|
Drug sponsors
would have to include a more thorough discussion of
their product's pros and cons in labeling if FDA
addresses concerns raised by Dartmouth Medical School
researchers that prescribers and patients are not told
enough about the debate that precedes drug
approval.
The drug
review process can raise important questions about a
drug's benefits and harms, but this information does
not make it into labeling where prescribers and
patients can easily find it, Lisa Schwartz and Steven
Woloshin conclude in the Oct. 21 New England Journal of
Medicine. Instead, it remains "practically
inaccessible."
Reviewers'
"sense of uncertainty about the net benefit of drugs is
almost always lost" in labeling, which is written by
sponsors, they note. Clinicians cannot "distinguish
drugs that reviewers endorsed enthusiastically from
those they viewed with great skepticism."
The article
continues the Dartmouth researchers' effort to find
better ways to communicate drug information. They have
worked with FDA to evaluate a Drug Facts Box that
quantifies a drug's benefits and side effects - an
approach supported by the Agency's Risk Communication
Advisory Committee ('The Pink Sheet,'
March 9, 2009).
FDA plans to
develop better guidance for agency staff on how to
write the Clinical Trials Section of labeling so it is
more helpful to health care providers, the agency said
in response to questions about the article.
Labeling
information on efficacy can be improved, FDA
acknowledged, saying it will explore current best
practices in this area. How to present benefits
information to patients was a topic at a recent FDA
workshop on patient medical information ('The Pink Sheet,'
Oct. 5, 2009).
As examples
of poor communication about a drug's net benefits,
Schwartz and Woloshin cite the insomnia drugs
Rozerem (ramelteon) and
Lunesta (eszopiclone), whose labels
provide no efficacy data.
For
Sepracor's Lunesta, labeling says the drug is superior
to placebo. It does not include information that on
average, Lunesta patients still met criteria for
insomnia and reported no clinically meaningful
improvement in next-day alertness or functioning, the
researchers point out.
In studies
supporting Takeda's Rozerem, the drug did not reduce
the proportion of cases meeting the definition of
insomnia or improve any of the secondary outcomes, they
say. The FDA medical review officer team leader in fact
concluded that ramelteon has "a statistically
significant treatment effect that is of marginal
clinical significance."
"The sense
that the FDA's decision was a close call was not
communicated in the label."
Review
documents can provide a sense of reviewers' struggle in
weighing benefit and risk, but this often does not get
into the label. Access to information in these
documents, which are lengthy and inconsistently
organized, could at least be improved by a standardized
executive summary, the researchers contend. The summary
should highlight reviewers' uncertainties, note whether
approval was conditional on a post-approval study, and
include data tables for the main results of the Phase
III trials, they suggest.
|
