Neumann F, Kastrati A, Pogatsa-Murray G, Mehilli J, Bollwein H, Bestehorn H, Schmitt C, Seyfarth M, Dirschinger J, Schömig A. Evaluation of Prolonged Antithrombotic Pretreatment ("Cooling-Off"
Strategy) Before Intervention in Patients With Unstable Coronary SyndromesA Randomized Controlled Trial. JAMA. 2003;290(12):1593-1599. doi:10.1001/jama.290.12.1593
Author Affiliations: Medizinische Klinik (Drs Neumann, Seyfarth, and Schömig) and Deutsches Herzzentrum (Drs Kastrati, Pogatsa-Murray, Mehilli, Bollwein, Schmitt, Dirschinger, and Schömig), Technische Universität München, Munich, Germany; and Herz-Zentrum Bad Krozingen (Drs Neumann and Bestehorn), Bad Krozingen, Germany.
Context In unstable coronary syndromes, catheter intervention is frequently
preceded by antithrombotic treatment to reduce periprocedural risk; however,
evidence from clinical trials to support antithrombotic pretreatment is sparse.
Objective To test the hypothesis that prolonged antithrombotic pretreatment improves
the outcome of catheter intervention in patients with acute unstable coronary
syndromes compared with early intervention.
Design, Setting, and Patients Randomized controlled trial conducted from February 27, 2000, to April
8, 2002, and including patients admitted to 2 German tertiary care centers
with symptoms of unstable angina plus either ST-segment depression or elevation
of cardiac troponin T levels.
Interventions Patients were randomly allocated to antithrombotic pretreatment for
3 to 5 days or to early intervention after pretreatment for less than 6 hours.
In both groups, antithrombotic pretreatment consisted of intravenous unfractionated
heparin (60-U/kg bolus followed by infusion adjusted to maintain partial thromboplastin
time of 60 to 85 seconds), aspirin (500-mg intravenous bolus followed by 100-mg
twice-daily oral dose), oral clopidogrel (600-mg loading dose followed by
75-mg twice-daily dose), and intravenous tirofiban (10-µg/kg bolus followed
by continuous infusion of 0.10 µg/kg per min).
Main Outcome Measure Composite 30-day incidence of large nonfatal myocardial infarction or
death from any cause.
Results Of the 410 patients enrolled, 207 were allocated to receive prolonged
antithrombotic pretreatment and 203 to receive early intervention. Elevated
levels of cardiac troponin T were present in 274 patients (67%), while 268
(65%) had ST-segment depression. The antithrombotic pretreatment and the early
intervention groups were well matched with respect to major baseline characteristics
and definitive treatment (catheter revascularization: 133 [64.3%] vs 143 [70.4%],
respectively; coronary artery bypass graft surgery: 16 [7.7%] vs 16 [7.9%]).
The primary end point was reached in 11.6% (3 deaths, 21 infarctions) of the
group receiving prolonged antithrombotic pretreatment and in 5.9% (no deaths,
12 infarctions) of the group receiving early intervention (relative risk,
1.96 [95% confidence interval, 1.01-3.82]; P = .04).
This outcome was attributable to events occurring before catheterization;
after catheterization, both groups incurred 11 events each (P = .92).
Conclusion In patients with unstable coronary syndromes, deferral of intervention
for prolonged antithrombotic pretreatment does not improve the outcome compared
with immediate intervention accompanied by intense antiplatelet treatment.
The best treatment option for most patients with unstable coronary syndromes
is angiographically guided revascularization, irrespective of the primary
success of medical treatment.1- 4 Compared
with a conservative strategy that reserves coronary angiography for persisting
spontaneous or inducible ischemia, an invasive strategy reduces risk of adverse
events including large myocardial infarction (MI),1- 3 severe
recurrent angina,4 and death.2 This
has been shown for patients with non–ST-segment elevation MI and for
those with unstable angina and ST-segment depression.1- 4
The risk associated with percutaneous coronary intervention as well
as that of coronary artery bypass graft (CABG) surgery is, however, increased
in patients with unstable coronary syndromes.5- 8 To
minimize this hazard, some have recommended passivating the plaque activity
by means of extended antithrombotic treatment before intervention.9 Several observational studies comparing swift intervention
vs antithrombotic pretreatment for unstable coronary syndromes have found
a lower rate of complications in patients undergoing percutaneous intervention
after more than 48 hours of antithrombotic pretreatment.10- 12 Pursuing
such a "cooling-off" strategy, investigators in the second Fast Revascularization
during Instability in Coronary Artery Disease (FRISC II) trial1 incorporated
a 4-day medical stabilization period preceding intervention into the FRISC
II protocol and speculated that this prolonged treatment period was a prerequisite
for the substantial benefit of the invasive strategy. Other trials have demonstrated
a reduction in intracoronary thrombus burden by prolonged pretreatment with
glycoprotein IIb/IIIa inhibitors.13,14 Nevertheless,
evidence in support of prolonged antithrombotic pretreatment is still largely
observational, and potent platelet inhibition at the time of intervention
only may suffice to optimize the outcome with early revascularization.15- 17 In the absence of
dedicated clinical trials, current guidelines are equivocal as to the optimal
timing of intervention in unstable coronary syndromes.12,18
To test the hypothesis that prolonged antithrombotic pretreatment is
beneficial before catheter intervention in patients with unstable coronary
syndromes, we conducted the Intracoronary Stenting With Antithrombotic Regimen
Cooling-Off (ISAR-COOL) trial.
Our trial included patients with angina pectoris at rest or with minimal
exertion, with the last episode occurring 24 hours or less before study entry.
Myocardial ischemia had to be verified by horizontal or downsloping ST-segment
depression of 0.1 mV or greater and/or cardiac troponin T concentration of
0.03 mg/L or greater (Roche Cardiac reader system, Roche-Boehringer-Mannheim,
Mannheim, Germany). We excluded patients with evidence of large myocardial
infarction, including ST-segment elevation of at least 1 mV in 2 or more contiguous
leads or elevation of the catalytic activity of creatine kinase and its MB
isoenzyme to 3 times the upper limit of normal or greater; those with hemodynamic
instability; those with contraindications to study medication; or those unable
to provide written informed consent for participation. The study was approved
by the ethics committee of the medical faculty of the Technische Universität
The trial was conducted at 2 German tertiary care centers. Allocation
to early intervention or prolonged antithrombotic pretreatment was made by
means of sealed envelopes containing a concealed computer-generated random
sequence, which was set in blocks of 50 for each of the participating hospitals.
The size of the block was preselected by the statistician and was unknown
to the investigators and medical staff caring for the patients. Dedicated
medical staff performed randomization and assigned participants to their groups
immediately after establishing eligibility and obtaining written informed
consent. Eligible patients were randomized to each of the 2 treatment groups
in equal proportions and in the order that they qualified. The 2 treatment
groups were studied concurrently. Time zero was defined as the time of randomization.
With the early intervention strategy we performed coronary angiography
as soon as possible, at least within 6 hours, during which time antithrombotic
pretreatment was instituted. With the prolonged antithrombotic pretreatment
strategy, we continued pretreatment for at least 3 days, to a maximum of 5
days, after which all patients underwent coronary angiography. Patients assigned
to the prolonged antithrombotic pretreatment strategy could undergo immediate
intervention for severe refractory angina, hemodynamic instability, or when
they reached the primary end point (see below). In both groups, our goal was
immediate revascularization after coronary angiography, preferably by ad hoc
percutaneous catheter intervention.
Apart from duration, antithrombotic pretreatment was identical in both
study groups and was started immediately after randomization. Antithrombotic
pretreatment consisted of (1) unfractionated heparin, initial bolus of 60
U/kg followed by infusion adjusted to a partial thromboplastin time of 60
to 85 seconds; (2) aspirin, initial intravenous bolus of 500 mg followed by
100-mg tablets twice daily; (3) oral clopidogrel, initial loading dose of
600 mg followed by 75 mg twice daily19; and
(4) intravenous tirofiban, initial bolus of 10 µg/kg followed by continuous
infusion of 0.10 µg/kg per minute.3 As
soon as the decision was made to implement percutaneous catheter intervention,
we administered an additional bolus of 60 U/kg of heparin and adjusted the
infusion rate of tirofiban to 0.15 µg/kg per minute.20 We
continued tirofiban for 24 hours and clopidogrel for 4 weeks after the intervention.
At day 4 after catheter intervention, we reduced the dose of clopidogrel to
75 mg once daily. In patients assigned to conservative treatment (ie, medical
treatment alone) or to CABG surgery, we stopped clopidogrel and tirofiban
after catheterization; heparin could be continued until surgery if indicated
clinically. For both groups, the study protocol mandated β-blockers,
angiotensin-converting enzyme inhibitors, and statins as concomitant treatment.
Nitrates were allowed if needed for suppression of angina.
We analyzed biochemical markers of myocardial damage every 8 hours before
catheterization and up to 48 hours after surgical or percutaneous catheter
revascularization. Thereafter and in patients treated conservatively after
catheterization, myocardial marker proteins were analyzed daily and for new
episodes of severe chest pain until hospital discharge. Patients underwent
conventional 12-lead electrocardiography on admission, immediately after revascularization,
and at least once daily thereafter, until hospital discharge. We performed
a telephone interview at 30 days. For patients reporting cardiac symptoms,
at least 1 clinical and electrocardiographic examination was performed in
the outpatient clinic or by the referring physician. Thirty-day follow-up
was complete in all surviving patients; in addition, 330 patients (80.5%)
subsequently had direct visits to our institutions. All information derived
from contingent hospital readmission records or provided by the referring
physician or by the outpatient clinic was entered into a computer database.
Our primary end point was the combined cumulative incidence of large
MI or death from any cause during 30 days of follow-up. Large MI was defined
by the occurence of any of the following: new Q waves in 2 or more contiguous
electrocardiographic leads, new left bundle branch block, or elevation of
the catalytic activity of creatine kinase and its MB isoenzyme to at least
5 times the upper limit of normal. After CABG surgery, we diagnosed large
MI if the catalytic activity of creatinine kinase MB isoenzyme exceeded 30
In addition, we assessed bleeding complications. According to the Thrombolysis
in Myocarcardial Infarction (TIMI) trial definitions,3 a
bleeding event was defined as major if it was intracranial or if there were
clinically significant overt signs of hemorrhage associated with a drop in
hemoglobin of more than 5 g/dL. To account for transfusion, an increase of
1 g/dL in hemoglobin was assumed for each unit of blood. All end points were
adjudicated by an independent blinded committee.
For calculation of sample size, we assumed a 7% 30-day incidence of
death and MI with prolonged antithrombotic pretreatment and hypothesized that
this represented a 60% relative risk reduction compared with early intervention.
We designed the study to have a power of 80% to test this hypothesis, with
a 2-sided P value of less than .05 indicating statistical
significance. According to these assumptions, 203 patients were required in
each treatment group. In a prespecified secondary analysis, we compared the
incidence of our primary end point after catheterization between the 2 treatment
strategies. According to the intention-to-treat principle, all primary analyses
were based on data from all patients as randomized.
Continuous variables are presented as median (interquartile range).
We report categorical variables as counts and proportions and assessed differences
between the groups with use of a 2-sided χ2 test or Fisher
exact test as appropriate. In addition, we report our primary end point as
an information-preserving composite end point.21 We
analyzed survival by the Kaplan-Meier method. We calculated a multivariate
logistic regression model adjusting for age, sex, smoking, serum cholesterol
level, hypertension, diabetes, insulin-dependent diabetes, prior angioplasty,
prior CABG surgery, prior MI, ST-segment depression, cardiac troponin T level,
coronary stenosis of 50% or greater, and left ventricular ejection fraction.
Some of the relative risks and 95% confidence intervals (CIs) were calculated
by conventional methods included in the crosstabs procedure of SPSS version
11.5 (SPSS Inc, Chicago, Ill). We also used S-Plus version 4.5 (Insightful
Corp, Seattle, Wash) for statistical analyses.
The trial profile is shown in Figure
1. From February 27, 2000, until April 8, 2002, we enrolled 410
patients, of whom 207 were allocated to receive prolonged antithrombotic pretreatment
and 203 to receive early intervention. There were no significant differences
between the study groups with respect to entry criteria and other baseline
characteristics (Table 1). Elevated
levels of cardiac troponin T were present in 274 patients (67%) while 268
(65%) had ST-segment depression. In both groups, half of the patients were
aged 70 years and older, and one third (136/410) were women. Patients with
diabetes mellitus constituted 29% (118/410) of our study cohort.
The median time to catheterization with prolonged antithrombotic pretreatment
was 86 hours; we catheterized only 12 patients in this group (5.8%) prematurely
according to the prespecified criteria. Of the patients assigned to early
intervention, 87.2% (177/203) underwent coronary angiography within 6 hours,
while the median time to catheterization was only 2.4 hours (Figure 2).
The angiographic baseline characteristics were not significantly different
between the 2 treatment groups (Table 1). More than two thirds (69.5% ) of our patient population
had multivessel disease and 89% (364) had coronary artery stenosis of 50%
or greater. The definitive treatment did not differ significantly between
the antithrombotic pretreatment and the early intervention groups (catheter
revascularization: 133 [64.3%] vs 143 [70.4%]; CABG surgery: 16 [7.7%] vs
16 [7.9%]), nor did concomitant medication (Table 2). We did not use drug-eluting stents, atherectomy, or brachytherapy.
At 30 days, the cumulative incidence of large MI or death from any cause
differed significantly between the 2 treatment strategies (Figure 3); this combined end point was reached in 11.6% (3 deaths,
21 infarctions) of the antithrombotic pretreatment group and in 5.9% (no deaths,
12 infarctions) of the early intervention group (relative risk [95% CI], 1.96
[1.01-3.82]; P = .04 [Fisher exact test, P = .05]). The observed incidences of both death and MI tended to be
higher with prolonged antithrombotic pretreatment as compared with early intervention
when we analyzed each component of the primary end point separately (Table 3). If we present our primary end
point in an information-preserving form, the number of patients in each of
3 categories (death, nonfatal MI, neither event) was 3, 21, and 183, respectively,
in the antithrombotic pretreatment group and 0, 12, and 191, respectively,
in the early intervention group.
In a multivariable logistic regression model that took into account
baseline variables listed in Table 1,
the adjusted odds ratio (OR) was 2.17 (95% CI, 1.01-4.76) (P = .047) for the primary end point comparing the prolonged antithrombotic
pretreatment strategy with the early intervention strategy. This was similar
to the unadjusted OR of 2.09 (95% CI, 1.01-4.30) (P =
.046) for the primary end point comparing prolonged antithriombotic pretreatment
with early intervention.
Findings in subgroups defined by entry criteria or definite treatment
were consistent with the main study results. We found no significant effect
on our primary end point when comparing prolonged antithrombotic pretreatment
vs early intervention, either in patients with elevated levels of cardiac
troponin T (OR, 1.65 [95% CI, 0.75-3.64]) or in those with ST-segment depression
(OR, 1.50 [95% CI, 0.67-3.37]). Similarly, in patients undergoing percutaneous
catheter intervention there was no detectable benefit of prolonged antithrombotic
pretreatment compared with early intervention (OR, 1.64 [95% CI, 0.73-3.68]).
Irrespective of the duration of pretreatment, the event rate after catheterization
was similar in both study groups (P = .92), with
11 events occurring in each of the 2 treatment groups (Figure 4). In the early intervention group only 1 event occurred
before catheterization, whereas the longer duration of pretreatment in the
prolonged antithrombotic pretreatment group resulted in 13 precatheterization
events. Major bleeding and severe thrombocytopenia occurred at a similar rate
in both groups (Table 3). None
of our patients experienced cerebral hemorrhage.
In our randomized controlled trial in patients with unstable coronary
syndromes, deferral of intervention for prolonged antithrombotic pretreatment
did not improve outcome compared with immediate intervention accompanied by
intense antiplatelet treatment. On the contrary, we observed a significant
increase in the cumulative 30-day incidence of large MI or death from any
cause in patients who underwent prolonged antithrombotic pretreatment. The
cardiac event rate after catheterization was almost identical irrespective
of the duration of antithrombotic pretreatment. The intense antithrombotic
treatment including triple antiplatelet therapy at the time of catheter intervention
may have contributed substantially to outcome after catheterization. Without
this antiplatelet regimen the prevention of excessive peri-interventional
MIs in the early intervention group might not have occurred.
In trials on the peri-interventional use of abciximab, the excessive
risk of catheter intervention in unstable coronary syndromes was reduced to
close to that observed among patients with stable angina.15- 17 Thus,
abciximab administered at the time of intervention appeared to achieve plaque
passivation.22 In our study, the high loading
dose of clopidogrel in combination with tirofiban afforded even stronger platelet
inhibition at the time of intervention.19 Whereas
earlier trials used the 300-mg loading dose of clopidogrel, the 600-mg loading
dose used in the ISAR-COOL trial had been tested in 2 more recent studies—a
large registry23 as well as the ISAR-REACT
(Intracoronary Stenting and Antithrombotic Regimen–Rapid Early Action
For Coronary Treatment) trial.24 The ISAR-REACT
trial addressed peri-interventional antithrombotic therapy in patients without
myocardial marker proteins or ischemic ST-segment changes. In the low to intermediate–risk
cohort of the ISAR-REACT trial, clopidogrel alone with a 600-mg loading dose
was highly effective because abciximab was no longer needed to reduce the
risk of peri-interventional ischemic events. With a 600-mg loading dose, a
2-hour delay, as in the early intervention group, suffices to achieve an almost
full effect of adenosine diphosphate–receptor inhibition.19 Within
the same time frame, tirofiban would have achieved a steady-state concentration
that enables a robust potent antiaggregatory effect.25 Our
study shows that the plaque passivation achieved by such intense antiplatelet
therapy at the time of intervention cannot be surpassed by extended pretreatment;
ie, extended pretreatment does not improve outcome.
Compared with the early intervention strategy, the excessive event rate
with the antithrombotic pretreatment strategy was incurred exclusively during
pretreatment, presumably as a consequence of its differential duration. During
the pretreatment phase (median duration, 3.6 days), 6.3% of the patients receiving
prolonged antithrombotic pretreatment experienced an event. Reflecting the
high-risk characteristics of our study population, this rate was higher than
the daily rate of 1.3% in the Chimeric 7E3 Anti-Platelet Therapy in Refractory
Unstable Angina Treatment (CAPTURE) trial26 or
the approximate daily rate of 1% found in other studies during conservative
treatment.27- 29 Compared
with other trials,26- 29 our
patient population was the oldest, comprised the highest proportion of patients
with ST-segment abnormalities, and had the highest prevalence of diabetes
mellitus. Our study demonstrates that despite recent advances in antithrombotic
therapy the risk of large MI or death from any cause is still substantial
during conservative treatment of unstable coronary syndromes in unselected
high-risk patient settings.
When we designed our study, we chose the duration of the pretreatment
phase based on FRISC II.1 Shorter pretreatment
phases lasting a median of 1 or 2 days have been instituted in more recent
trials.3,4 Nevertheless, the principle
message of our study would not have been altered had we chosen a shorter duration
of pretreatment. In our study, there was no trade-off between a reduction
in postcatheterization events and the incidence of precatheterization events
that could be optimized by choosing another duration of pretreatment. On the
contrary, our study suggests that the shorter the pretreatment phase the lower
the incidence of adverse events.
As in other contemporary studies, 30-day mortality in the entire population
was below 1%.1- 4,27 Thus,
our primary end point and the benefit from early intervention were driven
largely by the incidence of MI; one-third of patients with MI will develop
new Q waves and two-thirds will not. Consistent with the outcome for nonfatal
MI, the observed mortality rates also favored early intervention. These analyses
need to be interpreted cautiously, because our study was not powered to analyze
each component of our primary end point. In our admittedly underpowered subgroup
analyses we did not find any indication that the benefit from early intervention
was linked to patients with elevated levels of cardiac troponin T or to those
with ST-segment depression, or to a specific modality of definite treatment,
such as catheter revascularization or CABG surgery.
In patients with unstable coronary syndromes, intense antithrombotic
pretreatment with triple antiplatelet therapy reduces the risk of adverse
cardiac events during an unavoidable conservative treatment phase,29 but does not appear to have any further benefit.
Conservative pretreatment consumes considerable resources, including an increase
in length of hospital stay (median [interquartile range] for prolonged vs
early intervention, 7 [6-11] days vs 5 [3-7] days; P<.001)
is associated with a substantial risk of cardiac complications, and does not
reduce the risk of subsequent revascularization procedures. In patients with
unstable coronary syndromes, antithrombotic pretreatment should therefore
be kept to the minimum duration required to organize early cardiac catheterization