FDA Procedures Draw Scrutiny
Vytorin, Avandia Cases Fuel Debate About Drug-Approval Process

By ANNA WILDE MATHEWS and RON WINSLOW
January 25, 2008

Controversies about cholesterol drug Vytorin and diabetes drug Avandia
are reigniting debate over what evidence the Food and Drug
Administration requires to approve drugs -- and may generate pressure
on the agency to raise its bar.

The FDA's system for approving drugs has been criticized as scrutiny
intensifies over Vytorin, marketed by Schering-Plough Corp. and Merck
& Co., in the wake of a study that questioned whether the widely
advertised blockbuster worked better than a cheaper generic.

Yesterday, Democratic Reps. John Dingell and Bart Stupak, both of
Michigan and leaders of the House Energy and Commerce Committee, sent
the latest in a volley of letters about the drug. Sen. Charles
Grassley (R., Iowa), who has led several investigations of the FDA and
the drug industry, yesterday opened his own probe into Vytorin. Both
Messrs. Stupak and Grassley said they were concerned about the issue
of FDA approval standards.

A shift by the FDA toward tougher scrutiny of new drugs could add
hundreds of millions of dollars to the cost of developing a drug at a
time when some big drug makers are struggling to replenish product
pipelines.

In addition, any preapproval holdup eats into the manufacturer's
crucial time to sell the drug before protective patents expire and
generic competition leaps in. "If you're looking at a decadelong
trial, you may not have sufficient incentive...to pursue that drug,"
says Alan Goldhammer, a deputy vice president at the Pharmaceutical
Research and Manufacturers of America, a trade group.

The lawmakers' interest is the latest sign that the flap over Vytorin,
in addition to a recent controversy about the safety of
GlaxoSmithKline PLC's Avandia, is adding new fuel to a long-running
debate over FDA approval standards -- and, specifically, a mechanism
known as "surrogate markers."

Vytorin and Avandia went on the market based largely on evidence that
they helped control patients' cholesterol and blood sugar,
respectively. These lab measures were believed to reflect important
clinical benefits, such as reducing the risk of heart attacks. Thus,
the measures serve as proxy markers, or surrogates, for the drug's
broader and more important effect on the body. Using surrogates helps
speed drug approvals, because studies can generally be shorter,
smaller and cheaper.

But the proxy markers can be misleading. Sometimes a drug works on a
proxy but doesn't deliver the promised benefit for the primary health
problem it is supposed to attack. Or a drug can have side effects that
don't surface during initial proxy-marker studies but end up
outweighing its benefit.

"There are inherent risks in using surrogate markers," said Clifford
Rosen, a senior scientist at the Maine Medical Center Research
Institute who, after leading an FDA panel on Avandia, wrote a piece in
the New England Journal of Medicine that questioned whether the
surrogate endpoint was enough for its approval.

In the case of Vytorin, the new study showed the drug didn't slow the
progression of heart disease better than a cheaper generic, even
though Vytorin did have a bigger effect on so-called LDL, or "bad"
cholesterol.

Robert Califf, director of Duke University's Translational Medicine
Institute, says using surrogate endpoints is a reasonable strategy for
allowing drugs to come on the market. But he said that for major
chronic ailments such as cardiovascular disease, the large-scale, long-
term studies needed to show benefits against heart attacks and other
events should begin immediately after a drug is approved. Direct-to-
consumer advertising shouldn't be permitted until those results are
in, he said. Finally, prescribing information, or labels, for such
medicines should clearly reflect the uncertainty of the long-term
benefit until the data are in.

In the case of Zetia, one of the ingredients in Vytorin, Merck and
Schering-Plough didn't launch a major long-term study until 2006, four
years after the drug was approved. Data aren't expected until about
2011. Combined sales of Zetia and Vytorin hit $5 billion in 2007 in
part due to aggressive advertising.

In the case of Avandia, after concerns emerged about a potential link
between the drug and a heightened risk of heart attacks, some
researchers suggested the FDA should have demanded more data on the
drug's cardiovascular impact. Cardiovascular problems are common among
diabetes patients.

"The FDA needs to have more teeth," says John Buse, a professor at the
University of North Carolina at Chapel Hill and a president of the
American Diabetes Association. "They need to demand hard endpoint
studies at the time of approval and they need to have some system for
monitoring progress" on the studies after the drugs are on the market.
Dr. Buse said he wasn't speaking for the ADA.

Mr. Grassley said yesterday that "in light of what's happened with
Avandia and Vytorin, maybe the FDA needs to re-examine when it's
appropriate to use surrogate endpoints....These two cases also
highlight the importance of vigilant postmarketing surveillance and
the need for more postmarketing studies to address important safety
questions."

Mr. Stupak said that requiring clinical endpoint studies for all drugs
before approval would delay their marketing. But, he added, "people
taking Vytorin are doing so because they believe it will reduce their
risk of heart attack. It would make sense for FDA to require
manufacturers to conduct an endpoint study to determine whether
Vytorin just reduces cholesterol or if it also reduces heart attacks."

The FDA has indicated only that it will look at the issue. In a
statement about the Vytorin study, the agency said its officials "will
be considering what, if any, impact this new study will have on our
standards for approval of cholesterol lowering drugs and may seek
outside input at a future public meeting." FDA officials also wrote to
the New England Journal of Medicine to defend the use of blood sugar
as a basis for approval, saying that requiring long-term
cardiovascular benefits "would significantly delay the availability of
new drugs for diabetes."

The FDA has long struggled with when to use the surrogate markers. In
one dramatic example of their limits, a study published in 1991 showed
that drugs that had been approved to prevent deaths from a form of
heart arrhythmia were actually tied to a higher rate of death instead.
The medications had been approved based largely on a surrogate marker
-- a reduction in extra heartbeats after a heart attack -- and not on
whether patients were more likely to survive.

But starting in the late 1980s, the use of surrogate markers helped
speed the flow of drugs to fight AIDS. "It's an incredible advance for
AIDS drug approval to be able to rely on surrogate markers," says
Scott Hammer, a professor at Columbia Medical School. But, he says,
surrogate markers "almost never explain 100% of the treatment effect"
of a drug.

Write to Anna Wilde Mathews at anna.mathews@wsj.com and Ron Winslow at
ron.winslow@wsj.com