≈ 100% Group 2: ACS RCTs

A coronary revascularization could be performed as part of the index enrolling ACS but not required

Median Revascularization rate 62.8% Group 3: Other RCTs

Broad group of RCTs were a coronary revascularization was possible, but not expected

Median Revascularization rate 3.8 %

Supplementary Slide 1

MI Events in RCTs by Revascularization Groups

Interventional RCTs

(N=31 RCTs)

ACS RCTs

(N=13 RCTs)

Other RCTs

(n=11 RCTs)

Pro

po

rtio

n o

f M

I eve

nts

wit

hin

th

e p

rim

ary

EP

Supplementary Slide 2

Use of Troponin to Define MI According to Revascularization Group

Interventional RCTs

ACS RCTs Other RCTs0

5

10

15

20

25

30

35

NO Tn

YES Tn

Adjust. MI Rate in RCTs by Revascularization Groups

Interventional RCTs

(N=31 RCTs)

ACS RCTs

(N=13 RCTs)

Other RCTs

(n=11 RCTs)

MI

%*

N o

f co

mp

on

ents

Supplementary Slide 3

Key features of MI definition in the 10 largest RCTs studied

Supplementary Slide 4