Frans Van de Werf, MD, PhD
University of Leuven, Belgium
Risk Stratification of ACS Patients
Which type of ACS patients are we
talking about to day?
4/14/2011
STEMI and NSTEMI in the NRMI registry from 1990 to 2006
and % in whom a troponin assay was used to diagnose MI
Roger, V. L. et al. Trends in incidence, severity, and outcome of hospitalized myocardial infarction.
Circulation 121, 863–869 (2010)
,
Trends in incidence of hospitalized MI from 1987 to 2006 in
OlmstedCounty, MN, USA, by ST-segment elevation status
Roger, V. L. et al. Trends in incidence, severity, and outcome of hospitalized myocardial infarction.
Circulation 121, 863–869 (2010)
Trends in incidence of hospitalized MI from 1987 to 2006 in
OlmstedCounty, MN, USA, by ST-segment elevation status
Roger, V. L. et al. Trends in incidence, severity, and outcome of hospitalized myocardial infarction.
Circulation 121, 863–869 (2010)
Age-adjusted and sex-adjusted incidence of acute MI in
KaiserPermanente, Northern California, USA,
from 1999 to 2008
Yeh, R. W. et al. Population trends in the incidence and outcomes of acute myocardial infarction. N. Engl. J.
Med. 362, 2155–2165 (2010)
Copyright ©2010 American Heart Association Roger, V. L. et al. Circulation 2010;121:863-869
30-day case fatality rates for hospitalized MI overall and by age, sex, and time period
.Long-term survival among 30-day survivors did not improve!
.Causes of death shifted from CV to non-CV in most recent
year quartile (50% non-CV)
BUT
What is a myocardial infarction nowadays ?
4/14/2011
Copyright ©2011 American College of Cardiology Foundation. Restrictions may apply.
Lim, C. C. S. et al. J Am Coll Cardiol 2011;57:653-661
Flow Diagram of
Patient Recruitment
Copyright ©2011 American College of Cardiology Foundation. Restrictions may apply.
Lim, C. C. S. et al. J Am Coll Cardiol 2011;57:653-661
Peak CK-MB and Troponin Values
Copyright ©2011 American College of Cardiology Foundation. Restrictions may apply.
Lim, C. C. S. et al. J Am Coll Cardiol 2011;57:653-661
PMI Groups
Copyright ©2011 American College of Cardiology Foundation. Restrictions may apply.
Lim, C. C. S. et al. J Am Coll Cardiol 2011;57:653-661
Percentage Change From Baseline of CRP,
SAA, MPO, and TNF-Alpha Levels
The classical risk models
4/14/2011
Risk Models and Risk Scores
GUSTO-I Model for 30-Day Mortality in STEMI
TIMI Risk Score for STEMI and NonSTEMI
GRACE Risk Model for 6 Months Outcome in
ACS (STEMI , NonSTEMI and UA)
GUSTO-I : Independent Clinical
Predictors of 30-Day Mortality*
Variable
Age, y
Systolic BP, mm Hg
Killip class
Heart rate, bpm
Location of infarction
-----------------------------------
Previous infarction
Age-by-Killip class interaction
Height, cm
Time to treatment, h
Diabetes
Weight, kg
Smoking
Choice of thrombolytic therapy
Previous bypass surgery
Hypertension
Prior cerebrovascular disease
Adjusted c2
717
550
350 (3 df)
275 (2 df)
143 (2 df)
--------------
64
29
31 (4 df)
23
21
16
22 (2 df)
15 (3 df)
16
14
10
Lee et al. Circulation 1995
*Indicates the independent
contribution of each variable
after adjustment for all other
factors in the list. The first 10
factors are significant with
P<0.00001; the next four P
<0.0001; the last two P<0.01. Sex
(P=0.043) and US enrollment
(P=0.047) were marginal
predictors.
Lee, K. L. et al. Circulation 1995;91:1659-1668
Predictors of 30 Day Mortality in
> 40 000 STEMI patients (GUSTO)
GUSTO-I : Observed 30-Day Mortality vs 30-Day
Mortality Predicted by Regression Model
Lee et al.
Circulation
1995
GRACE: Predictors of 6 Month Mortality
4/14/2011
Predictors
Age (per 10 year increase)
Medical History
• Congestive heart failure
• Hypertension
• Peripheral vascular disease
• PCI
Χ2
505.7
34.2
8.8
21.8
8.3
HR (95%CI)
1.8 (1.68 to 1.84)
1.5 (1.32 to 1.73)
1.2 (1.05 to 1.33)
1.4 (1.21 to 1.62)
0.8 (0.64 to 0.93)
GRACE: Predictors of 6 Month Mortality
Predictors
Presentation characteristics
• Pulse (per 30 bpm ↑)
• Systolic blood pressure (20 mmHg ↓)
• Killip class
• Initial serum creatinine (per 88 µmol/l ↑)
• Initial cardiac markers or enzymes
• Cardiac arrest
Findings on electrocardiography
• ST segment deviation
• LBBB
• No of leads with ST segment deviation
Χ2
44.3
152
142.8
135.3
63.0
58.5
46.8
10.0
20.1
HR (95%CI)
1.2 (1.16 to 1.31)
1.2 (1.22 to 1.30)
1.5 (1.41 to 1.62)
1.2 (1.19 to 1.29)
1.6 (1.42 to 1.78)
2.6 (2.00 to 3.32)
1.6 (1.41 to 1.88)
1.3 (1.10 to 1.60)
1.2 (1.10 to 1.33)
Simplified GRACE Risk Model for
Death in ACS at 6 Months
Simplified GRACE Risk Model for
Death/MI in ACS at 6 Months
GRACE Risk Nomogram for Death
at 6 Months (simplified model)
GRACE Risk Nomogram for Death/MI
at 6 Months (simplified model)
GRACE Risk Nomogram for Death
at 6 Months (Simplified Model)
GRACE Risk Nomogram for Death/MI
at 6 Months (Simplified Model)
26
27
GRACE PDA Software
TIMI Risk Score: Independent Predictors of
30-Day Mortality in InTIME-II
Morrow et al.
Circulation
2000
TIMI Risk Score for STEMI
Morrow et al.
Circulation
2000
TIMI Risk Score for UA/Non-STEMI
TIMI Risk Score and all-cause mortality, MI, and severe recurrent ischemia
calculated for enoxaparin and UFH in TIMI11B and ESSENCE
Antman, E. M. et al. JAMA 2000;284:835-842
A recent case:Women, 66 y, chest pain, troponins negative, cholesterol : 266 mg/dl
ST segment depression in II, III, aVF and V6
enrolled in a NSTE-ACS study, coronary angiography: negative!
4/14/2011
Event rates after an ACS
4/14/2011
Cumulative death rates in 3721 ACS patients
from UK and Belgium at ± 5 year (GRACE)
0
5
10
15
20
25
STEMI Non-STEMI UA
% CV
CV
CV
n=1403 n=1107 n=850
65% after
discharge
83% after
discharge
95% after
discharge
14%
16%
13%
19%
TOTAL
22%
TOTAL
17%
TOTAL
Fox K et al. Eur Heart J 2010
Who Can/Should Use
the Risk Models/Scores ?
Risk models/scores can/should be used by clinicians to :
for triage decisions
To determine risk of an adverse event or
co-morbidity
to delineate treatment options
Risk models/scores can/should be used by guidelines committees to :
to determine relative treatment benefit/harm of certain therapies in different risk categories
Do (Good !) Clinicians Really need these
Risk Models/ Scores ?
78-year-old female patient with 6 mm ST-
segment elevation in anterior leads, history
of hypertension, type 2 DM, renal failure
(serum creatinine of 1.8 mg/dl), never
smoked, Killip class II on admission
vs
42-year-old male patient with 1 mm ST-
segment elevation in inferior leads, no
hypertension, no DM, heavy smoker (1.5
packet /day), Killip class I on admission
Estimated
30 day mortality
20% or 30%
Estimated
30 day mortality
1% or 2%
Do Guidelines Committees Use
these Risk Models/Scores?
Yes,to a certain extent but the usual
conclusion is that there is
“uncertainty” and that new
prospective and randomized studies
are needed in specific risk categories
such as eg the elderly, diabetics,
renal failure , hypertensive pts etc
Conclusions Risk models/scores for cardiac risk stratification in
ACS patients perform well in populations similar to the one from which they were obtained
Risk models/scores for mortality alone provide better discrimination than those for a composite endpoint
Risk models/scores tend to underestimate the risk in patients who did not participate in randomizedtrials
Risk models/scores for longterm outcome are lacking
The usefulness of these risk models/scores in daily clinical practice has not been demonstrated