Surgical guidelines around the
globe propose a simple strategy for minimizing the risk
of perioperative heart attack in at-risk patients: administer
beta-blockers. Now, evidence suggests that this practice
has increased the overall risk of death and disability.
The new study, led by researchers at McMaster University
and the University of Ottawa, was published May 31 in
The Lancet.
In the POISE trial, no fewer than
8,351 patients with cardiac problems or risk factors
were randomized to receive either the beta-blocker metoprolol
or placebo at 190 hospitals in 23 countries. Beta-blockers
did indeed cut down on heart attacks and atrial fibrillation
in surgical patients, but they were linked to an overall
33% increased risk of death. Sepsis was the only cause
of death that was significantly more common in those
on beta-blockers compared to controls. The researchers
believe that the drugs may delay diagnosis of infection
by preventing tachycardia, and impede the delivery of
antibiotics by slowing bloodflow.
BAD
TRADE-OFF
But perhaps an even more serious worry was the sharply
increased incidence of stroke, which was doubled in
the beta-blocker patients — affecting about one
in 150. Though few died, says Dr P J Devereaux of McMaster,
these strokes had a far more serious effect than did
the typical perioperative MIs that beta-blockers aim
to minimize. While most of the heart attack patients
had no major sequels like heart failure or revascularization,
more than half of the stroke victims were permanently
disabled. It's not looking like a good trade-off.
The advice to give perioperative
beta-blockers, even though it came from such august
bodies as the American College of Physicians and the
American Heart Association, actually rests on very little
evidence. Trace it to its origins and only two studies,
comprising a total of 312 patients, back it up. "We
knew how small the original studies were," says Dr Devereaux.
"We also knew there were adverse effects of giving perioperative
beta-blockers, including low blood pressure and bradycardia,
and we wondered if this could be linked to other complications
that we continued to see in surgical patients." He adds:
"We still expected beta-blockers to show an overall
survival benefit. We were surprised by the results."
FULL
DISCLOSURE
The carefully-audited trial struck a hitch when two
sets of data, from Iran and Colombia, were found to
be fraudulent. It wasn't about researchers with an agenda,
said Dr Devereaux: "The motives were pretty standard
financial ones. Patients had been randomized after surgery,
notes filled in after the event, things like that."
The trial's blinded steering committee threw out all
that data without ever knowing what it said. What was
left provided plenty of statistical power.
"When we went back and looked at
the previous studies in a meta-analysis, we found that
the signal was always there," says Dr Devereaux. In
fact, the relative risk of perioperative stroke in beta-blocker
patients in these previous studies was even higher.
But because the studies were small, each saw only a
few strokes or sepsis deaths, and the pattern never
emerged until now.
It shows you can't take anything
for granted until you've had a really large, blinded,
prospective study, says Dr Devereaux. "That's especially
true when you have 100 million major surgeries a year
globally. You've got to get it right."
|
