Sure to stoke the still-smoldering coals of the COURAGE debate into
fresh flames, a new meta-analysis comparing PCI with medical therapy
for stable coronary disease suggests that PCI might be better than
medical therapy for improving survival, published in the September
9, 2008 issue of the Journal of the American College of Cardiology
(JACC) [1]. The authors of the controversial analysis, which
includes the COURAGE trial and 16 other studies, report that
patients treated with an invasive approach experienced a 20%
reduction in all-cause mortality.

"These findings suggest that a PCI-based invasive strategy may
improve long-term survival compared with a medical-treatment-only
strategy in patients with stable coronary artery disease" Dr Albert
Schömig (Deutsches Herzzentrum, Munich, Germany) and colleagues
write. "This justifies the performance of a new randomized clinical
trial sufficiently powered for evaluating the impact of PCI on long-
term mortality."

The findings, however, are at odds with COURAGE, the largest
randomized controlled trial to address this issue. As previously
reported by heartwire, COURAGE found no advantage in terms of death
or MI for an initial strategy of PCI over an initial strategy of
optimal medical therapy in patients with stable coronary disease.
Since its presentation during the 2007 ACC meeting, COURAGE has
continued to rile interventional cardiologists, who insist the trial
does not mirror real-life patients with angina who can't be
medically controlled. But noninterventional cardiologists have
accused their interventional counterparts of discounting results
that might spell a dip in billings.

To heartwire, the senior author of the new meta-analysis, Dr Adnan
Kastrati (Deutsches Herzzentrum), said that it was precisely because
of this controversy that his group chose to reopen the question of a
mortality benefit with a meta-analysis that included the COURAGE
results.

"It's exactly because we now have included the COURAGE trial that
our meta-analysis should reduce the controversy," Kastrati said. "If
you consider the COURAGE trial, you can see also some reduction in
mortality, but it is not enough statistically. So the reason we did
the study was because no single study is able or sufficiently sized
to assess mortality, and the best, most scientific way to put these
results together is through a meta-analysis. I'm aware of the
critics of this analysis, but we have no other method to do this.
Despite the limitations, the best way to let us know the status of
the data about the effect of PCI on mortality is to pull all of the
studies together in a scientific way, and that is through a meta-
analysis."

What Defines a "Recent" MI?

Unlike other meta-analyses, Kastrati and Schömig's study attempted
to look at all studies that included data on PCI vs medical therapy,
including patients with previous, but not recent, MI. This was
necessary, they insist, to reduce the selection bias that has
plagued other studies.

But this is a sticking point for the COURAGE investigators, who were
asked to comment on the meta-analysis for heartwire.

"This is a flawed meta-analysis because, of the 17 studies that are
included, five are of post-MI patients, and one study is of silent
myocardial ischemia, combined with the 11 chronic stable angina
trials," COURAGE study chair, Dr William E Boden (Buffalo General
Hospital, New York) told heartwire. He notes that the study by Dakik
et al, as well as DANAMI, SWISSI-II, INSPIRE, and ALKK are post-MI
studies, while ACIP is the study of silent myocardial ischemia.

"This is a classic apples plus oranges plus pears meta-analysis,"
Boden continued. "Such a permissive inclusion of disparate clinical
trials violates the fundamental tenets of responsible meta-analysis
and is scientifically unjustified. In none of the 11 chronic stable
angina trials was there a mortality benefit with PCI. Based on this
current meta-analysis, I don't believe that the conclusion that a
PCI-based invasive strategy can improve long-term survival vs
medical therapy only in patients with stable CAD is justified."

COURAGE co-primary investigator Dr William Weintraub (Christiana
Care Health System, Newark, DE) declined to comment in detail but
said he was "entirely supportive" of Boden's comments.

Kastrati, however, stood by the inclusion criteria for his study,
pointing out that COURAGE itself had included post-MI patients with
stable angina. "Many people at some point in their lives have had an
infarction, and I don't see any reason to exclude patients who had
infarctions six weeks ago," Kastrati explained to heartwire. "If you
start to exclude them, what is the [cutoff]? Two weeks, three weeks?
It's a source of bias if you start to say I'm excluding this and
this and this."

To be included in the current meta-analysis, patients could have had
a previous MI but not within the past seven days.

Indeed, Kastrati continued, the major limitation of other meta-
analyses, including at least one that also included COURAGE, is
selection bias. "When you are doing a meta-analysis, you know in
advance the results of the trials, so you can distort the results if
you select the trials that you know lead in a certain direction. So
the best way to do a meta-analysis is to include all the trials."

Significant Reduction in All-Cause Death

In all, Schömig et al's meta-analysis included 7513 patients; within
their respective studies, 3675 were assigned to PCI and 3838 to
medical treatment. At 12 months, 271 patients treated with PCI had
died, compared with 335 patients treated medically, a 20% reduction
in the odds ratio of all-cause death, the authors report.

PCI was also associated with a reduction in cardiac death and a
reduction in MI, although these did not reach statistical
significance.

Schömig and colleagues also analyzed their data looking at rates of
all-cause death over different follow-up periods. They found that
the odds ratio for all-cause death in PCI-treated patients improved
with longer and longer follow-up.

"We know that PCI procedures are associated with an inherent risk
during the period around the intervention, but we need more time for
this disadvantage to be offset by the benefits of follow-up,"
Kastrati explained. "If you look at the results of our study, you
see a relation between follow-up length and degree of benefit with
PCI. I'm not saying you have to wait 20 years, but in patients with
an average age of about 65 years, waiting longer, having a longer
follow-up, you can expect more benefit from PCI."

The authors also point out that 28% of medically treated patients in
their meta-analysis ended up undergoing PCI at some point during
follow-up, a pattern that could have "blunted" the survival
differences between the two groups.

A Contentious Issue

Schömig et al's analysis is accompanied by two additional papers in
the same issue of JACC. In one, Dr Robert A O'Rourke (University of
Texas Health Sciences Center, San Antonio), a co-primary
investigator for COURAGE, lists the limitations of the analysis,
pointing to the fact that the definitions of MI, extent of CAD, and
presence of periprocedural or spontaneous MI are not given in the
paper [2]. Moreover the confidence intervals for mortality end
points in individual trials are wide, making it all the more
problematic to combine in a single analysis, O'Rourke notes.
Secondary prevention varied widely between trials, and patient-level
data were not used, he added. Particularly problematic was the
inclusion of patients with recent MI, he observed, echoing
Boden: "Some studies included patients with recent MI who were not
clearly in a stable phase of their disease, threatening the
generalizability of the findings to patients with stable coronary
disease," he wrote.

"These findings using a meta-analysis instead of a large randomized
controlled trial suggest but do not establish that the PCI-based
invasive strategy may improve long-term survival compared with
medical treatment strategy in stable coronary artery disease,"
O'Rourke concludes. "Based on the strength of the evidence, the
author of this commentary recommends more aggressive medical therapy
for patients with moderate to severe angina and PCI or CABG for
patients whose symptoms persist."

A second "state-of-the-art" paper also published in this issue of
JACC, however, takes a different view from O'Rourke's, permitting
the possibility that PCI yields survival benefit in the right
patients. But the focus of the study by Drs Demosthenes G Katritsis
(Athens Euroclinic, Greece) and Bernhard Meier (Swiss Cardiovascular
Center, Bern) is on the best ways of deciding which patients should
undergo PCI and which should receive medical therapy, something they
propose can be defined by a history of recent MI (less than three
months), a demonstration of significant inducible ischemia, the
detection of reduced fractional flow reserve, or the presence of
specific angiographic features of stenosis.

But like Schömig and Kastrati, they suggest that one of the reasons
earlier studies did not demonstrate a survival benefit with PCI is
limited follow-up.

"In light of the inadequate duration of follow-up, the lack of proof
of efficacy is likely not the proof of lack of efficacy," they write.

Kastrati, for his part, says he was using predominantly PCI in
stable angina patients before COURAGE, and he's seen nothing since
COURAGE to change his approach. And although he and his colleagues
highlight the need for a large study with longer follow-up, Kastrati
acknowledges that such a study is extremely unlikely, especially
given the difficulties COURAGE faced to finish enrollment and follow-
up.

"As long as we don't have a bigger study, this [meta-analysis] is
the best data we have in favor of a PCI strategy," Kastrati told
heartwire. "We have been doing PCI to relieve angina--even after
COURAGE, I think most people agree that PCI relieves angina--but
what our meta-analysis shows is that in addition to this beneficial
effect on angina, the data support a beneficial effect of PCI on
prognosis also."

In the paper, Schömig disclosed receiving unrestricted grant support
for the department of cardiology, Technische Universität; and from
Amersham/General Electric, Bayerische Forschungsstiftung, Bristol-
Myers Squibb, Cordis, Cryocath, Guidant, Medtronic, Nycomed, and
Schering. Kastrati has disclosed receiving lecture fees from Bristol-
Myers Squibb, Cordis, GlaxoSmithKline, Lilly, Medtronic, Novartis,
and Sanofi-Aventis.

Schömig A, Mehilli J, de Waha A, et al. A meta-analysis of 17
randomized trials of a percutaneous coronary intervention-based
strategy in patients with stable coronary artery disease. J Am Coll
Cardiol 2008; 52:894-904.
O'Rourke RA. Optimal medical therapy is a proven option for chronic
stable angina. J Am Coll Cardiol 2008; 52:905-907.
Katritsis DG, Meier B. Percutaneous Coronary intervention for stable
coronary artery disease. J Am Coll Cardiol 2008; 52:889-893.