Martha Rosenberg for Truthout: You were an
FDA drug reviewer from 1998 to 2008, working on well-known drugs
like Cymbalta, Zyprexa, Concerta, Invega, Provigil and Saphris,
and encountered the same kind of coercive working environment as
the device reviewers.
Ronald Kavanagh:
That's correct.
In the Center for Drugs [Center for Drug
Evaluation and Research or CDER], as in the Center for Devices,
the honest employee fears the dishonest employee.
There is also
irrefutable evidence that managers at CDER have placed the
nation at risk by corrupting the evaluation of drugs and by
interfering with our ability to ensure the safety and efficacy
of drugs. While I was at FDA, drug reviewers were clearly told
not to question drug companies and that our job was to approve
drugs.
We were prevented, except in rare instances, from
presenting findings at advisory committees. In 2007, formal
policies were instituted so that speaking in any way that could
reflect poorly on the agency could result in termination.
If we
asked questions that could delay or prevent a drug's approval -
which of course was our job as drug reviewers - management would
reprimand us, reassign us, hold secret meetings about us, and
worse.
Obviously in such an environment, people will
self-censor.
MR:
What are some of the ways in which safety risks were minimized
in drug evaluation and review?
RK:
Well, first of all I think most people would be shocked at how
malleable safety data is.
Human studies are usually too short
and the number of subjects in them too small to adequately
characterize the most dangerous risks. That's why even a single
case has to be taken seriously.
A safety signal from any study -
and not just safety data from short term efficacy and safety
studies (used for labeling) - needs to be evaluated.
This means
data from long term safety studies needs to be evaluated as well
as the data from even longer, ongoing safety studies and from
clinical pharmacology studies. Some of this information also
needs to be examined during development of a drug.
Yet I have
seen new drug reviews where none of this was done by the medical
safety reviewer.
MR:
Would you give an example?
RK:
For example, human clinical pharmacology trials are typically
done in Europe, yet clinical pharmacology reviewers at FDA have
been barred from analyzing this information prior to studies
being conducted in the US.
Without being able to do this, we are
unable to detect evidence of risks early and cannot provide
guidance that would help with the development of the drug in
terms not only of safety and proving efficacy, but also with the
efficiency and cost effectiveness of the drug's development.
New
labeling policies can also mask risks as they exclude the
labeling of adverse events if they are under a certain
percentage and/or not double the rate found with a placebo.
By
this rule, certain serious and potentially lethal adverse events
that eventually resulted in a drug being withdrawn from the
market would not have had any mention of the adverse events made
in the labeling at all.
On top of that, I frequently found
companies submitting certain data to one place and other data to
another place and safety information elsewhere so it could not
all be pulled together and then coming in for a meeting to
obtain an agreement and proposing that the safety issue is
negligible and does not need further evaluation.
MR:
Like they are trying to pull the
wool over the FDA's eyes?
RK:
During development, if reviewers say things that companies don't
like, they will complain about the reviewer or they will call
upper management and have the reviewer removed or overruled.
On
one occasion, the company even told me they were going to call
upper management to get a clear requirement for approval that
they did not want to fulfill eliminated, which I then saw
happen.
On another occasion a company clearly stated in a
meeting that they had "paid for an approval."
MR:
That is shocking. Wouldn't the FDA
managers want safety risks investigated?
RK:
Just the opposite. Sometimes we were literally instructed to
only read a 100-150 page summary and to accept drug company
claims without examining the actual data, which on multiple
occasions I found directly contradicted the summary document.
Other times I was ordered not to review certain sections of the
submission, but invariably that's where the safety issues would
be.
This could only occur if FDA management was told about
issues in the submission before it had even been reviewed. In
addition, management would overload us with huge amounts of
material that could not possibly be read by a given deadline and
would withhold assistance.
When you are able to dig in, if you
found issues that would make you turn down a drug, you could be
pressured to reverse your decision or the review would then be
handed off to someone who would simply copy and paste whatever
claims the company made in the summary document.
MR:
You have recounted that this is what
happened to you with the nerve gas drug
pyridostigmine.
RK:
Yes, pyridostigmine is intended to be given preventatively in
case of a nerve gas attack with the nerve agent Soman and it was
used experimentally on Gulf War troops.
After the first Gulf
War, there were concerns it was linked to Gulf War Illness.
Then, prior to Operation Iraqi Freedom, the Defense Department (DoD)
tried to have President
Bush waive informed consent for pyridostigmine, even though it was still an investigational
drug.
MR:
Why?
RK:
Possibly because there is less hassle medicating troops if no
informed consent is required.
When President Bush refused to
waive informed consent, the FDA approved pyridostigmine using
the "Animal Rule" which allows the approval of drugs for human
use based on animal data.
It was employed because it was
unethical to dose humans with the nerve agent Soman to see if
pyridostigmine would actually prevent death.
However, the way
the drugs were used in the animal studies didn't reflect how
they would be used in humans and resulted in misleading
conclusions.
MR:
Another FDA reviewer turned down pyridostigmine before you?
RK:
Yes. I was assigned to re-review his conclusions regarding
pyridostigmine and even before I began my review I was pressured
to approve it and this pressure continued through nearly two
dozen meetings with FDA management.
After it became clear that I
would not be pressured into an approval and it became apparent
that it would be approved according to the animal rule in spite
of the science, I raised an even stronger objection:
not only
did it not work against nerve agents other than Soman, but
pyridostigmine actually increased
lethality in the presence of other nerve agents and we knew that
Saddam Hussein was not using Soman and was instead using these
other nerve agents.
MR:
So, you were just stating what should have been obvious?
RK:
This information was not secret - both FDA and DoD public
documents acknowledge increased lethality with other nerve
agents such as Sarin, and DoD and other government documents
that are public also document that Saddam Hussein was not using
Soman and was instead using these other nerve agents
exclusively.
Yet because I raised this as an objection, I was
immediately replaced as the primary reviewer so that I could not
document my concerns and so that pyridostigmine could be
approved.
It's since been proposed that if we ever face the
prospect of nerve agents in the future, that this approval will
be used as a justification to convince the President at that
time to waive informed consent without presenting a full
picture.
Even though using pyridostigmine would likely only
invite the use of nerve agents.
MR:
Why would the FDA and DoD allow troops to be put in this kind of
harm's way?
RK:
I don't know and don't want to speculate.
However senior
managers made statements indicating knowledge that the approval
was illegal. In any case, it was clear and known that use of pyridostigmine would interfere with the operation of our troops.
MR:
Your training as a pediatric clinical pharmacologist has made
you especially sensitive to drug risks for children. What are
some of the unique drug risks children face?
RK:
Pediatric approvals are based on the assumption that children
will respond similarly to similar exposures.
Yet dosages that
are used for studies in children are often based on approved
adult dosages rather than a scientific determination of whether
children achieve the same or higher exposures than adults. This
is because companies don't want to develop lower dosages for
children if they don't have to.
Thus exposure studies in
children are done after the efficacy studies have been begun
instead of before when it's needed. The exposure studies then
may also use overweight children as well as too few children.
Since no allowance is made for race, age, puberty, or actual
weight and since there are differences in children's clearance
of drugs, there are often higher exposures to active and toxic
metabolites in children compared to adults.
Thus there are often
unnecessary risks with the doses that are approved.
MR:
Are there other risks with one-size-fits-all doses?
RK:
There are racial differences in drug metabolism that are not
taken into consideration.
For example, one anticancer drug
breaks down faster in African Americans, so patients don't get
sufficient exposure to the drug to kill tumors. Yet African
Americans were not included in the safety and efficacy studies.
When drugs break down faster by one particular pathway, the
patients will also sustain greater toxicity and even death from
the toxic metabolite that is formed.
This is especially true
when the company subsequently recommends higher doses to
overcome the lower exposure due to faster metabolism.
In one
case, this occurred with a drug used in pregnant women, where
hormonal changes during pregnancy cause a greater breakdown to a
metabolite that is suspected to cause mental retardation in
children exposed during the pregnancy.
Not only does the
labeling suggest possible use during pregnancy, the labeling
recommends a higher dose during pregnancy.
All the while, it
appears that the company was aware of the formation of a
metabolite that likely affects brain development from well
before the drug was ever submitted to the FDA.
MR:
Are the risks just ignored?
RK:
FDA's response to most expected risks is to deny them and wait
until there is irrefutable evidence postmarketing, and then
simply add a watered down warning in the labeling.
In fact, when
patients exhibit drug toxicity, it is usually attributed to an
underlying condition which we know is likely to make the drug
toxicity worse.
This also allows the toxicity to be dismissed as
being unrelated to the drug in any way. Consequently, toxicities
are only attributed to the drug when the evidence is
irrefutable. Thus the majority of cases where there is a
contributing factor are simply dismissed.
When you do raise
potential safety issues, the refrain that I heard repeatedly
from upper management was‚
"where are the dead bodies in the
street?"
Which I took to mean that we only do something if the
press is making an issue of it.
MR:
You have also spoken about the dangers of certain ADHD drugs and
presented some damning data about Cephalon's stimulant
Provigil.
RK:
In 2006, a medical reviewer found several cases of what he
thought might be Stevens Johnson Syndrome (SJS) in children who
took Provigil or modafinil.
SJS and the related conditions
erythema multiforme and toxic epidermal necrolysis (TEN) are
life-threatening skin conditions where huge swathes of skin
covering large sections of the body die and slough off and the
mucus membranes are also affected.
The diseases are incredibly
painful and kill 10 and 40 percent respectively of the people
who develop them.
The reviewer believed he was going to be
overruled and asked me for help. We were able to get an advisory
committee meeting in which I was allowed to present slides of
the data that supported a diagnosis of SJS in a child in the
study.
I also showed that a metabolite of modafinil was 16 times
higher in children than in adults and similar to the worst drug
that exists for causing SJS, Blephamide.
The drug company
doctors were unprepared for my presentation and claimed they had
no information on the child, including no photos and that they
had lost contact.
MR:
One of the pharma doctors actually tried to downplay SJS with
modafinil, saying a child was hospitalized, but was not in the
"burn unit," according to the transcript.
RK:
Yes. Largely because of my presentation, the advisory committee
voted 12-1 against approval, but Cephalon claimed in the press
that the rash was viral and was not from the drug.
The next
year, armodafinil, a related drug, was approved with a
contraindication for children with a contraindication following
three months later for modafinil.
Contemporaneously, Cephalon
agreed to pay $425 million for off-label marketing of modafinil.
That means that for 18 months, the FDA kept quiet about the
issue of SJS in children, while Cephalon continued off-label
marketing at full steam.
Later, I found that the FDA had
internal documents that had the same conclusion as my analysis
but they had been withheld from the advisory committee.
All drugs have dangers including
death, and psychiatric drugs tend to be particularly dangerous,
but as long as we make reasonable attempts to minimize risks,
and provide adequate information for prescribers and patients, I
am not opposed to them.
On multiple occasions I have stood up
for smaller drug companies against FDA management.
MR:
The recent revelations of reprisals against FDA device reviewers
must not have surprised you at all.
RK:
No they didn't. After FDA management learned I had gone to
Congress about certain issues, I found my office had been
entered and my computer physically tampered with.
I saw strange
cursor movements on my computer when I was just sitting at my
desk reading that I suspected was evidence of spying. After I
gave Representative Waxman's (D-CA) office a USB drive with
evidence, FDA staff was admonished that it was prohibited to
download information to USB drives.
Then, after I openly
reported irregularities in an antipsychotic drug review and FDA
financial collusion with outsiders to Senator Grassley’s office
and the House Committee on Oversight and Government Reform, I
was threatened with prison if I should release trade secret
information to Congress.
MR:
That is similar to the FDA's claim with the device reviewers.
Why do efforts to silence free speech always seem to be couched
as "trade secrets"?
RK:
Because much of the information we receive are trade secrets and
companies explicitly label everything they provide the FDA as
such and explicitly prohibit their dissemination.
In spite of
this, the Food Drug and Cosmetics Act explicitly allows
communication of trade secrets by FDA employees to Congress, but
since most people are unaware of this, FDA management can use
the threat of jail for violation of the Trade Secrets Act, not
only to discourage reviewers, but in my case they got Senator
Grassley's staff to destroy the evidence I provided them.
The
threats, however, can be much worse than prison. One manager
threatened my children - who had just turned 4 and 7 years old -
and in one large staff meeting, I was referred to as a
"saboteur."
Based on other things that happened and were said, I
was afraid that I could be killed for talking to Congress and
criminal investigators.
MR:
Still, the FDA transparency meeting transcripts indicate you not
only went to members of Congress, you appealed to the Health and
Human Services inspector general.
RK:
Congress did put me in contact with the Justice Department,
however, I don't believe my complaints were taken seriously by
the FBI or investigated.
I believe that actual felonies may well
have occurred. For example, I found evidence of insider trading
of drug company stocks reflecting knowledge that likely only FDA
management would have known.
I believe I also have documentation
of falsification of documents, fraud, perjury, and widespread
racketeering, including witnesses tampering and witness
retaliation.
MR:
And in addition to this alleged wrongdoing, the public is at
risk from unsafe drugs that were approved?
RK:
Yes. In fact, thanks in part to the
Prescription Drug User Fee
Act, [in which drug companies pay for expedited reviews]
thalidomide could not be stopped today.
