Allan M' Ross, MD FACC FAHA

1
Allan M. Ross, MD FACC FAHA

• Professor Emeritus of Medicine, George Washington
  University
• Washington, DC, USA

2
Rescue PCI in STEMI Outcomes and
Evidence Revisited
ACC Lake Louise, March 2008
Allan M Ross
3
Outcome After Rescue PCI and/or Late PCI in
STEMI Patients
ACC Lake Louise, March 2008
Allan M Ross
4
Wijeysundera, H et al JACC 2007 v49 p 422
Abbate, A et al JACC march 4 2008
p956
5
Outcome After Rescue PCI and/or Late PCI in
STEMI Patients
They Are Both a Matter of Time and Flow
6
The Principal Determinants of Outcome Following
Reperfusion Therapy in Acute MI Are
-Magnitude of IRA Patency Restoration and -Speed
of Patency Restoration
GUSTO - PATENCY
GISSI - TIME
mortality REDUCTION
30 d MORTALITY
7
Decreasing Mortality Benefit with Increasing
delay from STEMI Symptom onset to Fibrinolytic
Therapy
REASONS
RR. gt7.5/30min
n1791
Decreasing rate of lysis as thrombus ages, cross
links, etc.
Decreasing quantity of ischemic but salvageable
myocardium
8
Decreasing Mortality Benefit with Increasing
delay from STEMI Symptom onset to Any Reperfusion
Therapy
REASONS30min
12 mos Mortality With Primary PCI as a Function
of Delay from Symptom onset
RR. gt7.5/30min
Decreasing rate of lysis as thrombus ages, cross
links, etc.
n1791
Decreasing quantity of ischemic but salvageable
myocardium
De Luca et al, Circ 1091223-1225
9
Wijeysundera, H et al JACC 2007 v49 p 422
In summary, this meta-analysis of randomized
trials lends support to the use of rescue PCI for
failed fibrinolytic therapy in patients with
STEMI..
10
The Immediate Response to this Publication
ACC Cardiosurce CME online (.the meta
analysis..) supports the recommendation of
rescue PCI as the treatment of choice for STEMI
patients who fail to reperfuse with fibrinolytic
therapy.
The Heart.org Rescue PCI confirmed as best
option after failed thrombolysis A new
meta-analysis shows a reduced risk of re-MI and
heart failure and a favorable trend in mortality
in patients undergoing rescue PCI vs those
treated conservatively
11
Do meta-analyses deserve the persuasive power
they are afforded?
12
(My) Hierarchy of Publication Types leading to a
reasonably firm clinical conclusions
Large Randomized Trial(S)
Multiple concordant smaller randomized
Trials Very Large Registries Pooled
analyses Meta-analyses Expert Opinions
13
Small trials usually used in meta analyses (or
pooled analyses) almost invariably suffer from
all components of publication bias (authors,
sponsors, and publishers) and often are
subsequently recognized as having been incorrect.
14
Funnel Plot
Published Reports
Unpublished Reports
15
Meta-analysis of RCTs of GIK vs Control in AMI
Create-ECLA Investigators AHJ 2004
GIK n/N
CONTROLn/N
STUDY
High Dose Trials
Low Dose Trials
0
1
2
3
4
Favors GIK Favors controls
16
The CREATE-ECLA Randomized Trial of GIK
n20,195 JAMA 2005
 This Article
 Citing Articles
 Related Content
 
Full text
 
PDF
 
Send to a friend
 
Save in My Folder
 
Save to citation manager
 
Permissions
 
Citation map
 
Citing articles on HighWire
 
Contact me when this article is cited
 
Related letters
 
Related articles
 
Similar articles in JAMA
0.10
30 Day Mortality
0.08
0.06
Cumulative Hazard Rate
Hazard Ratio 1.03 95 CI 0.95-1.13 P0.45
0.04
0.02
0.0
0
5
10
15
20
25
30
Day
17
Wijeysundera, H et al JACC 2007 v49 p 422
Mortality
P0.09
18
Wijeysundera, H et al JACC 2007 v49 p 422
Heart Failure There was no significant
difference in its incidence between PCI and non
interventional therapy after failed thrombolysis
Reinfarction Data on this post MI complication
was available in only three of the trials used in
the meta-analysis. An advantage for rescue was
seen in only one of the three (REACT)
19
Wijeysundera, H et al JACC 2007 v49 p 422
20
Rescue angioplasty during myocardial infarction
has a beneficial effect on mortality a tenable
hypothesis.
Canadian Journal of Cardiology 199283855
Belenkie I Traboulsi M Hall CA Hansen JL Roth
DL Manyari D Filipchuck NG Schnurr L et al
21
Merlin the first randomized rescue trial with at
least a modest sample size (n307)

Sutton et al, JACC 2004
p.89
Dx of lytic failure was lack of ST seg resolution
60 min after lytic given Time from pain onset to
rescue was 5.4 hours
11
10
22
Merlin the first randomized rescue trial with at
least a modest sample size (n307)

Sutton et al, JACC 2004
RESCUE CONSERVATIVE
Transfusion 17 1
Stroke 5 1
Reinfarction 7 10
CHF 24 30 (Re-)PCI
7 20
Conclusion The benefits of rescue are small, the
principal favorable effect being reduced
subsequent revascularizations No effect
on survival or follow up ventricular function
23
REACT n427 Conservative Rx vs Repeat Lysis vs
Rescue PCI Time, Pain-first Lytic 145 min
First
Lytic SK 56
SK as Rescue Lytic 58
failed lysis Lack of ST Res.
at 90 min
Time, Lysis 1-Rescue 270 min
(2/3 of pts 1st admitted to a PCI capable hosp.)
Time, Pain onset to rescue 6.9
hours
Gershlick A et al. NEJM 2005
The 6 month endpoint was a combination of death,
ReMI, Stroke, and CHF. The only significant
difference between the groups was ReMI, 2 with
rescue, 9 with the other strategies but that
drove the combined endpoint to significance
plt.01
This trial (surprisingly) was generally perceived
as a strong endorsement of routine Rescue
24
Merlin the trial has now published a 3 year
follow up

Sutton et al, AHJ May 2007
N307
Dx of lytic failure was lack of ST seg resolution
60 min
Long term follow up for survival, rescue v
conservative Rx
3 year mortality
25
Heretofore there has been a visceral feeling
(like mine was) amongst interventionalists, and
perhaps most cardiologists that providing rescue
for failed lytic patients is logical,
intuitively effective and evidence supported. On
the other hand..
26
It is well documented and well accepted that in
STEMI patients the benefits of early reperfusion
are considerable, but then decreasing with
prolonged delay to reperfusion. Many Rescue PCIs
occur after a quite considerable delay

• Relationship Among the Duration of Symptoms of
  Acute MI Before Reperfusion Therapy, Mortality
  Reduction, and Extent of Myocardial Salvage

Gersh et al.
27
(My synthesis) the evidence suggests that the
benefits of rescue will be disappointing unless
performed very quickly after suspicion of failed
lysis is raised (absent ST segment resolution
possibly at 60 rather than 90 min) and in
patients who clearly are at elevated risk for a
poor outcome
28
AHA/ACC Guidelines for Management of Acute STEMI
(2004)
29
(No Transcript)
30
2007 Focused Update of the ACC/AHA/SCAI 2005
Guideline Update for Percutaneous Coronary
Intervention
RESCUE PCI it might be reasonable to select
moderate and high-risk patients for rescue PCI
after suspected failure of fibrinolysis (on the
basis of ST segment criteria) .
31
2007 Focused Update of the AHA/ACC 2004
Guidelines for the Management of Patients With
ST-Elevation Myocardial Infarction
A strategy of coronary angiography with intent
to perform PCI (or emergency coronary artery
bypass graft surgery) is recommended for patients
who have received fibrinolytic therapy and have
any of the following a) cardiogenic shock in
patients lt75 years who are suitable candidates
for revascularization, b) severe congestive heart
failure and/or pulmonary edema (Killip class
III), or c) hemodynamically compromising
ventricular arrhythmias.
32
Routine systematic rescue for suspected failed
fibrinolysis?
33
Abbate, A et al JACC march 4 2008
p956
Conclusion PCI of the IRA late (12 hours to 60
days) is associated with significant improvements
in cardiac function and survival.
34
Abbate, A et al JACC march 4 2008
theHeart.org A new meta-analysis of trials
comparing PCI with medical therapy in patients
randomized later than 12 hours after MI has shown
a significant benefit in cardiac function and
mortality in the PCI group
TCTMD
Late PCI After Infarction Still Improves Survival
Over Medical Therapy
35
Death or non-fatal recurrent MI, or Stroke
P.03
p.22
p.22
Abbate et al JACC March 4 2008
36
8-48 days, 12-48 hrs, gt48 hrs, gt24 hrs, 3-28
days, 12h-14 days, 3-58 days, gt4 days, 5-42 days
37
SURVIVAL
p.03
P.03
38
SURVIVAL
p.03
39
Change in Ejection Fraction, Baseline to
Follow-up
Plt.01
Where is the TOAT trial?
40
The OAT trial (also Tosca 2) 2006 pts 3-28 days
post STEMI PCI v Conservative management
41
Brave 2 Comparison of Late PCI to Conservative
therapy Schomig A et al. JAMA 2005 n 364.
Median time, Symptom onset to PCI 23 hours.
Baseline EF (SPECT) 50. Final Infarct size 8
v 13 (plt.001) No other significant outcome
differences (mortality, recurrent MI, CVA)
between groups
42
Randomized Comparison of Percutaneous
Transluminal Coronary Angioplasty and Medical
Therapy in Stable Survivors of Acute Myocardial
Infarction With Single Vessel Disease A Study of
the Arbeitsgemeinschaft Leitende Kardiologische
Krankenhausärzte (ALKK) n300 Uwe Zeymer,
Rainer Uebis, Albrecht Vogt, Hans-Georg Glunz,
Hans-Friedrich Vöhringer, Dietrich Harmjanz, and
Karl-Ludwig Neuhaus
Primary end point was the survival free of
reinfarction, (re)intervention, coronary artery
bypass surgery, or readmission for severe angina
pectoris at 1 year. The event-free survival at 1
year was 82 in the medical group and 90 in the
angioplasty group (P0.06). This difference was
driven by the need for a later (re)intervention
(20 vs 8, plt.03)
43
SURVIVAL
44
Change in Ejection Fraction, Baseline to
Follow-up
45
The SWISSI II Randomized Controlled Trial Erne P
et al. JAMA. 2007297(18)1985-1991.
Effects of Percutaneous Coronary Interventions in
Silent Ischemia After Myocardial Infarction
Demonstrated by symptom limited Max. ETT with
imaging that documented persistent ischemia
Within the previous 3 months
46
JACC v 40 5 2002
Conclusions PCI 1 month post STEMI had an
adverse effect on remodeling but tended to be
associated with an increase in exercise
tolerance and QOL (as measured by a single
((unblinded)) self-administered questioaire)
Yousef ZR et al.
JACC v40 5
47
CONCLUSION The available evidence does not
support endorsement of routine Rescue nor a
clear cut benefit for routine late PCI (beyond
the period of infarct size reduction) in stable
STEMI patients. One could say these approaches
are not well grounded..
48
(No Transcript)
49
CONCLUSIONS (2) The available evidence does not
support endorsement of routine Rescue nor a
clear cut benefit for late PCI (beyond the
period of infarct size reduction) in stable STEMI
patients. There is however support for selective
use of both approaches in several clinical
circumstances including hemodynamic instability,
definite evidence for residual ischemia, and
other clinical markers of a poor prognosis.
50
Thank you
51
(No Transcript)
52
(No Transcript)
53
DECOPI
Steg et al. EHJ 2004
Probability of survival free of non fatal MI or
ventricular arrhythmias
CONCLUSIONS Systematic late angioplasty of the
infarct vessel was associated with a higher LVEF
at six months, no difference in clinical outcomes
and higher costs than medical therapy
54
Two Questions Regarding Endpoints In STEMI Trials
In trials comparing Rescue PCI for one group with
any conservative treatment strategy in the other
(B) is it reasonable to include future need for a
PCI in group B as an endpoint equivalent to
death, CHF or CVA , or another MI etc?
Why do STEMI patients who after rescue have an
open IRA have a lower frequency of reinfarction
than those left with a closed one.ever happened
to double jeapordy?
55
While initially believing as did most that rescue
could be an effective response to failed
fibrinolyis I had the opportunity to acquire some
personal experience performing these procedures
while clinging to with a visceral belief that it
could be generally effective.
56
The earliest randomized trials of Rescue,
Belinke, the RECUE trial and TAMI had fewer than
100 pts each. Then came MERLIN and most recently
REACT
57
Rescue Success (n 174) 88 30d mortality
8.4 Rescue Failure (n 24) 12 30d
mortality 30.2 No Rescue (n 266) 30d
mortality 7.9 Successful fibrinolysis (n
1058) 30d mortality 5.2 Pain to Fibrinolytic
3.4h. Further delay 90 min to angiography/ PCI
(per protocol). Total time, pain to Rescue (1st
balloon) 4.9 hrs. 2/3 of failures and of deaths
were patients in cardiogenic shock when PCI was
attempted
58
But when I volunteered this lecture topic to
Peter following the widespread acceptance of the
2007 meta-analysis it was an opportunity and
necessity to look more closely into the plethora
of supportive statements and the details of the
relevant reports.
59
Wijeysundera, H et al JACC 2007 v49 p 422
P.09
The control group could include supportive care
or a second round of fibrinolysis but by the time
of this meta-analysis the latter (rescue lysis)
was not considered a viable strategy, not
superior to supportive care, hence folded into
the control group for a 2 way comparison of PCI v
conservative care
60





Wijeysundera, H et al JACC 2007 v49 p 422
61
2007 Focused Update of the ACC/AHA/SCAI 2005
Guideline Update for Percutaneous Coronary
Intervention
it might be reasonable to select moderate and
high-risk patients for rescue PCI after
suspected failure of fibrinolysis (ST segment
criteria) .
62
Wijeysundera, H et al JACC 2007 v49 p 422
63
Outcome After Rescue PCI and/or Late PCI in
STEMI Patients
They Are Both a Matter of Time and Flow
64
Rescue PCI A Meta-Analysis of Randomized Trials
Wijeysundera, H. C. et al. J Am Coll Cardiol
200749422-430
According to observations by Salvador Dali
65
Rescue PCI A Meta-Analysis of Randomized Trials
Wijeysundera, H. C. et al. J Am Coll Cardiol
200749422-430
According to observations by Salvador Dali
66
Heart,org A new meta-analysis of trials
comparing PCI with medical therapy in patients
randomized later than 12 hours after MI has shown
a significant benefit in cardiac function and
mortality in the PCI group
67
Some of our clinical beliefs and even guidelines
are not as well grounded as we might wish.
68
Randomized Comparison of Percutaneous
Transluminal Coronary Angioplasty and Medical
Therapy in Stable Survivors of Acute Myocardial
Infarction With Single Vessel Disease A Study of
the Arbeitsgemeinschaft Leitende Kardiologische
Krankenhausärzte Uwe Zeymer, Rainer Uebis,
Albrecht Vogt, Hans-Georg Glunz, Hans-Friedrich
Vöhringer, Dietrich Harmjanz, and Karl-Ludwig
Neuhaus
Three hundred patients with single vessel disease
of the infarct vessel and no or minor angina
pectoris in the subacute phase (1 to 6 weeks)
after an acute myocardial infarction were
randomized to angioplasty (n149) or medical
therapy (n151). Primary end point was the
survival free of reinfarction, (re)intervention,
coronary artery bypass surgery, or readmission
for severe angina pectoris at 1 year. The
event-free survival at 1 year was 82 in the
medical group and 90 in the angioplasty group
(P0.06). This difference was mainly driven by
the difference in the need for (re)intervention
(20 vs 8, plt.03)
69
(No Transcript)
70
(No Transcript)
71
Enrolment started in July 1998 at 16 university
hospitals in France and Belgium and finished in
December 2001. To be eligible for inclusion,
patients had to be aged 20-75 years, have a first
Q-wave MI, no spontaneous or low-level recurrent
ischaemia, and angiographic demonstration of
total occlusion of the infarct-related artery
(TIMI grade 0-1 flow) located on a proximal
segment with a reference luminal of at least 2.0
mm.
72
JACC 2002
73
DECOPI Steg et al EHJ
Enrolment started in July 1998 at 16 university
hospitals in France and Belgium and finished in
December 2001. To be eligible for inclusion,
patients had to be aged 20-75 years, have a first
Q-wave MI, no spontaneous or low-level recurrent
ischaemia, and angiographic demonstration of
total occlusion of the infarct-related artery
(TIMI grade 0-1 flow) located on a proximal
segment with a reference luminal of at least 2.0
mm.
74
AHA/ACC STEMI Guidelines (2004)
75
(No Transcript)
76
JACC v 40 5 2002
Conclusions PCI 1 month post STEMI had an
adverse effect on remodeling..
Yousef ZR et al.
JACC v40 5