Results of a randomized trial show no overall benefit from treatment with
acetaminophen (paracetamol) in improving outcomes after an acute stroke.

The researchers had hypothesized that lowering body temperature after stroke
using acetaminophen might limit damage and improve outcome, but in this trial at
least, there was no overall effect of treatment. "These results do not support
routine use of high-dose paracetamol in patients with acute stroke," the
researchers, with first author Heleen M. den Hertog, MD, from Erasmus Medical
Center, in Rotterdam, the Netherlands, conclude.

However, in patients who presented with a body temperature above the median of
37ºC, up to 39ºC, there was some signal of benefit, they note, "but this post
hoc finding needs further study."


Results of the Paracetamol (Acetaminophen) for Ischemic Stroke (PAIS) trial,
published online March 17, are in the May issue of the Lancet Neurology. They
were first presented last September at the 6th World Stroke Congress and
reported by Medscape Neurology & Neurosurgery at that time.

Reducing Temperature

Many patients with stroke have fever or subfebrile body temperatures, which have
been associated with case fatality and poor outcome, the authors write. For
every degree increase in body temperature, the odds of poor outcome are doubled,
they note.

The PAIS trial was a randomized, double-blind, controlled trial that included
1400 patients with ischemic stroke or intracerebral hemorrhage and a body
temperature between 36ºC and 39ºC, who were enrolled within 12 hours of symptom
onset. They were randomized to receive 6 g/day of acetaminophen or placebo.

The primary outcome was improvement beyond expectation on the modified Rankin
Scale (mRS) at 3 months, according to a sliding dichotomy approach.

The primary analysis showed that 37% (260/697) of patients improved beyond
expectation in the acetaminophen group vs 33% (232/703) in the placebo group, a
difference that was not statistically significant.

However, they did find a difference in treatment effect based on baseline body
temperature. For patients with an admission body temperature of 37°C and higher,
the odds ratio for improvement in functional outcome was 1.43.

There was no difference in secondary end points or in predefined subgroups; for
example, there was no evidence of benefit for any stroke subtype or time from
stroke onset.

Importantly, there was no increase in adverse events with acetaminophen
treatment, including the rate of infections or liver-enzyme disturbances. There
were 55 (8%) serious adverse events in the acetaminophen group vs 70 (10%) in
the placebo group. Fewer patients on acetaminophen died within 14 days of their
stroke, but no difference was found in case fatality at 3 months.

While their results do not support routine use of acetaminophen, Dr. den Hertog
and colleagues conclude, they call the post hoc finding "promising," although
they point out that it requires confirmation in an independent study. "If such
an effect can be confirmed, a simple, safe, and cheap treatment with a long time
window for start of therapy will be available for patients with acute ischemic
stroke or intracerebral hemorrhage," they write.

Efforts Better Focused on Other Treatments?

In a Reflection and Reaction commentary that accompanies the paper, Scott E
Kasner, MD, from the Comprehensive Stroke Center at the University of
Pennsylvania Medical Center, in Philadelphia, notes that this trial, which was
about 5 times larger than any previous trials of this intervention, showed no
overall benefit. However, the study, which originally was planned to enroll 2500
patients, was stopped at 1400 patients for insufficient financial resources, he
points out.

The statistical plan was changed prior to termination to measure improvement
beyond expectation using this sliding dichotomy approach, Dr. Kasner notes.
"This revised approach was a good choice to maximize power, but the results are
difficult to interpret as clinical reality," he writes. "Moreover, this outcome
probably overemphasizes very small differences between groups."

Dr. Kasner agrees that "the treatment is appealing, because it is easy to use,
inexpensive, and widely applicable." However, he points out that cost would be a
major obstacle to a trial to confirm this finding, since it would require
enrollment of between 10,000 and 20,000 patients.

"Our funds, resources, and efforts may be better focused on more promising
potential treatments of acute stroke," he writes. "In clinical practice,
existing data seem sufficient for us to comfortably conclude that paracetamol is
safe and will lower high temperatures in patients with acute stroke, and that is
exactly what we should use it for and what we should expect it to do."

The study was funded by the Netherlands Heart Foundation. The authors declare no
conflict of interest. Dr. Kasner reports no conflict of interest.

Lancet Neurol. 2009;8:434-440 Abstract, 415-416. Abstract