1
DEPARTMENT OF HEALTH AND HUMAN
SERVICES
FOOD AND DRUG
ADMINISTRATION
CENTER FOR DRUG EVALUATION AND
RESEARCH
JOINT MEETING OF
THE ARTHRITIS ADVISORY
COMMITTEE AND
THE DRUG SAFETY AND RISK
MANAGEMENT
ADVISORY COMMITTEE
VOLUME II
Hilton
2
P A R T I C I P A N T S
Alastair J.J. Wood, M.D., Chair
Kimberly Littleton Topper, M.D. Executive
Secretary
ARTHRITIS ADVISORY COMMITTEE MEMBERS
Allan Gibofsky, M.D., J.D., Chair
Joan M. Bathon, M.D.
Dennis W. Boulware, M.D.
John J. Cush, M.D.
Gary Stuart Hoffman, M.D.
Norman T. Ilowite, M.D.
Susan M. Manzi, M.D., M.P.H.
DRUG SAFETY AND RISK MANAGEMENT ADVISORY
COMMITTEE
MEMBERS
Peter A. Gross, M.D., Chair
Stephanie Y. Crawford, Ph.D., M.P.H.
Ruth S. Day, Ph.D.
Curt D. Furberg, M.D., Ph.D.
Jacqueline S. Gardner, Ph.D., M.P.H.
Eric S. Holmboe, M.D.
Arthur A. Levin, M.P.H., Consumer Rep.
Louis A. Morris, Ph.D.
Richard Platt, M.D., M.Sc.
Robyn S. Shapiro, J.D.
Annette Stemhagen,
Dr.PH., Industry Rep.
FDA CONSULTANTS (VOTING)
Steven Abramson, M.D.
Ralph B. D'Agostino,
Ph.D.
Robert H. Dworkin, Ph.D.
Janet Elashoff, Ph.D.
John T.
Farrar, M.D.
Leona M. Malone, L.C.S.W.,
Patient Rep.
Thomas Fleming, Ph.D.
Charles H. Hennekens,
M.D.
Steven
Nissen, M.D.
Emil
Paganini, M.D., FACP, FRCP
Steven L. Shafer, M.D.
Alastair J.J. Wood, M.D. (Meeting Chair)
3
P A R T I C I P A N T S (Continued)
FDA CONSULTANTS
(NON-VOTING)
Byron Cryer, M.D. (Speaker and
Discussant)
Milton Packer, M.D. (Speaker only)
NIH PARTICIPANTS (VOTING)
Richard O. Cannon, III, M.D.
Michael J. Domanski, M.D.
GUEST SPEAKERS (Non-Voting)
Garret A. FitzGerald, M.D.
Ernest Hawk, M.D., M.P.H.
Bernard Levin, M.D.
Constantine Lyketsos, M.S., M.H.S.
FDA (CDER)
Jonca Bull, M.D.
David Graham, M.D., M.P.H.
Brian Harvey, M.D.
Sharon Hertz, M.D.
John Jenkins, M.D., F.C.C.P.
Sandy Kweder, M.D.
Robert O'Neil, Ph.D.
Joel Schiffenbauer, M.D.
Paul
Seligman, M.D.
Robert Temple, M.D.
Anne
Trontell, M.D., M.P.H.
Lourdes
Villalba, M.D.
James Witter, M.D., Ph.D.
Steven Galson, M.D.
Kimberly Littleton Topper, M.S.,
Executive
Secretary
4
C O N T E N T S
Call to Order:
Alastair J.J. Wood, M.D.,
Chair 5
Conflict of Interest Statement:
Kimberly Littleton Topper,
M.S. 5
Interpretation of Observational Studies
of Cardiovascular Risk of Non-steroidal
Drugs
Richard Platt, M.D., M.S. 8
Review of Epidemiologic Studies on
Cardiovascular Risk with Selected NSAIDs
David Graham, M.D., M.P.H. 37
Committee Questions to Speakers 89
Arcoxia (etoricoxib)
Merck Research
Laboratories
Sponsor Presentation
Sean P. Curtis, M.D. 152
FDA
Presentation
Joel Schiffenbauer, M.D. 189
Lumiracoxib
Novartis
Pharmaceuticals
Sponsor
Presentation
Introduction
Mathias Hukkelhoven, Ph.D. 201
Gastrointestinal and Cardiovascular Safety
of Lumiracoxib, Ibuprofen,
and Naproxen
Patrice Matchaba, M.D. 205
Open Public Hearing 236
FDA Presentation (Lumiracoxib)
Committee Questions to Speakers 346
Committee Discussion 410
5
P R O C E E D I N G S
Call to Order
DR. WOOD: Let's get started and welcome
back to another day. We are going to begin as on
the agenda seeing we worked late last
night.
A couple of housekeeping things
first. As
they say in the movie theater, please
turn off your
cell phones. We don't have the one that
sort of,
you know, spars you into space if you do
that, the
ejector seat, but then please don't
answer your
calls in here, so we don't have to hear
the
beginning of your conversation.
Kimberly, are you going to read
the
conflict of interest? Okay.
Go ahead.
Conflict of Interest
Statement
MS. TOPPER: The following announcement
addresses the issue of conflict of
interest with
respect to this meeting and is made as
part of the
record to preclude even the appearance of
such.
Based on the agenda, it has
been
determined that the topics of today's
meeting are
issues of broad applicability and there
are no
6
products being approved. Unlike issues before a
committee in which a particular product
is
discussed, issues of broader
applicability involved
many industrial sponsors and academic
institutions.
All special government employees have
been screened
for their financial interests as they may
apply to
the general topics at hand.
To determine if any of the
conflict of
interest existed, the agency has reviewed
the
agenda and all relevant financial
interests
reported by the meeting participants. The
Food and
Drug Administration has granted general
matter
waivers to the special government
employees
participating in this meeting who require
a waiver
under Title 18, United States Code
Section 208.
A copy of the waiver statements
may be
obtained by submitting a written request
of the
agency's Freedom of Information Office,
Room 12A-30
of the
Because general topics impact
so many
entities, it is not practical to recite
all
potential conflicts of interest as they
apply to
7
each member, consultant, and guest
speaker. FDA
acknowledges that there may be potential conflicts
of interest, but because of the general
nature of
the discussions before the committee,
these
potential conflicts are mitigated.
With respect to FDA's invited
industry
representative, we would like to disclose
that Dr.
Annette Stemhagen is participating in
this meeting
as a non-voting industry representative
acting on
behalf of regulated industry.
Dr. Stemhagen's role on this
committee is
to represent industry interests in
general, and not
any one particular company. Dr. Stemhagen is vice
president of Strategic Development
Services for
Covance Periapproval Services, Inc.
In the event that the
discussions involve
any other products of firm not already on
the
agenda for which FDA participants have a
financial
interest, the participants involved and
their
exclusion will be noted for the record.
With respect to all other
participants, we
ask in the interest of fairness that they
address
8
any current or previous financial
involvement with
any first whose products they may wish to
comment
upon.
Thank you.
DR. WOOD: Thank you.
Let's go right to the first
speaker, Dr.
Platt, who is going to tell us about
observational
studies.
Interpretation of Observational
Studies of
Cardiovascular Risk of
Nonsteroidal Drugs
Richard Platt, M.D.,
M.S.
DR. PLATT: Thanks.
The framers of the
meeting thought it would be useful at
this point to
have a discussion about observational
studies to
put us all on the same page.
There was a view by some that
the
expertise around the table might be
uneven and it
would be worthwhile to have some
discussion about
some of the basics. It is clear that that is not
the case.
I realize that a number of the
people here
have written a book and several of my
teachers are
9
here, so to that extent, I think we can
either make
this a quick discuss or use this as an
opportunity
for a real interactive discussion,
because there
are some hard questions here and no
matter how we
sort we out, we are going to be left with
less than
in the way of firm answers than we would
like.
I also understand that there is
a point of
view that says that there are lies, damn
lies, and
observational studies, so part of what I
think is
worth doing is using this time maybe to
take our
temperature about whether and under what
circumstances we can put weight on
observational
studies.
We saw a version of this slide
last night
actually in the last presentation about
why perform
observational studies at all, because I
subscribe
to the general view that all things being
equal, a
clinical trial, a randomized trial is
more
credible, provides more information than
an
observational study.
The problem is all things
aren't always
equal and so there are reasons to ask
what we can
10
learn from observational studies.
I think the most important of
them is no
matter how well a clinical trial is
designed, the
individuals who are recruited and
consented to a
clinical trial are inherently going to be
different
from the actual population of users, and
if we want
to understand how an agent performs among
real
users in the way they actually use the
drug, then,
I think there is no escape but to look to
observational studies.
Additionally, observational data
is by
definition there, so when a pressing
question
arises, sometimes observational data is
the first
way we can get insight into the
relationship
between the drugs we care about and the
exposures.
I think in that regard, these
studies can
often be thought of as helping us
identify the
areas in which it would be most fruitful
to invest
in full-blown randomized trials. We will never
live in a world where we are able to do
all the
randomized trials we care about.
I know that Charlie Hennekens'
landmark
11
randomized trial of aspirin was preceded
by, as I
recollect Charlie, a large number of
observational
trials, it made you think that it was
reasonable to
do those randomized trials, so
observational
studies can be useful in that regard.
Finally, when we are talking
about trying
to understand effects that are relatively
unusual,
we stress even the largest clinical
trials. We
talked yesterday about the fact that the
most
recent drug approvals have used much
larger
populations in the NDA phase than had
been studied
in the old days, and yet they are still
small
compared to the numbers needed to parse
out
relatively small differences.
There are a lot of different
kinds of
observational trials. I have listed a few of the
most common. The ones between the lines here are
the ones that are really the subject for
discussion
here.
Tom Fleming made the absolutely
correct
and somewhat counterintuitive point that
it is
often more difficult to do good
observational
12
studies of relatively common outcomes
than rare
ones, and because of that, the group of
studies
that I think at least are reasonable to
consider
for looking at relatively common outcomes
are
case-control studies, nested case-control
studies
and cohort studies.
We have examples of each in the
materials
that have been handed to us. The study by Kimmel
is a pretty traditional case-control
study. The
studies by Ray are cohort studies, as is
the Aramis
study.
The study by Dave Graham, the Solomon study
are nested case-control studies.
Just as a quick reminder, the
distinguishing feature of cohort studies
is the
fact that the study population is defined
on the
basis of whether people are exposed to
the drug or
not, and then we look forward to what
happens to
them.
In that way, they are exactly comparable to
clinical trials, with the big difference
that the
assignment to drug is not randomized.
The strengths of those compared
to
case-control studies are you have a
reasonable shot
13
at the outset of selecting individuals
who are
representative of the group that you are
trying to
study, and if you organize the study
properly, you
have a reasonably good chance of getting
unbiased
exposure assessments.
The weaknesses, particularly of
observational cohort studies is that just
because
individuals had the right drug exposure
at the
outset, they may change that. You can deal with
that with an intention-to-treat design,
but you pay
for a price for that, and in
observational studies,
loss to followup is a big problems.
We are particularly plagued by
that
because the large majority of the
observational
studies we are working in are ones that
use
administrative data from one sort of
health plan or
another, and individuals move in and out
of health
plans, so that it becomes difficult to
follow them
over time.
Case-control studies, remember
are ones
that start with individuals who have the
outcome we
care about, myocardial infarction or
myocardial
14
infarction and sudden death, and compares
them to
individuals who haven't had that
experience, then,
you look back and ask what their drug
exposures
are, the reasons for doing those studies
are that
they are, first of all, very efficient
studies.
You don't have to study
thousands and
thousands. You can study as many cases as
you find
and a reasonable number of controls, and
you can
look back and classify exposure however
is most
useful, and that is a very convenient and
versatile
feature of case-control studies.
The big weaknesses are that it
is very
hard to assure oneself that the cases and
the
controls are really representative of the
populations that you care about, and for
conventional case-control studies, for
instance,
the study by Kimmel that we are going to
look at,
it takes a lot of work to be sure that
people who
know what they have already experienced
an MI don't
differentially report their exposure to
the drugs
that we care about.
That can be for all sorts of
reasons and
15
it might not even be wrong, but the
individual who
has had an MI and might be just thinking
harder
about whether he or she had been exposed
to a drug
that we care about.
By the way, nested case-control
studies,
for instance, the study that David Graham
did is a
hybrid that really, in my view, draws
many of the
strengths from both designs, that is,
because
nested means the case-control study is
nested in a
defined population, so it has a lot of
the
strengths of cohort studies and some of
the
efficiencies of the case-control studies.
The differences between the
observational
studies and randomized studies are pretty
clear.
Randomized trials have the tremendous
advantage
that there is lots more reason to expect
the
treated and untreated groups to be
comparable to
one another.
There is a lot more opportunity
to be sure
that the outcome assessment and adherence
to
treatment are good or at least well
known, and we
have reviewed the difference for the
observational
16
studies.
I think it is worth making the
point that
there are a substantial number of
similarities
between observational and randomized
studies. Just
because we randomize individuals in
randomized
studies, it doesn't mean that the treated
and
untreated groups are comparable.
We talked about a study
yesterday that was
a randomized trial where there was a
substantial
imbalance in important risk factors. So, it is
incumbent no matter what kind of study
you do, I
think to look for comparability, and both
studies
have as potential weaknesses that there
are risks
of false positive results and doing
subgroup
analyses and multiple comparisons
increases that
risk.
We talked a fair amount about
that
yesterday, and both are at risk for false
negative
results.
That can be partly because the studies
may not be powered well enough either
because there
is insufficient sample size or
individuals aren't
studied for a long enough duration to see
the
17
biological effects that we care about, or
a
vulnerable group just isn't included.
That is a problem with both
kinds of
studies and I think all studies have to
be
evaluated on their own merits, so let's
just step
through the various places where
observational
studies might be into trouble or at least
the
things that need careful assessment when
we look at
these studies.
The first is are we studying
the right
outcomes. It is essentially impossible in
any of
these observational studies to use the
kind of
rigorous adjudication that is a hallmark
of the
randomized study, so I think we are going
to have
to ask ourselves are these outcomes good
enough.
The several kinds of outcomes
in the
studies that we have been asked to look
at are
hospitalized MIs. The case-control study by Kimmel
uses survivors. It had to use survivors because
they were collecting the exposure
information by
interview after the individuals had left
the
hospital, so if we care about all MIs,
then, that
18
study isn't going to tell us what we want
to know.
Some of the studies use MI and
out-of-hospital sudden death by linking
to vital
statistics records. I think that is probably the
closest we can get in observational
studies to the
intention-to-treat all outcome designs of
the
randomized trials, and some of the
studies use
composite designs.
You have to ask are these
outcomes
measured appropriately. Most of the studies that
we are looking at use some form of
automated
medical record or claims data that have
been, in my
view, reasonably well validated. That is, there is
a moderate literature showing that claims
data are
not so bad for studying acute myocardial
infarction. They have sensitivities in
the 90s and
positive predictive values in the 90s.
So, they are not perfect and I think we
will have to ask as we review the studied
can the
amount of uncertainty that we know exists
in those
account for the effects that we see, or
could they
obliterate effects that we would like to
see and
19
which aren't there.
My sense is that that is
probably not a
sufficient explanation to dismiss the
studies that
we are looking at. The issue of bias is
one that I
think always has to live as a sub-text,
but quite
frankly, in the studies that do outcomes
in the way
we have been describing, I don't think
that is a
serious problem.
For cohort studies, we have to
ask are we
studying the right population, and here I
think we
really do have to stop and ask
carefully. One is
are these people selected from the
population under
study.
I think in most of these examples, they are
reasonably representative, that is, a
study of the
people of
plan.
I think that the data systems
that are
used to identify the individuals in the
cohort are
good enough to give us reasonable belief that
we
are identifying either all the people or
a
representative sample of them.
I think there is a fair
question of
20
whether they are representative of the
larger
population. We could ask are health plan members
systematically different from the general
population of individuals who are taking
these
medications.
The range of studies we have
include
health plan members. I think that there is
reasonable information that they probably
are
representative, at least with respect to
the drug
myocardial infarction outcomes that are
studied.
Studies in Medicare and population-based
studies,
such as those in
reason to think that they are
representative.
But there is an important
consideration
about whether there are issues about the
way
clinicians practice in those setting that
might
have a serious impact on selecting
individuals. In
particular, to the extent that
formularies are
restrictive of, say, newer or more
expensive drugs
like the COX-2 inhibitors, but I think we
have to
ask very carefully whether the factors
that would
influence the prescribing of one class of
drugs
21
over another is likely to seriously
impact the risk
of these outcomes.
Additionally, if there are cost
differentials for these drugs, it may be
that there
is some form of self-selection that
causes
individuals who are sicker to receive these
drugs,
and I think that it is incumbent on us to
expect
that to be a problem in every one of
these
observational studies and to ask how well
do these
studies do in adjusting for that. I will circle
back to that in a moment.
I think we have to be concerned
about
whether we are studying people who have
had prior
NSAID exposure, in which case we would be
worried
about survivor biases, of finding the
individuals
who are relatively immune to these
problems.
Finally, there are study design
issues
about whether there are restrictions of
eligibility
that might importantly color the
data. For
instance, at least one of the studies we
are
looking at requires individuals to have
received at
least two dispensings of a nonsteroidal
agent in
22
order to be eligible.
That means that you have to
live long
enough to have two dispensings, so it
certainly
doesn't tell us anything about the early
effects of
these drugs, and it might in an important
way color
the results with regard to later
exposure.
There is an important question
which is
not unique to the observational studies,
which is
who are the right comparators. We had a number of
discussions about that yesterday. I think that all
the issues that we discuss with regard to
the
clinical trials are applicable here. In
particular, there is a lot of reason to
want to
compare to other nonsteroidal users
because that
gives the best chance of having a group
that is
similar with regard to underlying disease
status
and presumably risk of myocardial
infarction.
Similarly, it is possible to
say that if
you really care about COX-2 selective
agents, you
should compared one COX-2 selective agent
to
another.
That leaves us in the uncomfortable
23
situation of not knowing what is the risk
compared
to no use at all, so we have some
comparisons that
do look at non-users or at least remote
users, and
that has its strengths. It has the big weakness,
of course, of putting us at risk of
making
comparisons against groups that are
unrelated.
So, we are really talking here
of mostly
about a study like the Kimmel study, not the nested
case-control study. The other kinds of concerns
that raise red flags are the real concern
about
losing cases who make the group who are
studied
unrepresentative.
I would point out to you, for
instance,
that in the Kimmel study, only half of
the MI
survivors who were identified were
actually
interviewed and therefore part of the
formal
analysis.
We already talked about the
fact that
since that study was limited to MI
survivors, that
restricts us to a less serious set of
outcomes.
The other problem that really
bedevils
conventional case-control studies is
knowing
24
whether the group of people who are
selected as
comparators are really comparable.
I think that is one of the
reasons that
there is so much interest in doing nested
case
control studies, because at the end of
the day it
is really extremely difficult to satisfy
oneself
that controls really are appropriate.
Much of what we need to be
concerned about
in these studies is understanding
exposures. Part
of the issue is understanding how to
characterize
exposure.
This is both a strength and a weakness
of these studied.
You will remember I made the
point at the
outset that if we want to understand how
drugs work
in actual practice, that we have to do
observational studies. On the other hand, that
means we have to find a reasonable way to
characterize these drugs.
We talked yesterday I think
about all the
important issues of understanding whether
we had to
look at absolute dose or cumulative
effects or
whether the effects start early or
whether they
25
start late.
I think that the best of the
studies that
we are looking at tackle a number of
these issues.
I will mention in a minute some of the
ways that
these studies have gone about that.
I think in terms of
ascertaining exposure,
it is probably reasonable to put the most
reliance
on the studies that use administrative
databases of
pharmacy dispensing, but I will just make
the point
that we have to be clear that these studies
are
done in situations where we have reason
to expect
that the administrative databases are
correct.
I think all the studies we are
reviewing
are ones where the investigators were
careful to
know that the individuals really had a
drug benefit
that was operating at the moment, that
would likely
find the prescription drug exposures that
we care
about, but as a general proposition, you
can't
assume that that is the case.
Most health plans have some kind of
restrictions on benefits that might lead
individuals to change their benefit
status, so
26
there would be periods of time when we
might know
that they had an MI, and we might not
know that
their drug exposure is at the moment.
I will return to a point that
we touched
on yesterday, which is that although
almost all of
the studies that we are talking about
report their
results as relative risks, a 2-fold
increase in
risk, a 70 percent decrease in risk. What we
really care about is the absolute
difference in
risk.
So, that is not different between
observational studies and randomized
studies, but I
think it is really a critical piece of
our thinking
about the problem that we are dealing
with.
The second thing that is just
worth
recalling is that when we talk about a 95
percent
confidence interval, that our expectation
about
where the true value lies is not
uniformly
distributed over that interval.
Our best guess about where the
true value
lies is around the point estimate, and if
that
point estimate is wrong, the large
majority of the
27
uncertainly is pretty close to that point
estimate,
so that it is particularly not helpful,
in my view,
to pay enormous attention to p values.
The difference between a p
value of 0.05,
as shown here, and a p value of 0.01 and
a p value
of 0.13 is not all that enormous in terms
of the
biological impact.
I think one of the things that
is a
particular concern that we need to pay
attention to
in these studies is the fact that it is
easy to
look at a lot of different comparisons,
and to the
extent that we do that, we are going to
have to
just be careful to know that the strength
of any
one comparison is weaker than it appears
to be.
For instance, this is a quote
from one of
the studies that we are looking at. We undertook
an observational study examining the
association
between rofecoxib, celecoxib, other
nonsteroidals
and myocardial infarction.
Well, there is no primary
hypothesis
there, and the results for all of the
nonsteroidals. They are all interesting to look
28
at, they are all associated with p
values. Those p
values are all relatively too extreme
given the
fact that there are so many comparisons.
It is a problem for randomized
trials. We
talked about subgroup analyses. It is important to
do those studies, those subgroup
analyses, but
absent having specified a principal
hypothesis at
the outset, I think that we have
difficulties in
knowing how much weight to put on any
particular
one.
We talked a lot about
confounding. That
is one of the most important concerns in
randomized
trials.
I know you all know what confounding is.
It wasn't obvious to me when I was making
these
slides that everyone knew that, but the
example, so
that we have it in mind is if what we
know is drug
A versus drug B, and MI or no MI, and we
don't take
into account important confounders, we
can get
importantly incorrect results.
So, here is an example of an
aggregate
analysis with a relative risk of 1.5
among 2,000
people who are exposed to two drugs. If you break
29
it apart and see that in the high-risk
group, drug
A accounted for 80 percent of the
exposure, and in
the low-risk group, drug B accounted for 80
percent
of the exposure, you see that in each of
those two
categories, the high-risk group and the
low-risk
group, that, in fact, there is no
association
between drug and outcome, but you have to
take them
apart to do that.
Well, the good news is if you
know what
the confounders are, and you have
measured them
accurately, it is possible to adjust for
them, and
all of the studies we are looking at do a
pretty
job of adjusting for the confounders that
we know
about, so I guess one of the questions is
how well
do they do at identifying the important
confounders.
I would say not bad on a lot of
that.
That is, if you take, for example, the
Graham study
or the studies that Wayne Ray did in
Medicaid, there are a number of
strengths. I will
sort of stop and back up on the things
that make
these look like relatively more credible
studies in
30
the scheme of the factors that we care
about.
They are inception cohorts of
nonsteroidal
users, that is, they are individuals who
had to
have been members of the health plan for
at least a
year before they received their
nonsteroidal.
There was a lot of information
about their
underlying medical status that was
available to the
investigators using both claims data and
medical
record data to ascertain cardiovascular
disease
along a number of dimensions, utilization
of
procedures like surgery or angioplasty or
diagnostic procedures that are intended
to find
cardiovascular disease, hospitalizations,
emergency
room visits, and a substantial amount of
information about the medications that
these
individuals took that was related to or
plausibly
related to cardiovascular risk factors.
Those large number of factors
were used to
create separate risk models using only
the
unexposed, and then to use those risk
models to
create risk indexes for the individuals
to use as
an adjuster for underlying cardiovascular
risk.
Is it perfect? No. Is
it pretty good?
It seems to me that it meets the sniff
test of
saying that it has a reasonable chance of
31
identifying important confounding.
Unfortunately, there are a
number of
important confounders for which health
care systems
typically don't have good data, like
smoking, OTC
NSAID use, obesity, family history, and
those are
typically much more problematic.
Some of these studies have
worked pretty
hard to try to either deal with it or
understand
whether it could be an important
problem. One of
the handouts we had, for instance, was
the study by
Schneeweiss and colleagues who looked
back at one
of the studies by Solomon that was
performed in the
Medicare data set, and asked how
important could
these unmeasured confounders be.
They actually had access to
information
from the Medicare Beneficiary Survey that
asked
representative Medicare beneficiaries
detailed
questions about many of the things that
we would
are about. They weren't the people who were
32
involved in that case-control study, but
if you
assume that the beneficiary survey,
members were
representative and they gave plausible
answers, it
is possible to extrapolate back to the
source
population, and the take-home message from
that
work, the answer didn't change very much,
which is
really what we want to know, not sort of
the
absolute difference, but whether those
unmeasured
confounders are important enough that
they could
cause a difference.
I think we still have to be
concerned at
the end of the day, we still have to be
concerned
about residual confounding as a
potentially
important problem.
One way I think that we can
draw relative
assurance from that work of adjusting for
confounding is to ask how much did the
estimate of
risk change between the unadjusted and
the adjusted
result.
I think there is a world of
difference
between an unadjusted result of 10 and an
adjusted
result of 1.5, and having an unadjusted
result of
33
1.6 and an adjusted result of 1.5. The former, I
think the reasonable assumption is we
arguably
haven't been able to deal with
confounding in a way
that would let us believe that 1.5 means
something.
I think there is a much
stronger case to
be made when adjusting for important
confounders
that we know about doesn't change the
risk estimate
very much, that that is a relative more
credible
answer.
Having said that, I think that
observational studies are best at finding
relative
risks that are more than 2. I think that I would
pay some attention to relative risks of
1.5. I get
very nervous about adjusted relative
risks of 1.2.
That doesn't mean that they are
not right
and I don't ignore them, but if we ask is
that for
sure the answer, my response to that is I
am just
less certain about that.
I think we are always left at
the end,
while we spend a lot of time thinking
about and
adjusting for confounding, and I think we can
do a
pretty good job of that, it is much
harder to
34
adjust for misclassification, and it is
essentially
impossible to adjust for bias.
So, I think one of the things
we have to
ask about is are there plausible sources
of
misclassification and bias, and if there
are, in
which direction do they work and would
they
seriously change our interpretation.
We talked about the fact that
absolute
differences are the important ones that
we care
about.
We have already started to look at data
that talks about person level risk and
population
level risk, so beyond saying that at the
end of the
day, I think these are the answers that
we really
need to talk about, not about relative
risk.
Personally, I think that we
need two kinds
of answers. One is what is the information that
patients and their physicians need to
have to make
decisions for them personally about
whether to
accept certain kinds of treatments in
exchange for
certain kinds of anticipated benefits.
I think there is a population
level
concern that we have to have that emerges
from the
35
same set of analyses, but takes on a
different
form.
So, you will be pleased to know
that I am
wrapping it up now, and I would say that
both the
cohort and nested case-control designs,
which are
the bulk of the observational studies
that we are
looking at, are relatively strong ones
and I think
deserve the committee's real attention.
I am sorry that not every one
of these
studies prespecified a primary hypothesis
that we
can attend to, but we should whenever
possible do
that.
Even though we don't find important effects
in some of these studies, I think it is
important
to recognize that they don't exclude one.
As I have said, I am least
certain about
attaching great weight to relatively
small excess
risks even understanding that when they
are
extrapolated to a large population, they
could
account for very important public health
problems.
Finally, I would say that the
things that
support the studies' conclusions are the
fact that
when we do subgroup analyses and look for
36
dose-response effects, that they
strengthen the
cause-effect relationship, and I think
that there
is reason to look for consistency across
studies.
I take the point that was made
yesterday
that it is possible that a dozen studies
of
naproxen could all have the same
underlying bias
that shift the point estimate in the same
direction, but it is not so clear to me
what that
bias is.
So, I think that we would have
to have a
reasonable idea of what might explain
consistent
differences across studies and ask if
they are of
sufficient magnitude to explain
that. As I say, I
am not clear that there are those kinds
of biases.
I think we have to be cautious
about the
fact that residual confounding bias and
misclassification are all issues with
these
studies.
So, I think that while they add to our
discussion, they have to be considered in
light of
the fact that they are imperfect
vehicles.
Thanks.
(Applause.)
DR. WOOD: Thanks very much.
Let's just go straight on to
the next
speaker and then we will take questions
for Dr.
37
Platt after David Graham's talk.
The next speaker is Dr. David
Graham from
the FDA.
Review of Epidemiologic
Studies on
Cardiovascular Risk with
Selected NSAIDs
David Graham, M.D.,
M.P.H.
DR. GRAHAM: Good morning. Today, I will
give a review of epidemiologic studies
and
cardiovascular risk with selected
NSAIDs. I will
be evaluating epidemiologic data from the
published
literature plus two currently unpublished
studies
that I have evaluated.
My focus will be on providing
estimates of
risk of acute myocardial infarction in
the setting
of the use of COX-2 selective NSAIDs or
naproxen,
although I will have some comments in
light of
yesterday's discussion about other NSAIDs
on those,
as well.
The methodology was to do a
PubMed search
38
by specific NSAIDs and then cross-check
the
citations in those articles to see if
there are
other articles I had missed.
I would also like to take this
moment to
thank Dr. Crawford for his leadership in
making it
possible for me to present some of our
preliminary
data from a study in California Medicaid,
which Dr.
Gurkiepal Singh from Stanford and I
recently
completed.
Before I get into the substance
of my
talk, I just want to comment a little bit
on excess
cases and projecting to the national
population
what was the impact of rofecoxib use, and
I am
doing this for two reasons - one, because
it has
been a source of controversy and
concern. We cite
a number in a paper that I and others
have
published from Kaiser Permanente in which
we made
an estimate of the impact of rofecoxib
use.
Tomorrow, FDA will present its
estimation
of the number harmed by rofecoxib,
modeling
randomized clinical trial survival
curves. A
couple of things I would like the
Committee just to
39
be aware of when they see that data
tomorrow. It
assumes a grace period at the beginning
of use that
is based on the VIGOR study and the
APPROVe, 6-week
grace period in which there is no
difference in MI
or increased risk of MI, and the first
six weeks of
high-dose use with the first 18 months of
low-dose
use of rofecoxib.
As I will show later in my
talk, I believe
that this is unreliable due to low
statistical
power early on, because we are only
talking about
in each of these studies a handful of
cases early
on in the study. Two or three cases of MI and wide
confidence intervals, you could have
divergence of
the curves very early.
The epi studies, however, that
I will
present will show that there is a 3- to
50-fold
more events to work with, more
statistical power,
and it suggests a different outcome.
The second is, is that the
patient
enrolled in randomized clinical trials
are
generally healthier than patients in the
real
world.
So, if you are going to model what is the
40
number of people who have been harmed in
the
population, you have got to assume what
is the
background rate that you are modeling off
of.
If you use a background rate
from healthy
people to model what is happening in the
population
of people who really aren't so healthy,
who have a
higher background rate, you will
underestimate the
actual population impact.
So, in any event, now on to the
substance
of my talk.
The next three slides provide a
very dense
overview of the major features of each of
the
epidemiologic studies that I
reviewed. I am
looking at COX-2 usage in acute
myocardial
infarction.
You can see that they are
grouped in
several groups. The top three studies I consider
from an epidemiologic perspective to be
stronger
studies to have been done better. In terms of the
things that Dr. Platt just talked about,
I thought
that these studies were the stronger
studies.
The next two studies from the
published
41
literature I thought were less strong,
and I will
describe why. Finally, I have separated out these
last two studies, one submitted by Merck
to the
FDA, performed by Ingenix, and the other,
the
Medi-Cal study that Dr. Gurkiepal Singh
and I have
recently completed of unpublished
studies, so they
are separated out from the group.
You can see we are talking
about different
source populations, and so if we can see
consistency of results across different
populations, different age groups, and
different
study designs, I think that that adds
support to
the notion that there is a real effect.
If we begin to see that there
is a lack of
consistency across the studies, then,
many of the
things that Dr. Platt talked about before
need to
be considered sort of the individual
level of the
studies, so what might explain why one
study shows
something and another one doesn't.
This next slide shows the case
definitions
and in a number of cases that we were
working with
to come up with the relative risk
estimates that I
42
will show you.
All of the studies began with
hospitalized
acute myocardial infarction. Several of the
studies were able to link members of
their base
cohorts to death certificate data to
identify
sudden cardiac deaths, as well. So, those are the
ones that have the +Sudden Cardiac Death.
The asterisk next to the Kimmel
study is
to remind me and to remind you that the
Kimmel
study was based on nonfatal MIs
only. By their
design, they had to interview their cases
in
person, so the patient had to survive
their
myocardial infarction to be
interviewed. So, there
are those differences in study design.
In the end, what is very
important in an
epidemiologic study in dealing with this
issue I
think in particular, is what is the
statistical
power of the study, and that is driven
primarily by
the number of events in the exposed group
that we
have to deal with.
So, in this column here, you
will see the
total number of cases of myocardial
infarction that
43
were identified in each of the studies. The
asterisk next to the Ingenix study 628 is
to remind
me that in that study, they identified
about 1,700
MIs in total, but they excluded 1,100 of
the MIs
because they occurred in people who
weren't exposed
to an NSAID at the time of the myocardial
infarction. So, as a result, they left them out,
because in the previous slide, when we
look at the
reference group, most of these studies
used either
non-use or remote use as the comparator. The
Ingenix study used active treatment with
either
diclofenac or ibuprofen.
I would like to say one thing
about
reference groups. Dr. Platt brought it up before.
In this issue, I don't believe that there
is a
single best or optimal reference
group. What you
really want to do is get as close as you
can to a
placebo group that has been randomized
and has all
the risk factors of the people who are
getting the
drug.
In the observational world we can't get
there, and so at the end of the day, if
you want to
44
do a study, you are in a sense forced to
pick among
the least evil of that you think, and
then it has
to do with how you define things.
So, non-users, for example,
could be
viewed as being close to the placebo
group, they
are not getting the drug. The problem is people
who don't use drugs tend to be healthier
than
people who do use drugs, so that raises a
host or
problems.
Yes, we can try to adjust for
confounding
and the like, but you are still left with
that
concern that they may be, in some way
that we can't
measure, different from the people who
get the
drug.
In the study I did, and in
several other
studies that people have done, we opted
to use
people who had been treated with NSAIDs in
the
past, but weren't currently taking an
NSAID at the
time of the event or the study, the
reasoning there
that whatever the selection factors are
that lead
to a patient getting an NSAID, that some
of those
selection factors are there in people who
45
previously received NSAIDs.
That is still not a perfect
group, though,
because you could argue that patients who
are no
longer taking NSAIDs might be healthier than
people
who are currently taking NSAIDs.
Finally, the problem that is
posed by
using an active comparator. If you have an active
comparator, and I am comparing another
drug to an
active comparator, and I see a
difference, I don't
know what it means. I need some place to anchor
the result, and for that reason, although
none of
them are perfect, I believe that the
non-use and
the remote use analyses at least give us
a way of
pegging results, and if we want to
compare one drug
to another drug, if we had that common
reference
point, at least it allows us to
accomplish that.
The one other thing I would
like to point
out about the number of cases is that for
rofecoxib, especially at the high doses
of
rofecoxib, most of these studies had
relatively few
exposed cases. The exception is the
Medicaid study where we had 157 exposed
cases to
46
the higher dose of rofecoxib.
Now, this is a very busy slide
and I won't
spend a lot of time going over it, but I
will be
happy to answer questions later.
Basically, before we heard
there are
unmeasured risk factors in automated
databases that
frequently can't be accounted for,
aspirin use and
smoking are among the most common. So, you can see
here that most of these studies, that
information
isn't obtainable.
Kimmel was able to get both
because they
interviewed the patients, the cases and
the
controls.
In the Medi-Cal study, it turns out that
aspirin is reimbursed, and so we have a handle on
it there.
In the Graham study, a survey
of controls
was done to see what these unmeasured
factors might
look like in the source population. The Solomon
study did the same thing, relying on the
Medicare
Beneficiary Survey that Dr. Platt talked
about
before.
Important limitations I think
that need to
47
be highlighted are that in the Mamdani
study, they
excluded patients who had less than 30
days of
NSAID use, so the survivor bias Dr. Platt
talked
about before, in my view, is big concern with this
study, and for that reason I ranked it in
sort of
that category of low quality studies.
In the Kimmel study, as Dr.
Platt also
mentioned, there was low participation
rate.
Basically, half of the cases and half of
the
controls who approached volunteered to be
in the
study.
More importantly I think in that study, and
it's unfortunate, is that there was what
I would
refer to as the potential for, in quote
"reverse
recall bias."
Normally, with recall bias, we
think oh, I
have had a heart attack, I am going to
remember
more efficiently what happened to me
immediately
before the heart attack compared to some
control
where I say to the control what were you
doing four
months ago on this particular day.
That is the classic recall
bias. This
situation I think had what I would
describe as
48
reverse recall bias. They interviewed the people
who had heart attacks within four months
of getting
out of the hospital - what happened to
you the day
and the week before you had your heart
attack four
months ago.
For the controls, they call
them on the
phone and they way what happened to you
yesterday
and the week before, so it is actually
the reverse.
The controls actually would have better
recall of
what they were actually doing than the
cases
potentially, and we will see how this is
reflected
in some of the results.
Finally, with the Medi-Cal
study, I think
the single greatest concern for the
committee in
considering these data (a) that it is
preliminary
data, and (b) that this is a new database
for
research purposes.
For that reason, I am just
including a
slide to orient people to that. The other
databases are ones that have been used
before.
This is a database that only in the last
two years
has come online to be sort of a quality
sufficient
49
to begin contemplating doing studies.
Its strengths are that it is
very large,
it
captures aspirin use, it doesn't censor people
by age.
It combines Medicare coverage when you go
over the age of 65 with the prescription
benefits
of Medicaid, so you get the drugs and the
outcomes.
Matching has been done to
multiple cause
of death tape, so that we have death data
in this
database up through 2002. We didn't include it in
the data I will show today because we
really want
the information up through 2004.
Once people get into Medicaid or Medicare,
they don't tend to drop out. The limitations are
that we can't get medical records, and
that is
something to understand, and that is a
very
complicated database. Dr. Singh from Stanford who
is the principal investigator for our
Medi-Cal
work, and who has worked to bring this
database
online, spent two years putting things
together and
working out the kinks in it before
contemplating
doing research with it, so at least you
understand
the limitations of that.
There is always the concern
about
unmeasured risk factors and Dr. Platt
talked about
that.
I want to review for you very briefly some
50
of the evidence from the published
literature where
efforts were made to look at what
unmeasured
confounding looked like and did it differ
across
NSAID type.
In our study using Kaiser
Permanente data,
we did a survey, a random survey of
random sample
of controls, and we looked at aspirin
use, smoking,
and over-the-counter NSAID use. You say see by
NSAID that there really was not significant
or
substantial differences in the
distribution of
these risk factors.
So, if they don't vary in the
control
group, they can't really confound that
observation
that you see very much.
In the Solomon study, these are
the data
from the beneficiary survey. Dr. Platt already
mentioned a further analyses of these
data that
showed that the actual impact of all
these
unmeasured confounders on the measure of
the
51
relative risk at the end was measured in
the
hundredths of an odds ratio, so if the
odds ratio
was 1.34, adjusting for these things and
projecting
it out would change it to maybe 1.35 or
1.33. We
are talking about minuscule differences,
not
qualitatively important differences.
Finally, in the Kimmel study,
they also,
through their interview, were able to see
that for
most of these factors, there was
similarity across
NSAID groups except for current smoking
where the
rofecoxib group had much lower current
smoking than
any of the other NSAID groups, but for
past
smoking, it was more than the other NSAID
groups or
the remote groups, and if you added these
two
together, the rofecoxib was very similar
to these,
but the celecoxib group had more smoking.
My own conclusion from this is
that yes,
it is possible that some of these
unmeasured risk
factors could be influencing the
results. I don't
think that there is strong evidence that
there is a
systemic bias that would sort of lead to
interfering with trusting the results and
thinking
52
that these factors are confounding the
observations
that we see.
So, first, I will talk about
rofecoxib,
then I will talk about celecoxib, then I
will talk
about valdecoxib in terms of
epidemiologic data.
These studies on the left, with
their
reference groups, are the ones that
looked at
myocardial infarction with
rofecoxib. What I have
shown is for all doses and where it was
present
less than or equal to 25 milligrams and
over 25
milligrams, what the fully adjusted odds
ratio and
95 percent confidence intervals were.
These studies varied in the
extent of
adjustment that they did. The Ray and the Graham
studies each adjusted for about 30
cardiovascular
risk factors. The Solomon study was a somewhat
smaller number, Mamdani was a somewhat
smaller
number.
Kimmel, they adjusted for somewhere in the
20s, the Ingenix study somewhere in the
20s, the
Medi-Cal study adjusted for about 40
cardiovascular
risk factors.
What you can see is when you
look across
53
the All Doses is that, in general, the
point
estimates were elevated and for many the
95 percent
confidence intervals excluded 1.
More importantly, though, is
looking at
the low dose and the high dose data
because we know
from the clinical trials data, and we
would suspect
it on just pharmacologic grounds, that if
there is
an association that it might be worse
with the
higher dose than with the lower.
So, four studies provide us estimates at
the low and the high doses, the Wayne Ray
study and
our study from California Medicaid, and
then the
two unpublished studies, one from Ingenix
and the
other from California Medicaid.
We see there that in three of the four
studies, there is an elevation in the
point
estimate.
In the Graham study, it included one.
When we look over 25 mg, we see greater
consistency
although in the Ingenix study, there is
this
paradoxical finding of sort of basically
a neutral
relative risk. I don't have an explanation for why
that happened, but it makes me concerned
to some
54
extent about what was going on in that
study,
because it is a result that goes in a
very
unexpected direction.
What I would like to point out,
because I
will come back to it again, is that when
we are
dealing with drug safety, and the goal
now is what
risk can I exclude, if my job is--now I
am not
talking about efficacy anymore, what I am
talking
about is safety--if my job is to protect
the public
from harm, what risk can I exclude based
on the
data that I have, I believe that is much
more
relevant to look at the upper bound of
the
confidence interval than the lower bound.
What traditionally happens is
we look at
the lower bound of the confidence
interval and we
say if it includes one, there isn't a
problem, but
the biggest reason, as Dr. Platt showed
in his
previous slide, for a wide distribution
and a wide
confidence interval in your study, is
that the
study doesn't have enough statistical
power to get
you a narrow enough confidence interval
to say that
you have the 95 percent certainty that
you want.
So, if your mission is above
all else I
want to do no harm, that I want to
protect patients
from harm, then, based on the data you
have, I
55
would submit that the upper bound of the
confidence
interval provides greater assurance to
patients,
and then if you are going to compare a
benefit to a
drug, that you might want to consider
that benefit
against that upper bound of the
confidence
interval, because that is compatible with
the data.
In any event, that is my view, and not
the FDA's.
This is a slide from California
Medicaid.
It is preliminary data and I wanted to
present it
to you, because what it shows is a
dose-response to
rofecoxib from 12.5 mg up to and through
50 mg.
You can see that we have very
wide
confidence intervals for some of them,
and that is
a reflection of the limited number of
cases, but I
want to point your attention to the very
narrow
confidence intervals in the 12 to 25 mg
and in the
25 to 50 mg, just to point out that in
the previous
slide here, where we are talking about
what are
these point estimates, that now you can
what we
56
have done is we have fleshed them out a
little bit
more.
Another comparison that I think
is
important to consider, certainly it was
for us,
when we did our study in Kaiser
Permanente, was at
the time there were two COX-2 selective
inhibitors
on the market, celecoxib and rofecoxib.
The bigger study raised a
question about
high-dose rofecoxib. Our question as researchers
was, and public health scientists, was,
well, let's
suppose that rofecoxib increases the risk
of
myocardial infarction.
We don't know that it does, but
let's
suppose that it does, what about
celecoxib, because
it actually had a larger share of the
market, and
if it turned out that these drugs have a
benefit,
and that benefit is worthwhile, then, it
would make
more sense from a practical perspective
to use the
drug that had a better safety profile.
So, to us, it was very natural to want to
compare rofecoxib to celecoxib, and so
several of
the epidemiologic studies felt similarly
and in
57
their design they included that analysis,
and some
of them it was, as Dr. Platt said, part
of a we are
going to make comparisons of everything
against
everything.
The Solomon study, for example,
did that.
They did not state in that study what
their prior
hypothesis was. In our study, we did
state it. I
mean yes, in a sense we had multiple
comparisons,
but we were interested in two different
things. We
were interested in rofecoxib versus
remote use, and
we were interested in rofecoxib versus
celecoxib,
but we thought it beforehand and we
planned that
analysis.
But in any event, what we say
is, when you
look at the all dose analysis, in all of
the
published studies, rofecoxib increased
the risk
compared to celecoxib. When we looked at low dose
rofecoxib, we see the increased
risk. When we look
at the high doses of rofecoxib to
celecoxib, again,
we see the same pattern.
Dr. Platt, in his talk before, talked
about relative risks, risk differences,
individual
58
risk, and population risk. The next two slides are
intended to address this at the level of
the
individual and at the level of
population.
What I have done on this
slide--and these
slides now, no one should interpret this
as meaning
this is what actually happened in the
population--the next slide is going to
have numbers
on it that are for illustrative purposes
only, to
help the committee understand what does a
relative
risk of 1.3 translate into at the
individual level
and at the level of population.
Your typical COX-2 user is
somebody in
their 60s who has several other health
problems, so
I went to the National Center for Health
Statistics
and got the myocardial infarction rate
for 65- to
74-year-old men in the United
States. That rate
turns out to be 1 per 50 per year.
What I did is I took that as
the
background rate and I said if I have an
individual
using this drug with that background rate
and then
I applied to that person the relative
risks or odds
ratios found in these studies that are
shown in the
59
previous slides, what would the excess
risk to the
person be, sort of what would that risk
difference
translate to for the individual.
For example, in the Ray study,
if you
remember, for 25 mg or less, the odds
ratio was
1.02.
Basically, it doesn't change. If
we based
it on the point estimate, that 0.02 would
translate
to 1 out of 2,500 in a year increased
risk of heart
attack.
Another way to view that number
is, is
that is the number needed to harm. If I treated
2,500 65- to 74-year-old men for a year
with
rofecoxib, and the rate was 1.02 that Ray
found,
treating 2,500 patients would produce 1
extra heart
attack.
Now, with the other studies
that found
higher estimates for the lower doses of
rofecoxib,
you can see that the number needed to
harm ranges
from about 90 to 200. That is saying for every 90
people to every 200 people I treat with
low-dose
rofecoxib, I would generate 1 other case.
For high doses, because the
relative risks
60
were higher, the number needed to harm
becomes
lower.
I have also shown it based on
the upper
bound of 95 percent confidence interval
to show you
that based on the data we have at hand,
these are
the excess risks that are consistent with
the data,
and from a public policy perspective,
from a public
health perspective, that is what I react
to, and
when I want to see a benefit and say does
benefit
exceed the risks, well, I want to know
what is a
real benefit in the population in terms
of reduced
hospitalization, lives saved, and does
that benefit
exceed what I can say is possibly the
risk of these
products.
At the population level, now we
have gone
from an individual. Remember in the Wayne Ray
study we said it is 1 out of 2,500. Well, that
would translate to 400 additional cases
of heart
attack if we treated a million men who
were 65 to
74 years old, and we treated them with
rofecoxib
low dose for a year.
With the others, you can see
that those
61
relative risks that might not look so
impressive,
that 1.23, that 1.30, that 1.4, that it
projects
out to a substantial number when you
multiply it by
the large number of people who use these
products.
For high doses it ends up being
even
greater, and then if we focus on the
upper bound of
the confidence interval, we again see
that the
numbers are larger still. This very high number in
our study was the result of our having
low
statistical power in addressing the high
dose
rofecoxib.
One other question that I think
is
important to consider is when does the
risk of
myocardial infarction with rofecoxib kick
in. Now,
we have seen data yesterday presented by
both FDA
and by Merck of various survival curves.
We saw the bigger curve that
showed the
separation after about 6 weeks with an
overall
relative risk of about 5. We saw, for the APPROVe
study, this close overlapping line at
about 18
months, and then they diverge with an
overall
composite hazard ratio of about 2.
I would submit to the committee
that the
reason for the failure of these studies
to show
divergence of the line shortly after the
drugs are
62
used are low statistical power, that they
just
don't have enough events to show it, and
as a
result, you can interpret because of the
low
statistical power you basically--how to
describe
it--you presume that there is nothing
there, and
you err on the side of the drug rather
than erring
on the side of what could the risk be to
the
population.
If you really want to know what
is going
on in the population, then, you want to
reduce the
uncertainty. The more uncertainty you have, if you
act basically on the lower bound of that
confidence
interval, which is what you are doing
when you are
saying the risk doesn't begin until 18
months, you
are basically saying that the absence of
evidence
is evidence of absence.
I would say that in safety,
what it is, is
you just don't have enough power.
Looking at the epidemiologic
studies, I
63
think that we have evidence to suggest that
the
risk begins much earlier. I will point it out, and
you guys and women can consider it for
yourselves.
In the Graham study, when we
looked at low
dose and high doses of rofecoxib, 50
percent of our
cases at the low dose and at the high
dose had used
at the time--remember these are inception
cohorts,
so these people, their total use, this
was 1.8
months, this was 2.7 months--50 percent
of our
cases occurred within 2 to 3 months of
starting the
drug.
That is a lot of power and that
really
speaks against the notion that the risk
is
backloaded, you know, it is for the low
dose, that
the risk doesn't happen until after 18
months.
Nobody in our study was on rofecoxib for
more than
about 15 months. I think that was the longest
duration of use we had in our study.
Now, in the Solomon study, they
looked at
the low dose and the high dose, and they
presented
data in several ways. One is that they grouped
things in 1 to 90 days, and what they
showed was
64
that for both the low dose and the high
dose, there
was evidence or risk early on.
The Kimmel study, for all its
deficiencies, most of it was low dose
rofecoxib,
and almost all the patients used it for
less than
12 months. So, their finding on rofecoxib, if
anything, would also speak to that the
low dose
effect kicks in long before 18 months.
Finally, the Solomon and the
Ingenix study
looked at the first 30 days of use of
these
products, and both of them found elevated
odds
ratios of 4 for cardiovascular risk in the first 30
days.
Now, in both of these studies,
they didn't
separate it out by low dose and high
dose, so this
is a composite, but in both studies,
about 85
percent of the use to 90 percent of the
use was low
dose.
So, basically, what I am
concluding from
this slide is that risk of myocardial
infarction
with rofecoxib begins when rofecoxib use
begins,
and that the inability to separate out
those curves
65
is based on the fact that if you were to
count the
actual number of events in the bigger
study in the
first 6 weeks, we are probably talking
about 3 or 4
events, and if you look at the confidence
intervals, you are going to see they are
wide.
For the APPROVe study, the same
thing
holds, that you have too few events. The whole
study had 45 events, and I don't recall
how many of
those were on rofecoxib and how much of
those were
on placebo, but when you think about it,
compare
that and then look at the epidemiologic
studies,
and look at the number of cases that were
in the
epidemiologic studies, and for all their
problems,
and we can talk about those, they suggest
there is
a big discordance, and I think the
answer, the
reason is absence of statistical power in
the
clinical trials.
In the epidemiologic literature,
this has
been recognized, and people have written
papers
saying that when you are trying to
summarize the
overall risk from a survival study, and
you want to
look at specific time periods, that you
are better
66
off taking the overall risk estimate for
the entire
study than focusing on a small segment at
a time
because of this issue of low statistical
power, so
I didn't invent this.
Now, switch over to
celecoxib. There are
a number of studies that have been done
to look at
celecoxib risk. What I have tried to do here is
plot out for you the relative risk or the
odds
ratio, the author of the study, and then
the point
estimates in the 95 percent confidence
intervals.
What you will see basically is
that for
most of these studies, there is no
evidence of a
protective or an injurious effect except
for the
Kimmel study that found a substantial
protective
effect.
Remember the Kimmel study and
what I
believe is this reverse recall bias, as
well as the
low participation rate, and I personally
discount
that study. The committee can decide for
themselves that they want to do.
What about celecoxib lower dose
versus
higher dose? Well, unfortunately, the only place
67
where this is adjusted, is looked at are
in the two
unpublished studies. We have the Ingenix
study and
we have the Medi-Cal study.
What I would focus your
attention on are
the
low dose and high dose, the low dose and the
high dose. What we see is in both studies,
evidence of a dose response. Now, the 95 percent
confidence interval in the Ingenix study
includes
1, but the point estimate is pretty
elevated. That
is 1.18 or so at 400 mg.
In the Medi-Cal study, we go
from 1.01 up
to about 1.24. Here, you can see the 95 percent
confidence intervals.
What I would conclude from
this, although
they are unpublished studies, that there
is
evidence of a dose response at the higher
doses of
celecoxib do confer an increased risk of
myocardial
infarction.
I should point out that in the
Medi-Cal
study, the methodology that we used in
that study
is the exact methodology that we used in
our Kaiser
Permanente study that Dr. Platt before
was gracious
68
enough to say is one of the better done
studies.
There are no published studies
on
valdecoxib, so what do we do? Well, preliminary
data from Medi-Cal, we had 54 exposed
cases and we
found a point estimate of 0.99. Now, this was
mostly 10 and 20 mg use. I think that out of all
the patients that we had in the study,
there were 2
or 3 who had 40 mg valdecoxib use.
In Medi-Cal, they only
reimburse for the
10-mg tablet, and they do this in an effort
to try
to discourage people having larger dose
tablets and
then taking more of it.
So, this is all the
epidemiologic
information that I am aware of, that I
have had an
opportunity to review on valdecoxib.
I will now move to
naproxen. The issue of
naproxen is important for several
reasons. One,
with the VIGOR study, the medical
community was
confronted with the hypothesis that
naproxen was
the single greatest and most effective
cardio-protectant in the history of
mankind, that
it was far better than aspirin.
We heard yesterday that aspirin
reduces
cardiovascular risk about 20 to 25
percent.
Naproxen, if we were going to believe the
VIGOR
69
results, would have to reduce the risk of
cardiovascular events by about 80 to 85
percent.
So, this stimulated a lot of
research.
Here, I have summarized in the same fashion
as I
did for the rofecoxib studies, the
various studies
that have been done. Again, I have
separated them
out by the studies that I think are
better done,
the studies that have more significant
limitations,
and then the two unpublished studies.
I point out the Rahme study to
say that
the only reason the Rahme study is listed
among
this group of suboptimal studies is that
its
reference group was other NSAIDs,
primarily
ibuprofen, because ibuprofen was the
predominant
other NSAID used in Quebec during the
study.
Again, we have the various
outcomes that
were done. What I would point is that you can see
the
number of cases that we had to work with in
these various studies, and I would point
out that
70
for the Solomon study, they had about 240
MI cases
that they studied overall, but as you
will see in a
few minutes, that exposure could occur
anytime in
the past 6 months, so they don't see in
the paper
how many people were actually on naproxen
at the
time they had their event, so I can't put
down a
list of how many people were currently
exposed.
The Watson study is the only
study that
used a composite outcome. It included myocardial
infarction, stroke, subarachnoid
hemorrhage, and
subdural hematoma. Why subarachnoid hemorrhage and
subdural hematomas are in there is beyond
me. In
any event, 26 cases of that composite
outcome and a
much smaller number of actual myocardial
infarctions. So, that is why that asterisk is
there.
With the Ingenix study, the
asterisk next
to the 179 is that this included both
prevalent and
incident cases, and the best studies, the
best
results come if you base it on incident
cases only
or incident use only as opposed to
prevalent use,
because prevalent use can have survivor
bias. But
71
in any event, in the Ingenix study, they
had a
number of different analyses, and they
didn't
always use their full number of cases.
There are important limitations
to note.
I think the one to focus is to realize
(a) there is
no perfect study, we have talked about
that before,
and, two, that among all the limitations
listed
here, I think the most important one to
note was in
the Watson study, was this composite
outcome which
really just makes it very difficult from
an
epidemiologic perspective to study things.
Myocardial infarction is very
well
validated in claims data, and Dr. Platt
has already
gone over that with you. Stroke is notoriously
difficult to work with in claims data,
and subdural
hematomas most commonly occur because as
people get
older, their brains shrink. They bump their heads
and then they get a little bleeding on
the surface
of the brain. What that has to do with myocardial
infarction risk, which is what we are really
concerned about today, is beyond me.
I have got two slides on the
results.
72
This slide shows the studies that found
no
protective effect. There is four studies that
found a protective effect, and I am
saving them for
a separate slide, because I want to look
at those
individually.
What you can see from the
majority of
these studies, and I would point out that
the
studies that were the best done studies
in the top
tier, they are on this slide, that all of
them sort
of suggest that there is no
cardio-protective
effect of naproxen. Several of the studies point
to the possibility of a small increased
risk with
naproxen.
But we have four studies of
positive
results, and we will probably all
remember the
Archives of Internal Medicine publishing
three of
the articles in the same issue with an
accompanying
editorial that stated the issue is
solved, naproxen
is cardio-protective.
I want to look at those studies
and just
describe to you my view of them. The top three
studies were the ones that were--well,
no, not the
73
Kimmel study--Rahme, Solomon, and Watson
were the
Archive studies.
In the Rahme study done in
Quebec, they
compared current naproxen use versus
other NSAIDs.
That other NSAID was, by and large,
ibuprofen, and
they found a protective effect. Well, if
ibuprofen
increases the risk of myocardial
infarction, let's
just say that it does, and naproxen doesn't,
naproxen could look like it's protective
compared
to ibuprofen, but not be protective
really.
The data presented in that
paper, if we
re-analyzed it versus non-use, we get an
odds ratio
of 1.28, statistically significant. Now, this is
not adjusted. It is not possible from the data
there for me to adjust this result, but
based on
what is in the paper, when you compared
the
unadjusted to the adjusted point
estimates, they
don't change very much, and what that
suggests to
me is that this effect, this 0.128 is
probably not
far off the mark.
That would then make it
comparable to the
analyses I showed on the previous slide,
that all
74
of these slides use non-use or remote
use, so then
it would add a fourth study to an
elevated point
estimate for naproxen.
Now, the Kimmel study, we have
already
talked about low participation rate and
this
reverse recall bias, and a small number
of NSAID
cases.
In fact, they don't even tell us in the
paper how many cases they had.
We move on to the Solomon study. This was
the result that was reported in the paper
and was
picked up by the press, a 16 percent
reduction in
heart attack risk with naproxen. The problem, in
my view, was that their definition of
exposure in
the study was any use of naproxen in the
past 6
months, which means that if I took
naproxen 6
months ago and stopped it, I could be
included in
this study as being exposed to naproxen.
So, the question is then, you
know, how do
we interpret the study. Well, Solomon was good
enough to present data by current use and
in recent
use, and recent use included people who
stopped
their naproxen. Their naproxen prescriptions day
75
supply ran out between 1 day and 60 days
before the
MI or the index date for their controls,
and remote
users, their NSAID use, their naproxen
use ended
from 61 days to 180 days prior to the
event.
So, let's look at what those
results are
then, and what we see is they are
identical. So,
unless the committee is prepared to
believe that
naproxen confers lifetime immunity to
cardiovascular disease, I think we have
to conclude
from these data that what we really have
here is
selection bias, and it is not the fault
of the
investigator. Dr. Platt talked about
before that
there are some things you can't adjust
for. You
can't adjust for bias. What you
can try to do is
identify bias, and if you identify it,
then at
least you know what you are dealing with.
Here, I think we have what is
classic
selection bias. It is not naproxen that protects
you again myocardial infarction, it is
some other
factor that in this health plan, that
they used to
study this drug, the patients who were
being
treated with naproxen happened to have
lower
76
cardiovascular risk.
I can't explain why that
happened. Dr.
Solomon probably can't explain why it
happened, but
it's not due to naproxen.
Finally, the Watson study. This study was
sponsored by Merck, and it was authored
by Merck
investigators. The result that was published as
being the basis for the conclusion was
this top
result, a 39 percent reduction in
cardiovascular
risk.
First, I just want to remind
everybody,
composite outcome here, subarachnoid
hemorrhage,
subdural hematoma, stroke, as well as
heart attack,
26 events total, much smaller number of
heart
attacks.
For this event, you can see the
checkmarks. These are the various variables that
they adjusted for in the study. The way they
handed cardiovascular risk, if you read
the paper,
I would have to say that it doesn't measure
up to
the standards that were set by Dr. Wayne
Ray.
We modeled our study in Kaiser
and in
77
Medi-Cal, and Dr. Wayne Ray, I think that
he has
set the standard for how one needs to go
about
adjusting for cardiovascular risk. It is
not enough
to rely on diagnoses. You have to use the
medications, because medications are much
more
accurate predictors of disease than
diagnoses in
these administrative claims data.
In any event, they didn't
adjust for
cardiovascular risk, and they didn't
adjust for
smoking although they had that data. Then, they
present later on another analysis that
now includes
cardiovascular risk and it is no longer,
in quotes,
"statistically significant,"
and then they include
smoking, and again it is not
statistically
significant.
My conclusion on the Watson
study was that
(a) they have got a composite outcome
that, in my
view, isn't very informative towards the
question
of myocardial infarction; (2) that it is
very small
numbers; (3) that a variety of approaches
were used
in the analysis that inadequately account
for the
risk factors that could confound the
result, so I
78
have discounted that, as well.
So, a conclusion when I look at
these, in
quotes, "4 positive studies," I
conclude that none
of them provide credible evidence of a
protective
effect.
In light of yesterday's
discussion in the
afternoon about other NSAIDs and what
might explain
the differences, let's say, celecoxib and
rofecoxib
studies, the rofecoxib studies used
naproxen as a
background, a comparator, the celecoxib
studies
using ibuprofen or diclofenac.
Dr. FitzGerald is talking and saying,
well, you know, all of these drugs could
increase
the risk because what is happening, you
know,
biochemically, with the balance of
prostacyclin,
could be influenced by these different
drugs in
ways that aren't immediately obvious or
detectable
in a clinical trial.
I thought I would just share
some of that
information on other NSAIDs with the
committee,
recognizing a couple things that no
single study is
definitive and what you want to look for
I think is
79
consistency across studies, but as far as
randomized trials go, I would like just
to mention
that there are generally too small, too
few events,
and you are not going to get the answers
that you
need from them unless you make these
clinical
trials substantially larger than anything
people
have contemplated up to now.
So, from our California
Medicaid study, it
is all preliminary and it has not been
published,
for ibuprofen we found a small but
statistically
significant increased risk. For
indomethacin we
found a risk of 1.7. I would like to say on
indomethacin that we found an increased
risk with
indomethacin in our Kaiser Permanente
study. It
was 1.3 and it was highly statistically
significant.
In at least two other studies
that I
reviewed in preparation for this advisory
meeting,
indomethacin is noted to have an
increased risk of
myocardial infarction.
It is not commented on in the
text because
that wasn't a primary analysis, but what
I am
80
talking to you about now is consistency,
and I
would submit to the committee that
indomethacin is
a lot of smoke, there is a lot of smoke
for
indomethacin.
In our study, in our Kaiser
study, for
example, we did not think in advance to
look at
indomethacin separately. I mean we knew
we were
going to look at it, but it wasn't a
primary
hypothesis. We didn't adjust for gout. I mean
everyone knows that indomethacin gets
used in gout.
Gout increases the risk of cardiovascular
disease.
Well, in the Medi-Cal study, we
adjusted
for gout. Yes, gout increases the risk of
myocardial infarction. It didn't change the odds
ratio here.
I think this next finding,
Meloxicam, is
important. Meloxicam is now the number one selling
branded NSAID in the country. With the removal
from the market of rofecoxib, the medical
community, shying away from the coxibs,
are moving
to other drugs that they perceive would
have the
advantages of COX-2 selectivity without
the bad rep
81
that coxibs appear to be acquiring.
So, you now have a shift in the
marketplace to Meloxicam. There have been articles
in the Wall Street Journal and the New
York Times
on this.
The company recently raised the price on
the tablets.
In any event, we are presenting
these data
just to say that we found an increased
risk. It is
one study, but I think it is the only
study. We
looked at this in Kaiser. Meloxicam is almost not
used in Kaiser, so we couldn't study it.
In our California Medicaid
study, we only
looked at drugs that had more than 50
currently
exposed cases. Nabumetone came out in this study
as not showing a whiff of a problem. Sulindac,
there was an increased risk.
Regarding ibuprofen, in our
Kaiser study,
we found an increased of 1.06, which
sounds really
trivial.
It wasn't statistically significant, but
the confidence intervals were pretty narrow. It
was 0.96 to 1.17.
My concern is, as Dr. Platt
talked about,
82
you know, above 2 you feel really
comfortable,
above 1.5, you can believe it, below that
you begin
to get really edgy. The problem is most of the
risks that we are probably facing, if it
turns out
that the non-coxib NSAIDs increase the
risk of
cardiovascular disease, that is where the
risk
level is going to be, and that is what we
are going
to have to contend with, because it has
tremendous
effects on the population.
Finally, dose response. This slide shows
for diclofenac. This is from California Medicaid.
What we wanted to do was show evidence of
dose
response, consistency in the data. Remember we
pointed out diclofenac before. Diclofenac in this
study overall did not have an increased
risk, but
at the high doses there is a suggestion
of a dose
response.
I will skip that. This slide was to say
that depending on your reference point,
you can get
different results, if I use an active
comparator
versus remote, and this is showing the
three NSAIDs
from California Medicaid compared to
non-coxib
83
NSAIDs, and you can see the rofecoxib is
different
than them, and the other two aren't
necessarily
that different.
My conclusions, and I am sorry
to have
gone so long. Celecoxib, we believe that based on
the evidence we have at hand, that there
is no
apparent effect of risk at doses of 200
mg or less.
Above 200 mg, we think that there is
evidence of
increased risk.
For rofecoxib, we believe that
there is
evidence of increased risk at both the
lower doses
and the higher doses, and that risk begin
early in
therapy and is apparent during the first
30 days of
use.
With valdecoxib, there is a
paucity of
information, but the information we have
at this
time suggests that the risk is not
increased at
doses of 20 mg or less.
As a class, non-coxib NSAIDs
may increase
the risk with differences between each of
the
NSAIDs.
I don't think we are going to be able to
talk so much about class effects. In the
end, it is
84
going to have to be looking at individual
drugs.
The COX-2 hypothesis may be
true, but if
it is, we are still going to have to look
at these
other drugs in terms of their individual
properties
and what they do.
Finally, naproxen is not
cardio-protective.
Thank you.
(Applause.)
DR. WOOD: Thanks very much. David, it
will come as no surprise to you that
every time
practically I pick up a newspaper, I read
about
what you are not going to tell us.
So, my question to you is what
have you
not told us that you think we should
know, because
I would like to make sure. Lots of other people
have shown up here without slides that
they forgot,
so I just want to be sure that if there
is anything
else we need to hear, we hear it.
DR. GRAHAM: Well, as far as the science
goes, I think I presented the evidence that I am
happy to be able to share with the
committee that I
85
thought it was important for the
committee to have
an opportunity to hear.
The source of controversy
surrounding my
presentation related to the unpublished
studies
that I was going to be permitted to
present or
asked, actually asked to present the
Ingenix
results, the unpublished study from
Merck, but that
I was being told not to present the
unpublished
data from the California Medicaid study,
and
personally, I had great difficult
standing here
before this committee as an investigator
and as a
scientist, as a physician, and telling
you the
information that I have, that I am
allowed to talk
about, and remaining silent on things
that I know
about that I am not allowed to talk to
you about.
Fortunately, Dr. Crawford
exercised great
leadership in making it possible for me
to present
that data, recognizing it's preliminary,
but the
methods that we used are identical to our
Kaiser
study for the California Medicaid, and
for me, I
think the big reservation is, is that
it's an
untested database, but I think that
everything that
86
could be done to develop the database and
to do
quality assurance and to work out the
kinks has
been done.
If you look at the findings in
the
California Medicaid study and you compare
them to
the clinical trials data, and the
anomalies and the
questions that you were discussing
yesterday about
the clinical trials' data, you look back
at the
California Medicaid data, and you are
going to see
I think great consistency between the
findings that
might help explain and interpret some of
the things
that seemed questionable or uncertain
yesterday.
So, in any event, I have been
able to
present what I thought was important to
present,
and I am happy to have had that
opportunity.
DR. WOOD: So, the answer is we have seen
it all, is that right?
DR. GRAHAM: You have seen it all.
DR. WOOD: Okay, good.
Let me ask you a
question. If you go back to your slide
that showed
the excess population risk, put that in
proportion
for us in terms of, say, the other drugs
that have
87
been withdrawn from the market. I mean what sort
of numbers would we be expected to see?
DR. GRAHAM: That is a great question.
The typical drug that has come off the
market in
the United States, like the leading cause
of drug
withdrawals in the United States in the
last 20
years has probably been acute liver
failure.
Rezulin came off the market because of
it,
troglitazone, bromfenac, a number of
other drugs.
Acute liver failure in the
general
population has a background rate of about
1 per
million per year. We are talking about that is the
rate of being struck by lightning, 1 per
million
per year, and these drugs were pulled off
the
market because it increased the risk of
that. It
might increase the risk 5-fold, it might
increase
the risk 10-fold, it might increase the
risk
100-fold.
The fact is the background rate was 1 in
a million and what that means is that the
actual
number of people affected is sort of
measured in
the tens and the hundreds for the liver
failure
that could be life-threatening.
In this situation, and this is
why the
lower relative risk becomes so critical,
we are
talking about a serious event that has a
very high
88
background rate. Heart attack is not a rare event,
and as I pointed out before, there is a 1
in 50
chance that the average American male age
65 to 74
is going to have a heart attack this
year, 1 in 50.
That is an extraordinarily high
risk. You
increase that risk 5-fold with a high
dose. That
is what happened with VIGOR. If I have got
millions of people taking the high doses,
and that
is what had in the United States, and I
have
increased the risk 5-fold, you are going
to get
numbers that balloon out like this.
So, there is no comparison in
terms of
what the population impact is of the
typical drug
that has come off the market in the United
States
and what we are dealing with here, and
that is
because of the high background rate of
the
underlying event that we are talking
about.
DR. WOOD: So, this would produce many
more cases from what I understand.
DR. GRAHAM: Many more.
Committee Questions to
Speakers
DR. WOOD: From the committee, we have
questions. Let's start with Dr. Shafer.
DR. SHAFER: Dr. Graham, tomorrow we are
going to be asked, as a committee, to
consider the
89
question about a class effect for the
selective
COX-2 antagonists and for the
non-selective NSAIDs.
One of the things that I am
finding, that
I am having trouble putting together
here, is we
have a lot of conflicting data, and for
the COX-2
antagonists we have a lot of data from randomized
controlled trials.
Certainly for the NSAIDs, we
are going to
have to go with a lot of these
observational
studies because we don't have a lot of
data on the
topic at hand from randomized controlled
trials.
As I look at this, if we come
up with some
sort of common warning as a class, and it
applies
to everything, we have, in fact,
communicated no
relevant information. On the other hand, if we are
going to come up with individual
drug-specific
90
recommendations, we are going to have to
have very
different evidentiary standards in some
ways,
because for some of these, we have very
little
information, as you pointed out, and yet
your data,
particularly the unpublished data from
the Medi-Cal
trial, and I appreciate that there is all
the
issues of not being previewed and stuff,
but we are
all familiar with that process and know
how it
works.
What can you tell us to guide
us? Should
we try to go drug by drug specific? How do we set
our evidentiary standards when we talk
about class
effects where in some cases, we are just
not going
to have a lot of data here?
DR. GRAHAM: Right.
What you are going to
be getting now, of course, is my opinion,
not FDA's
opinion. Probably if you were to talk to
Bob Temple
or John Jenkins, or anybody else,
everybody is
going to have a slightly different
answer.
What we talking about now I
think to some
extent is philosophy, so what that
preamble, first,
I believe based on the evidence that there
is a
91
COX-2 effect and that that COX-2 effect
is dose
dependent, and that we see evidence of
that with
rofecoxib, with celecoxib, and with
valdecoxib.
The difference between rofecoxib and
the
other two coxibs on the market is that a
safe dose
for rofecoxib wasn't identified, the dose
wasn't
low enough. That raises a question in my mind
about what is an appropriate therapeutic
index for
a drug.
I am giving you my opinion now,
but when I
listened to Dr. Cryer's presentation
yesterday, the
bottom line conclusion I came to at the
end of that
was there really doesn't appear to be a
need for
COX-2 selective NSAIDs based on what I
heard
yesterday. There is probably other information out
there why I am wrong, but that was the
conclusion I
came from.
So, in any event, that is
answer one. I
believe there is an effect and it's dose
related,
and with celecoxib and valdecoxib, I
think we have
evidence.
You said before we have a good
evidentiary base based on clinical trials
for the
92
COX-2s.
I would challenge that in the sense of the
survival curves and the things that I
talked about
there, that we have a very weak
evidentiary base
for things like protective, you know, is
there a
grace period for use, and also on the
dose issue,
we really don't have a great evidentiary
base. But
that being said, you understand me.
Now, for the non-coxib NSAIDs,
my own view
is that as an epidemiologist first, I try
to report
the phenomenon I observe and leave it to
brighter
minds to figure out why what I observed
happens.
You are asking me sort of what
do I think
is happening underneath it all. I am attracted to
the COX-2 hypothesis personally. Dr. Gurkiepal
Singh, my colleague and co-author in
Medi-Cal, he
has a different view on that, but I think
that we
can these in vitro tests that say, oh,
this is the
COX-2 selectivity of this NSAID, you
know, in a
test tube.
What happens in the human body
could end
up being surprisingly different. We saw yesterday
that the dynamic response of these
differences,
93
that the platelet effect is very quick,
the
thromboxane effect is a very quick
effect, the
prostacyclin effect seems to be a more
gradual
effect, that this creates very complex
interactions
that ibuprofen, that any of these drugs
could, in
the end, end up with a deficit, a
prostacyclin
deficit that results.
I think Dr. FitzGerald showed
that slide
yesterday with the normal distribution of
the time
area under the curve and then this little
sliver
where they are not protected, and that
may be the
reason why, for these different drugs,
that we end
up with these different relative risks
and these
different odds ratios.
In the end, for the non-selective NSAIDs,
my own advice would be let's look to see
are there
somewhere in studies--it is going to be
observational studies--in observational
studies
that we believe have been reasonably well
done.
By "well done," here,
they have to be
large.
The literature is full of really small
studies.
I mean I could have presented Meloxicam
94
studies, 5 patients, no risk. Well, da, you know,
you have got a confidence interval that
goes from
zero to infinity. They need to be large. Look in
a systematic way to identify what the
body of
evidence is.
Can we identify bad actors? I believe
indomethacin, for example, is clearly a
bad actor,
and if people looking at the data
concluded that,
take appropriate action, weed the garden
of the bad
actors.
Try to identify drugs that
based on the
evidence we have, appear to be less risk
in the
totality of their evidence, looking for
consistency
study to study to study, and then, in a
rational
way, suggest these are the drugs we think
that the
public should use, and these other drugs,
well,
then you have to decide do you want them
on the
market or not.
I am not really going to
comment on that,
but I think that is the approach I would
take. I
would be trying to sort of identify right
off the
bat the bad actors and let's get rid of
them.
Things that look like they may
actually be
safe, and when I say "safe"
now, I mean that they
don't appear to have cardiovascular risk,
identify
95
them and shift the market towards that,
and then
deal with the others.
DR. WOOD: Dr. Friedman.
DR. FRIEDMAN: Thank you.
Several
comments.
First, as both Dr. Graham and Dr. Platt
have mentioned, observational studies are
essential, but they have a number of
limitations,
and because of those limitations, it is
easy after
the fact to critique away those whose
results you
don't much care for as we have seen.
But a couple of other
points. One, can
these particular drugs, their primary
use, we are
dealing with chronic conditions,
conditions that
last years, sometimes many years, and so
the drugs
are intended for use over those many
years
potentially.
Yet, most of the clinical
trials we heard
reported yesterday are 12, 18 weeks, a
few of them
go longer. You mentioned that one of the reasons
96
we didn't see the problems early on may
be numbers,
and I agree that is potentially it, but
the fact is
we didn't see problems arise in the
studies until
14, 18 months.
We often see analyses by
patient years of
exposure.
In this particular setting, I don't know
whether patient years are always equal to
patient
years, and therefore, I guess I would say
why
aren't we doing more bigger, longer
randomized
clinical trials for these chronic
conditions?
DR. GRAHAM: I am not speaking for the
agency now.
DR. WOOD: We got that.
Don't say it each
time.
DR. GRAHAM: Okay.
I think they are
incredibly expensive and companies don't
want to do
them.
There is not an incentive for them to do
them, and you would have to talk to the
people from
the
new drug side of the house, but the fact is
that they are not requiring them.
So, that is a very legitimate
question.
You know, working as an epidemiologist,
we try to
97
make do with what is, and so we use the
observational data. You are going to get better
quality data if you are able to do this,
but just
to give you a sense of the size of the
studies that
I think you would need to do, I mean you
talked
about before that you have the APPROVe
study and we
see no effect until 18 months, but there
was study
090 that was talked about briefly by Dr.
Villalba
yesterday. It was a 6-week study at 12.5 mg, and
it showed a difference, the suggestion of
a
cardiovascular risk within the 6-week
study at the
lowest dose. Now, it's a small study, as well.
But I am just saying that to
say that I
think the epidemiologic data, in my mind
at least,
answers the question about when the
effect begins.
The question is if you want to have--this
is the
philosophy--how much certainty do you
need to make
a decision.
Right now, when it comes to
efficacy, the
effect, does the drug work, you are
looking at the
lower bound of the confidence interval,
and you
want to see is that different than 1,
because if it
98
is, then, I will conclude with 95 percent
certainty
or greater that the drug actually has an
effect.
When it comes to safety, you
are doing the
same thing. You are looking at that lower bound.
You
want this 95 percent certainty that the drug is
harmful.
You are presuming that the drug is safe
rather than let's presume we want to do
no harm to
patients.
Let's start off at the
beginning assuming
that the drug isn't safe, and we want to
have a
certain level of confidence about how bad
this drug
could be, and that is still tolerable to
us. We
want to cap the risk. It will be a completely
different way of looking at studies for a
safety
perspective, one that actually gives a
priority to
safety and it maximally protective of
patient
safety, just as that high standard for
efficacy is
maximally protective of patient safety,
because by
keeping drugs off the market that don't
work, I am
protecting patients from unsafe drugs,
and if I
have pneumonia and I am given a drug that
doesn't
work, well, I get a harm from that.
But that's philosophy, and I
think it's an
outcropping, it's a development, a
natural
extension of the development of clinical
trials in
99
the United States where the focus has
always been
on efficacy.
DR. WOOD: Let's try and keep both the
questions and the answers reasonably
short,
otherwise, we will be here until after
midnight.
DR. GRAHAM: I apologize.
DR. WOOD: That's okay.
Let's go on to
Dr. Elashoff.
DR. ELASHOFF: First, I have one comment
and then one question. In terms of confounding,
just because you put a lot of variables
in some
model doesn't necessarily mean that you
have
adequately removed the confounding
effects even of
those variables.
The second has to do with Dr.
Graham's
slide 13, the excess population
risk. I note that
the Ingenix data has been left out of the
bottom
category.
DR. GRAHAM: That's right, because for the
100
high dose.
DR. ELASHOFF: Yes, but the negative sign
needs to be on the slide, otherwise, it's
a biased
presentation.
DR. GRAHAM: Well enough.
I take that
correction. Okay, fair enough.
DR. WOOD: Dr. Bathon.
DR. BATHON: Yes.
As we weigh the
risk-benefit ratio of these drugs, one
consideration is that there are subgroups
of
patients in which the benefit might
outweigh the
risk possibly.
With that in mind, it would be
helpful for
us who are not cardiologists or
epidemiologists to
be able to put the relative risks that we
have been
seeing over the past day or two in
context with all
the cardiovascular risk factors that
exist.
So, for example, if you were
take the
presumed relative risk of rofecoxib of
1.5 to 2.0,
at least at the higher dose, and put it
into some
context for us of the 20 to 40
cardiovascular risk
factors that exist in a sort of rank
order, where
101
would you put the COX-2 drugs?
DR. GRAHAM: For the high dose it would
be probably more significant than smoking
or
diabetes or hypertension, maybe more
important than
the combination of several of those
factors in a
patient.
For the lower dose, it is probably more
than hypertension, a little less than
diabetes, and
a little less than smoking.
I know, David, you know the
cardiovascular
risk factors much better than I do, and
so does Dr.
Hennekens, but that would be my ballpark
on that.
DR. WOOD: Dr. Abramson.
DR. ABRAMSON: Yes. I
want to go back to
the question Dr. Shafer asked about if
these
classes of drugs or this group of drugs
could be if
there was a hierarchy of risk, and you
first
answered that you thought the coxibs were
more
risky, but I would challenge you a bit
simply on
your own presentation.
I would like you to discuss your data,
because you then went on to talk about
how
indomethacin has a risk, Meloxicam has a
risk.
102
Based on your data, the message that came
through
is that there was a dose response risk
for
cardiovascular outcomes, that we saw it
within the
coxibs, but we also saw it where the data
were
available in the non-selective NSAIDs.
There are data that we have
seen that
ibuprofen might increase risk. We didn't talk
about the McDonald and Way paper that in
cardiovascular discharge patients, people
given
ibuprofen had a higher mortality 2-fold.
So, as the smoke clears, I am
not sure
that the simple answer that the coxibs
were
different was actually supported by your
data, nor
your ultimate explanation. Can you defend that?
DR. GRAHAM: I think you are accurate.
What I was saying was I was referring, I
think, to
the underlying COX-2 hypothesis and that
it is
clearer, I believe, and, well, maybe it's
an
overgeneralization, because we have the n
that we
are viewing is so small, that looking at
rofecoxib
as sort of the example where we can see
very
clearly the dose response at all the
levels and its
103
progression, and understanding its
mechanism of
action, and then seeing similar things with
celecoxib and valdecoxib.
I think what you are saying is
fair.
Maybe a better thing to say is, in the
end, that
you do need to look at it drug by drug.
What I was saying, though, in
that answer
that I gave to Dr. Shafer, I was really
talking
more about sort of the COX-2 mechanism
and the
coxibs as being, in quotes, "COX-2
selective," but
I think your observation is correct.
DR. ABRAMSON: Add to that, that although
there is a hazard that we don't
accomplish a lot by
simply saying the class of NSAIDs may
have risk, I
think we have under-appreciated that over
the last
10 years.
It is not that different from the
mid-nineties recognizing that there was a
class GI
effect of these drugs, and that compared
to
placebo, whether it's hypertension or
long-term
potential adverse outcomes, this is
something that
doctors have to be aware of, even the
simple thing
104
of checking blood pressures when you put
people on
any nonsteroidal drug.
So, I don't know that it is
necessarily a
bad outcome to call attention to this
class effect
until we get better information on each
of these
individual drugs.
DR. WOOD: Dr. Day.
DR. DAY: I have a comment about recall
bias and reverse recall bias. There is a huge
research literature on how memory works
both in the
laboratory and in the every-day world,
and there
are two phenomena that have been very
heavily
studied that I think might be relevant
here.
One is called flashbulb memory,
and the
idea is when an emotional spectacular
event
happens, such as when you first learn
that JFK had
been shot, or the Challenger blew up, or
the World
Trade Center had been hit, it is as if
the old-time
flashbulb from an old-time flash camera
went off
and captured all the details, and you
remember all
of those details forever afterwards
associated with
the event that you might otherwise have
just not
105
even noticed or forgotten.
So, there is a lot of research
on
flashbulb memory that shows many of those
details
are indeed correct, but some are
notoriously false.
For example, there are accounts of people
who
remember a certain even with great
emotional
aspects to it, and they remember
listening the
world series when so-and-so is pitching
and it was
the
bottom of the 9th, da-da-da, all these details,
and when you go back and check the
evidence of what
was going on, on that day and time, that
particular
game was not on.
So, that phenomenon number one,
flashbulb
memory, and the second is eyewitness
testimony.
How you ask a person a question will
affect what
answers you get. So, if you have in the courtroom,
someone who has witnessed a car accident,
if the
lawyer asks this witness, "Did you
see the broken
glass," then, the witness is more
likely to say yes
than if you ask, "Did you see any
broken glass,"
because the broken glass presumes that
there was
some, and so forth.
So, I take your points
seriously about
potential recall bias and reverse recall
bias, but
we would have to look at both, whether
there is an
106
emotional component or not. Those who have had an
MI, for example, would have that most
likely, but
also how the questions are asked in these
surveys,
and it is not trivial how you ask people
questions
about were you taking any medications or
were you
taking medication X, and for how long,
and what was
the dosage, and so on.
So, I don't think that these
details are
always published with the studies, and I
would like
to encourage people who ask people about
their
experiences with drugs, take a look at
the memory
literature for some of these points.
DR. WOOD: Dr. Gibofsky.
DR. GIBOFSKY: Dr. Graham, I am wondering
if you separated out your populations
based on the
indication for which they were taking the
drug. I
ask that because we heard yesterday, and
it's well
known, that rheumatoid arthritis is
itself a risk
factor for cardiovascular disease, and
higher doses
107
of coxibs, in particular celecoxib, are
usually
given to patients with rheumatoid
arthritis as
opposed to osteoarthritis.
So, I am wondering if you look
at that in
your breakdown.
DR. GRAHAM: Several of the studies that I
reviewed have looked at the indication,
but in
automated claims data, it is very
difficult to be
sort of be sure does the patient have
rheumatoid
arthritis, and there are different
algorithms one
could use, but in general, what has been
found in
the studies where they have looked at
that, that
the prevalence of rheumatoid arthritis in
the study
populations has been low, very low, and
that its
impact on the results when they adjusted
for it
didn't materially affect things.
Now, in the California Medicaid
study, one
difference in that study was that our
base
population was limited to patients who
had
diagnoses of osteoarthritis or rheumatoid
arthritis. Now, these are diagnoses, and so does
that mean that they really had
osteoarthritis or
108
rheumatoid arthritis, I don't know, but
when we did
try to eliminate in that study at least
were the
people who might be using an NSAIDs for a
muscle
injury, a short-term complaint as opposed
to a
chronic illness.
In none of those does the
presence of
rheumatoid arthritis seem to affect
things, but
again I think the prevalence is pretty
low in all
of these studies.
DR. GIBOFSKY: One quick question for Dr.
Platt, if I might. I need to understand the
concept of survivor bias somewhat in that
I think
there is a difference between a patient
who is
drug-naive, then put on a drug, and then
an event
happens versus a patient who may have
seen a drug,
perhaps seen another drug after that, 3 or
4 agents
of the class, and is then switched to
another agent
and something happens.
I think we have talked about
remote versus
current, but there is also this issue of
sequential
effect, and I am wondering how you deal
with that
as a survivor, particularly because of
the paper we
109
saw a few weeks ago in the Archives
suggesting that
discontinuation of an NSAID may itself be a
risk
factor for a thrombotic event.
DR. PLATT: Your point is exactly right.
I think that the concern about survivor
bias is
that if we think that some people are
particularly
susceptible, which is almost certainly
the case,
then, if we start the clock after a
person has
already been exposed to a drug or to one
that has
the same effect, then, it is very much
less likely
that those individuals will have a
problem.
That may be the explanation,
for instance,
for the reason that the literature was so
badly
wrong about postmenopausal estrogens and
heart
disease, that most of the epi studies
started with
prevalent users.
I think the majority of the
studies that
we were reviewing here, these were
individuals who
are known to have had at least a year of
prior
experience without exposure to the
nonsteroidals.
Your study in Kaiser I know was an
exception cohort at least with regard to
a year of
110
prior history, but I am not aware that
any studies
have a longer drug-free prior interval
than that.
DR. WOOD: Dr. O'Neil, do you want to
comment particularly on this?
DR. O'NEIL: Yes, this is an important
point and a lot of things have been
covered in
Richard's and David's presentation, but
one thing I
think that is relevant that Richard did
not cover,
that is, the value of a randomized trial,
is the
ascertainment and follow-up, and knowing
the status
of individuals in the sense of who goes off
therapy
and how long they stay on therapy.
That is very critical relative
to the time
dependency of the risk. It was mentioned, for
example, the use in the observational
sense of
recent and remote and current use. Those are all
terms that are nice, but they don't get
at the
issue that we are trying to get at with
regard to
the clinical trials, and that is
essentially when
does time zero start for you.
So, I think the appropriate
question to
ask is what is the duration of exposure
since your
111
initial exposure to the drug, because I
think that
is very relevant to the interpretation of
the three
clinical trials that we have, two of
which are in
placebo-control populations.
There is a rofecoxib-naproxen
control
trial for one years, there is a
placebo-control
trial in polyp prevention for three
years, and
there is a placebo-control trial in
Alzheimer's
disease for four years, and the time
dependency
from time zero matters as you have seen
in the
plots.
It is relevant to the excess
risk
calculation. So, I would ask the committee, as
well as I would ask David, of the
observational
studies that you have reported, how many
of them
are cohort studies, and how many of them
are able
to identify new initial use, and then
track
continued use for that individual, so
that one
could look at the relationship between
the hazard
rates and the hazard ratios that we are
identifying
in the randomized trials and match that
to the odds
ratios that are being reported in the
observational
112
studies.
DR. GRAHAM: On one of my initial slides,
you can see what the cohort studies were,
and in
some of the nested case control studies,
you are
also able to get the time on drug. Actually, in
Wayne Ray's cohort study, most of these
cohort
studies include prevalent and incident
users, so
they will do what is called a "new
user"
subanalysis, which is to try to get to
this issue
of when does time zero begin.
We addressed that problem in
our study
here by the inception cohort design in
our base
population, so that we can identify what
time zero
was for the cases.
Now, none of those studies
presented data
in the form of a survival analysis, which
I think
in the end, that is what Dr. O'Neil would
like to
see.
DR. O'NEIL: No, my question is not so
much in survival. I don't believe, and again that
is why I am asking you, I don't think any
of those
studies were designed or able to capture
the
113
question I am asking.
In fact, if I am not mistaken,
in the
Wayne Ray study, he defined new use, but
he did not
define any time from new use, which is
essentially
critical to when those risks start.
DR. GRAHAM: That study isn't cited as one
of the studies where we are able to
derive that
information. This slide was a slide that I
presented to show that from the
epidemiologic
literature, those studies where the investigators
had identified when time zero began for
rofecoxib
use, and they didn't present the data as
a survival
analysis, but they identified when time
zero began
and then, in various ways, showed you
either what
the distribution of the cases were, so
that you can
see that it was impossible for the risk
to have
been delayed for 18 months, because
nobody in the
study used the drug for 18 months, or
they parsed
time out and looked at the first 30 days
of use
from time zero, and found the risks that
they found
down here.
But you are right, those
studies aren't
114
designed that way, and we haven't had
time in our
Medicaid study to do these analyses yet,
but we
have the data to now do the cohort study
and time
to event, so we will have an opportunity
actually
within the data to actually compare and
look to see
exactly the question you are driving at.
But I would say that from the
published
data, in each of these studies, time zero
for
rofecoxib was identified and in some way
or
another, information that I think could
be useful
to
the committee in establishing when does risk
begin was contained in those studies.
DR. O'NEIL: Well, the other point here,
which is the value of clinical trials,
and it was
the question that was discussed yesterday
with
regard to the intent-to-treat analysis,
and that is
to say to analyze all outcomes once
randomized to
the trial regardless of whether you want
to track
the individual to 14 days post-exposure.
You can't really maybe get
access to this
information in the observational
studies. That is
a conjecture, but it's one or the other
biases, and
115
it was interesting to the comment,
whether one
would believe this or not, that
discontinuation,
discontinuation from an NSAID alone
raises risk.
If that were to be the case,
that is a
different analysis altogether.
DR. GRAHAM: In that actual paper, it
could be that people were discontinuing
the NSAIDs
because they were having chest pain and
it was
being interpreted as dyspepsia or
something, and
then they go to have their infarct.
I mean you are right about
that, but this
is the nature of how epidemiology is
done, and I
can't change it. I didn't make the rules, I am
only following them. Nobody is arguing that
clinical trials, if they could be large
enough,
that they would give all of us answers
that we
would have greater comfort trusting what
they are
saying.
What I am proposing is that we
don't have
that kind of data in the clinical
trials. As large
as the clinical trials are, for the
questions that
this committee is facing, you don't have
the data
116
you need, and what I presented is the
epidemiologic
data, and it is imperfect and it has its
warts, and
that is why I would emphasize looking at
consistency and trying to sort of derive
from that
a general sense.
I mean does it make
pharmacologic sense
that you would have an 18-month
delay? I mean I
guess I suppose it depends on what you
think the
mechanism of action is for the underlying
disease,
but even in the clinical trials, study
090 was 6
weeks long, 12.5 mg, and it had a
cardiovascular
effect.
DR. WOOD: I am happy to facilitate a
discussion among the FDA, but I think we
would
rather hear from the committee right
now. Dr.
Farrar, you are next.
DR. FARRAR: I think that the
recommendations of the committee tomorrow
are going
to depend on the assessment of the
overall risk and
the overall benefit of this class of
drugs.
As a researcher and after all
the data
that has been presented, I am more than
happy to
117
accept the fact that there are serious
risks even
of death from taking NSAIDs. In fact, though,
there are serious risks in taking any
medication at
all.
For some of the NSAIDs, it is
cardiovascular risks, for some of them it
is
clearly GI bleeding. As a doctor, though, who
takes care of patients, I know that
treating pain
or not treating pain and not treating the
disability of arthritis also has very
serious risks
even of death.
Given the extensive work that
you have
done, on the risk of both the
cardiovascular and
the GI bleed, I wonder what level of risk
is
acceptable you, and remembering that the
only other
drugs that are really available is
analgesics or
narcotics, and the only other drugs that
are really
available in terms of limiting
inflammation are
biologics or immunosuppressants, I wonder
what drug
is safe enough that you would recommend
that I
actually would be able to use it in
patients to
prevent some of their suffering.
DR. GRAHAM: Well, I am not going to give
a product endorsement. A couple of things, though.
DR. WOOD: Try and make it brief.
118
DR. GRAHAM: One, the benefits of the
treatment for the traditional NSAIDs
compared to
the COX-2 selective NSAIDs with GI bleed,
we have
clinical trial evidence that suggest that
there may
be a difference, but here, to me, is an
anomaly.
Rofecoxib got the indication
for being
GI-protective, celecoxib didn't based on the
clinical trials data you guys looked at
yesterday.
There are two published studies
in the
literature looking at what I would say is
actual
benefit.
There, they were looking at
hospitalization for GI bleed--they didn't
look at
death from GI bleed, but I wish they
had--but
hospitalization for GI bleed, and what
they found
was, in both of these studies, that
celecoxib was
actually more beneficial, you know, lower
rate of
hospitalization for GI than
rofecoxib. So, that is
the population, two large studies.
You have got your clinical
trials that
119
would have said it should be the
reverse. So, I
throw that out as one sort of conundrum.
The second is that I don't
think that the
actual benefits of these drugs are
understood well
enough to sort of try to weigh these very
well.
The case fatality rate for myocardial
infarction in
the United States approaches 40
percent. The case
fatality rate for hospitalized GI
bleeding is
probably somewhere around 5 or 10, it is
a much
lower case fatality rate.
Nobody that I have seen
anywhere has sort
of worked this out very well, so I would
submit to
you and to the committee that you
actually know
very little about the actual population
benefit of
any
of these products.
DR. WOOD: I don't think we are going to
get an answer to that question, so let's
move on.
Dr. Nissen.
DR. NISSEN: Let me briefly answer the
earlier question about what does the
hazard ratio
of 1.5 to 2 mean. Before I came to the
meeting, I
made a point to look this up, because I
thought it
120
would be very relevant.
It is equivalent to raising a cholesterol
from 200 to 260, or taking up
smoking. Another way
for the committee, I mean as a
cardiologist I have
to deal with this all the time, the most
effective
drugs we have for prevention of morbidity
and
mortality are statins, and they reduce
risk about
35 percent.
So, a hazard ratio of 1.5 to 2
is really a
very, very big effect when you are
talking about
the most common cause of mortality, and
that is why
this discussion is so important.
Now, my question is this. We are going to
be asked to balance risk and benefit, and
so the
magnitude of the hazard ratio is very
important to
all of us, and I am trying to reconcile
what we see
in the randomized control trials with,
let's take
rofecoxib for a moment, where it looks
like the
hazard ratio in the randomized trials is
in the
range of 2, 3, 4, maybe even higher, and
in the
observational data it is significantly lower.
I would like to propose a
hypothesis to
121
you and just ask you if you think this is
right.
In your observational data, you are
looking at
mostly short-term exposure, so you are
looking at
less than 12 months typically of
exposure.
It may well be that the hazard
increases
over time, so that by the time you get to
18
months, you can actually see it in a much
smaller
randomized trial, and so it doesn't rule
out the
possibility that, in fact, both
observations are
right, that, in fact, there is an early
hazard, but
that early hazard has a smaller hazard
ratio than
the hazard at 18 months or 24 months or
even 36
months, and if we ever were to look out 5
years, it
might still be increasing.
Do you think that is a
reasonable
hypothesis?
DR. GRAHAM:
I think more likely it is,
that in your clinical trials, early on
you don't
have enough power to distinguish the
risk. The
hazard is the same, but the lines are
closer
together, because we are closer to the
origin.
I think one other explanation
for the
122
lower risk ratios in observational
studies, I would
think is more likely due to
misclassification of
exposure and misclassification of
outcome. It is
likely to be nondifferential, so it would
tend to
reduce the odds ratios and relative risks
towards
1.
Exposure, because people are
going to take
it, a lot of these people are taking it
on a prn
kind of basis. In a clinical trial, you have a
greater certitude that they are actually
taking it
every day. That introduces a lot of
misclassification, so the a priori
hypothesis going
into an observational study, with
misclassification
going on, you are fighting an uphill
battle to see
an effect.
DR. WOOD: We have got lots of people who
want to ask questions. I want to make sure that
the people who are asking questions have
questions
they want to ask for clarification of the
speakers
who have spoken rather than just general
points.
Dr. D'Agostino.
DR. D'AGOSTINO: I have a couple of
123
questions along the way here. I have spent a good
part of my career in the Framingham Heart
Study,
and it's an epidemiological study and a
cohort
study, and we take joy when somebody runs
a
controlled trial on hypotheses and then
later on
confirms it.
The first question is I am
concerned that
even though you have gone through this
careful
analysis, your conclusions are no
apparent effect,
probably increased effect, probable
increased risk.
They really don't help us in the sense of
pinning
things down. We have a couple of very strong I
think good studies, the APPROVe study and
the APC
study as placebo-controlled trials.
Tell us quickly where is the
weight of how
we should look at these two pieces, the
controlled
trials we have versus what you have
produced.
DR. WOOD: Really quickly.
DR. D'AGOSTINO: Really quickly, it can be
done quickly.
DR. GRAHAM: My belief is that for the
controlled clinical trials, for the
levels of risk
124
that we are concerned about, that they do
not have
the statistical power early on to show
risk
differences.
DR. D'AGOSTINO: I think Bob O'Neil's
comment is very important here.
The other two points, and again
I will
make them quick, I am very concerned
about the high
dose effect you have, and I am really
concerned
about the MI and the number of
cases. I mean blood
pressure, cholesterol, diabetes, smoking,
this is
what drives people to have heart attacks
and what
have you, and that is completely missing
on your
assessment of how many new cases, so I
guess it is
more of a comment that I am really
concerned that
that sheet needs sobering interpretation.
DR. GRAHAM: But it was based on the odds
ratios and relative risks where those
factors were
adjusted for, so as well as they are
adjusted for,
that is what the projection represents,
the excess
after adjustment.
DR. D'AGOSTINO: Yes, but I mean the
comment was made by you, throwing in the
analysis
125
doesn't necessarily adjust for them.
The last one, you made a very
nice point
about the cardio-protective effect, and
you tried
to show that these uses, and what have
you, somehow
or other all have the same risk, and your
interpretation that there must be some
confounding
going on, why doesn't that hold for all
the studies
you gave, why don't that hold for the
Solomon
study, which you thought was a great
study, yet,
this one result you don't like?
DR. GRAHAM: For what, the Kimmel study?
DR. D'AGOSTINO: Wasn't it the Solomon
study that had the naproxen as the
cardio-protective?
DR. GRAHAM: That is because the cardio
protection was present when they were on
the drug
and
when they weren't on the drug.
DR. D'AGOSTINO: I understand what you are
saying, but if that's a problem, then, it
means
there is some confounding going on.
DR. GRAHAM: No, it's selection bias.
DR. D'AGOSTINO: Well, it's selection
126
bias, but why isn't it for the whole
study? Why do
you throw out a result you don't like and
keep all
the results you like?
DR. GRAHAM: No, that is not what I did.
I pointed out a result where they showed
the
presence of the selection bias. In other studies,
the Ingenix study is the only other study
that
looked at this. I don't have a slide of it.
DR. D'AGOSTINO: I don't know if it's a
selection bias or misinterpretation of
the data.
DR. GRAHAM: Well, to me it looks like
selection bias.
DR. WOOD: Let's continue that
conversation later.
Dr. Morris.
DR. MORRIS: David, would you go to slide
14.
That is the risk, the duration of use.
I
think one of your points was that if you
look at
your study, tell me if I understand this
right,
that with the lower dose, that the median
time to
an AMI is sooner than with a higher dose,
did I
understand that right?
DR. GRAHAM: Yes.
DR. MORRIS: A month?
DR. GRAHAM: Had more cases, a greater
127
proportion of our cases, but the other
thing is
remember, down here, we are talking about
18 cases
or so.
The N here is small, the N here is like 58,
and the N here is 10. So, I wouldn't read too much
into the difference.
The more important point is
that at the
low dose, nobody was out there beyond 18
months, so
all the action happened before 18 months,
and the
same for the others. I see what you are saying. I
can only say that is what our data were.
DR. MORRIS: One interpretation is what
you said earlier, that for this
particular drug, we
are talking about, as you said, no safe
level. I
was wondering if that is the way you
interpreted
it, that because we are talking about
Vioxx here,
and there is no safe level, that
something is going
to happen sooner, or is it something with
the
populations are different.
DR. GRAHAM: The populations could be
128
different, but I think, you know, you
would expect
the
higher dose to have a shorter latency to onset
than the higher dose, but the numbers are
so small.
DR. MORRIS: Okay, it's a small number
problem.
DR. WOOD: So, the answer is too small
numbers at high dose.
Dr. Boulware.
DR. BOULWARE: I just want to make sure I
understand something that you had
proposed in your
excess population risk slide, if you
would put that
back up.
As a rheumatologist, I use
these drugs in
a population much greater than what you
have here
with a 65 to 74 where the risk of an MI
is fairly
high in that group.
Did you want us to believe that
this
excess risk that you are proposing would
be
extrapolated to other population groups,
too?
DR. GRAHAM: Well, no.
DR. BOULWARE: Do you have any numbers
that may demonstrate that?
DR. GRAHAM: Well, the answer to the
second is no. This was an example in
conversation
with people planning the talk, to try to
help
129
people connect with what it means.
Cardiovascular risks go
up. I mean in the
next age group higher, the risks are
higher. In
the age groups lower, they are lower, but
cardiovascular risk begins to increase in
the 40s.
DR. BOULWARE: I understand, but it
wouldn't be a linear type of thing.
DR. GRAHAM: No, the background risk isn't
linear, the relative risks, though, are
adjusted
out.
DR. BOULWARE: Because one of the
questions we will be faced with is are
there
subpopulations or groups that these may
be safe in,
and I just want to make sure I understand
the
relative risk in different age groups.
DR. GRAHAM: Nobody in any of the studies
where they have looked at it have
reported effect
modification, which would be that the
level of risk
differs at different ages.
DR. BOULWARE: One more question here. I
want to make sure I understand. I think I heard a
comment that says when the risk
approaches
2.0--maybe I just assumed that you said
this--that
it was an unacceptable level of risk.
Is there ever a case where a
drug may have
130
a clinical benefit in which that risk is
acceptable, because for the patients I
see, not
giving them any of these drugs will
confer a great
deal of risk on them, and physical
impairment, and
we have studies that show that the
functional
classification of rheumatoid arthritis
patients
carries with it a significant mortality
as that
class goes up?
DR. WOOD: I think that is a question for
the committee to answer rather than Dr.
Graham.
Let's move on to Dr.
Cryer. Do you have a
question?
DR. CRYER: I do.
The comment and
question I have of Dr. Graham addresses
an issue
that I think is an important difference
between the
observational studies and the prospective
studies,
131
and this difference relates to assessment
of drug
compliance and missed doses, and I think
it is
critical as it relates to assessing drugs
which
potentially affect platelet function.
A huge difference, as you know,
between
aspirin's effect and every other NSAID
including
the COX-2 inhibitors, is that with the
non-aspirin
NSAIDs, as soon as you remove the drugs,
whatever
potential effect they would have had on
the
platelet are immediately reversed.
So, with naproxen specifically,
my
preconceived bias, which may be wrong,
but my
preconceived bias based upon everything I
know
about the pharmacology and the things
that Dr.
FitzGerald has reviewed for us, is that
it should
have some mild anti-platelet effects
which would
only be present when the drug is on board
in the
system.
So, the specific question is,
in the
observational studies, recognizing that
in clinical
practice people miss doses of their
NSAIDs, they
are not taking their NSAIDs consistently,
how do
132
you account for the missed doses in the
observational studies recognizing that
this could
potentially lead to a mitigation of
whatever
negative effect or positive effect that
they may
have?
DR. GRAHAM: It ends up being
misclassification. Generally, what that
means is it
will force the observed level of risk,
the relative
risk of the odds ratio closer to 1. So, if we had
an increased risk, it would make it
lower, if we
had a protective effect, it would sort of
make it
higher, closer to 1.
DR. CRYER: Right, we agree on that. The
specific question is, is there a way to
actually
recognize or to account for when people
do not take
their doses in the observational
databases?
DR. GRAHAM: No, there isn't, so when you
are studying, say, an increased risk,
that is why I
said if you find something, you have to
realize you
found it despite the misclassification.
DR. WOOD: Okay.
Dr. Domanski.
DR. DOMANSKI: I will save it for
133
tomorrow.
DR. WOOD: Okay, great.
Dr. Furberg.
DR. FURBERG: No.
DR. WOOD: Okay, great.
Dr. Temple, who does speak for
the FDA.
DR. TEMPLE: I am just asking questions.
A couple.
Actually, one point is it seems to me
that since we expect that people are
going to be
getting one drug or another, comparisons
with other
NSAIDs seems like as good a comparison as
we should
make.
You might want to leave out indomethacin if
you are worried about it. That's one thing.
I guess my main question,
though, is
everybody has paid appropriate lip
service to the
idea that very small differences are hard
to
interpret in epidemiology.
People have said 1.5, 2. Actually, I
notice in one of his editorials, Dr.
Furberg cited
a paper of mine where I said anything
less than 2
really needs a lot of questions. Jerry Cornfield,
who sort of invented all this stuff, used
to say 3.
Well, we are talking about
differences
134
here that are 0.1 differences, not that
they
wouldn't be hugely important if they were
true,
that is absolutely true. So, I guess I want to
know what Richard and you make of all
this, because
the numbers are very small, and yet, just
as an
example, there is a very great consistency
that you
cite that celecoxib looks sort of okay,
but you
found one study where there is a little
hint that
maybe the higher dose is a problem, and
since
probably we all think dose response is
likely, that
looks good to you.
DR. GRAHAM: Two studies, there were 2.
DR. TEMPLE: Okay, 2.
The valdecoxib
data, which shows nothing, doesn't look
so good
because we probably all believe that
there is
likely to be a class effect.
What I am asking is, with
numbers like
this, how do you know what to do with
them? That
seems very fundamental for the
epidemiology.
DR. WOOD: But, Bob, there are 4
randomized clinical trials here, and your
comments
don't apply to them, I assume.
DR. TEMPLE: No, they don't, although they
are not perfectly consistent either. But, no, I am
asking, what do we make of differences of
this
135
magnitude with everybody having given lip
service
to the idea that small differences are
hard to
interpret, and yet we seem to be
enthusiastically
endorsing them, so I just want to know
what Richard
and David think about that.
DR. GRAHAM: Rich, do you want to go
first?
DR. PLATT: I think we have to be cautious
about how we interpret it, so I would say
the
finding of a relative risk of 3 in an
epidemiologic
study, as David found, is meaningful--
DR. TEMPLE: For high dose rofecoxib.
DR. PLATT: For high dose rofecoxib.
DR. TEMPLE: I would not dispute that at
all.
DR. PLATT: It seems to me that in that
context, that a dose response effect,
that the
information about lower doses gains
weight by
borrowing from that. I think that is also worth
136
keeping in mind when, in other studies
that are
working in that range that make us all
nervous,
there appears to be a dose response
effect.
It is the kind of consistency that
makes
the study, in my mind, be worth more
attention. I
think there is something to be said for
giving more
weight to relatively small excess risks
if they are
seen in a number of different
environments when we
can't have good reason to think that
there is a
similar kind of biases that might be
contributing
to it.
After that, I agree with
you. We are in
relatively difficult terrain. I think that it is
not the same as no data, though. I think we ought
to distinguish between the situation in
which we
have no evidence from ones in which we
have
relatively weak evidence.
We didn't talk at all, for
instance, about
the enormous number of spontaneous
reports of
myocardial infarction following exposure
to
nonsteroidals. There are thousands and thousands
of them.
In my mind, they don't contribute at all
137
to the discussion, whereas, I think these
need to
be weighed in the mix when we don't have
clinical
trial information to depend on.
DR. GRAHAM: My answer is similar to his,
but I think that what you are identifying
is, is
that we are hitting or at least right now
the
frontier is the limits of what the
available tools
we have to define the levels of risk that
we are
talking about.
We are talking about small levels
of risk
that turn out for this particular event
to be
enormously important in a population
level. If you
are talking liver failure, we wouldn't be
having
this conversation. For that reason, it becomes
important and what I would say is sort of
emphasizing what Rich said, is I would be
looking
for consistency across different studies,
and if I
found a number of studies, say, as with
Indocin,
for example, to me, that is more
persuasive.
If I found a number of studies
that
pointed to a particular set of NSAIDs
that seems to
have low risks, I would take comfort in
that in the
138
absence of perfect information. I mean some light
in a storm is probably better than no
light In a
storm.
DR. TEMPLE: I take it if the differences
were at the level of 10 percent, 1.1
versus 1.2--
DR. GRAHAM: I am thinking more in a very
qualitative sense of things that they
seem to
cluster around 1. I mean 1.1 for ibuprofen, it
could be that, for example, may naproxen
increases
the risk 3 percent in the real world, we
are never
going to figure that out, maybe ibuprofen
increases
it 10 percent or 15 percent, maybe we
could figure
that out, I don't know, but there is
going to be a
place where qualitatively, if we see
enough studies
kind of sort of pointing to the same
place, you
know, most of them, they are not all
going to say
the same thing, there is going to be
these
conflicts, just like we have in clinical
trials
data.
But if most of the compass
arrows are sort
of pointing in the same direction for
particular
NSAIDs, I think those are the ones that
at least
139
that I sort of place on a suspect list.
DR. TEMPLE: So, very low hazards need at
least multiple support before they are
credible.
DR. GRAHAM: I think so, and I think that
you want to try to encourage to collect
that
information sort of to test that out.
DR. TEMPLE: Alastair, could I take half a
second to answer a question Larry raised
before?
DR. WOOD: Sure, a second.
DR. TEMPLE: Well, it's a very good
question, you know, if the drug is going
to be used
forever, why don't you study them
forever. The
only thing I would point out here is that
what sort
of started people thinking was VIGOR, and
VIGOR
didn't take 3 years to show anything, it
showed up
in 9 months.
So, what you have seen is for,
say,
lumiracoxib, a humongous study of about
the same
length, but, of course, they didn't know
about
APPROVe, did they, and whatever you think
APPROVe
means, whether Bob is right that it's
late, or
David is right that there weren't enough
cases,
140
people were pointing toward a study that
by every
reasonable thought, if you think
platelets are
involved, ought to be long enough to show
things
up.
But then you form a new
hypothesis once
you have APPROVe, and you have to adapt
it, and I
think that goes on all the time. It would not be I
must say for most things my first thought
unless
you are looking for cancer that you need
a 3-year
study to find it, but maybe you learned
that it
does.
Just for what is worth as an
example, you
can't get an anti-arrhythmic drug
approved in this
country without showing that you don't
alter
survival unfavorably. One result is there are
hardly any being developed, but, you
know, we had
bad experiences, we didn't like the
results of
CAST, so you change.
I think there is no doubt that
things
evolve and you have to expect that, and
APPROVe,
depending on what you think of it,
changes the
nature of what you expect.
DR. GRAHAM: Bob, just one point on that.
I think if the APPROVe study had been 5
or 10 times
larger than it was--I am talking about
retrospect
141
now--you would be able to answer with
much greater
confidence what is happening month 1 to
18. I
guess what I am saying is that you could
also
shorten the latency to identification of
a problem
if it turns out that the risk is early
on.
DR. TEMPLE: David, I think that is
entirely possible, and if it involves
platelets, I
would believe you, but if it involves a
small,
long-term increase in blood pressure,
then, I am
not so sure.
DR. GRAHAM: Right, but we saw yesterday--
DR. TEMPLE: We don't know.
DR. GRAHAM: We don't, but if it's
prostacyclin, that effect could occur
immediately.
DR. TEMPLE: Yes, but the blood pressure
effect could be delayed.
DR. WOOD: Right.
So what, Bob, you are
saying is that it is easy to be a Monday
morning
quarterback, but the data were not there
before.
DR. TEMPLE: I would never be that rude.
DR. WOOD: I think you are right.
Dr. Stemhagen.
DR. STEMHAGEN: I would like to clarify a
couple things. First, I am a little concerned in
terms of the unpublished data. I appreciate that
142
we are able to get data very quickly,
right at the
minute that it is being generated, but
none of us
have had a chance to really review that,
so I do
have some concerns about the weight putting
on this
unpublished data when the rest of us
haven't had a
chance to look at it.
I think there needs to be some
clarification. There was some discussion
about the
recall bias, and so on. Certainly, there is
a major
concern about that in case- controlled
studies, and
we don't have the questionnaires, but
there were a
lot of sort of subanalysis done in the
Kimmel
study, about trying to look at whether
recall bias
is a problem, and I am not sure that you
have
highlighted that enough that looking at
all those
different things, there were really no
differences
143
found.
Similarly, in the Watson study,
it's a
GPRD study, it is different than a lot of
the large
databases, the automated databases.
There is a lot more personal
involvement
in terms of the data and the data
collection and
the
adjudication of results, and I think it just
needs to be clear that all of these
studies are not
the same in terms of a Medicare study
where we
can't go back and validate records. A lot of them
had a much more careful review, and I am
just not
sure that that was totally clear and if
you hadn't
read each of the papers.
I would like to just ask a
question in
terms of your definition of the inception
cohort,
if you could just go over that again,
because of
your comments about the short-term use.
DR. GRAHAM: Inception cohorts are where
people enter the cohort with their
first-time use
of a specific agent, so it's basically
like an
incident cohort, it's new users. That is to be
distinguished from a prevalence cohort
where
144
starting January 1st, everybody who was
on an NSAID
is in our cohort.
Some of those could be people who were
on
it before January 1st, and others could
be people
who start an NSAID after January 1st, so
you are
mixing people who are prevalent on the
drug, who
may have survived, or whatever, and
people who are
newly starting it.
In those types of cohort
studies, a new
user analysis was designed to focus on
those people
who, during the study window, were new
initiators
of the particular drug under study, so
that time
zero could be identified for those
people.
That is what Alec Walker &
Company did in
their Ingenix study. It was a prevalence cohort,
but they did a new user analysis in which
they
identified new users, and it was that new
user
analysis that showed the 1 to 30-day
increased
risk.
Wayne Ray did the same thing in
terms of
new user analysis, and in our study, the
nested
case control, everyone was an inception
user in the
145
base population.
DR. STEMHAGEN: I guess just a comment in
terms of people thinking about clinical
trials
where we have washout periods, is that
people are
really switching.
If they are RA or OA patients,
they are
not starting new with the drug, they have
been on
something for a long time, and they are
switching.
So, we have to think about those risks in
terms of
the weight we are putting onto that
inception
cohort, as well.
I guess the last point is based
on the
question that Ralph had about the other
studies. I
just want us to keep in mind also that a
lot of
those studies come from very unique
populations -
the randomized clinical trials, the colon
polyp
study, and the Alzheimer's disease
patients, so are
very different.
We can't tease out in any of
these
observational studies whether we have
patients that
meet those criteria or have those
indications, as
we also pointed out.
DR. WOOD: Tom.
DR. FLEMING: I think Drs. Platt's and
Graham's presentations were informative,
but with
146
certainly a lot of complexities for
methodologic
issues that I assume tomorrow, we will
give our
perspectives about, so let me ask a
question and
then a clarification.
The question relates to the
slide on the 4
positive naproxen studies, I think slide
22. While
you are getting that, just very quickly,
these
large linked databases certainly are very
useful
from the perspective of getting defined
populations
with numerators and denominators, but
have many
challenges that people have been talking
about
along the lines of lack of randomization,
no
confounder information, specificity and
sensitivity.
Bob O'Neil got at a point that
I think is
critical, and that is the complexity of
not having
a time zero cohort with the ability to do
what
would be the analogous ITT analysis with
complete
follow-up or minimize loss to follow-up.
You bring out in the Solomon
example
there, David, a very nice illustration of
this very
point that you recognized, which is the
selection
bias that can go on when you are
characterizing
people into these groups, and it's
misleading to
think that you are really seeing the
causal effect
147
of any use versus current, versus recent,
versus
remote, the causality could be going in
the other
direction.
Intrinsic differences in
patients could be
influencing whether they are, in fact, in
those
four categories. But don't you, in essence, even
though your conclusion might be right,
aren't you,
in essence, doing the same thing at the
top when
you are looking at naproxen, say, when
you are
looking at other NSAIDs, it is
protective, but you
don't know whether it's, in fact, truly
the harmful
effect of the other NSAIDs, so you try to
get in a
non-use population, you are trying to
simulate a
placebo, but how do you know that those
non-use
people weren't intrinsically better? Isn't it the
same issue?
DR. WOOD: I think we have had this
discussion.
DR. FLEMING: But this is important, I
want to get his views, because it's
important for
naproxen.
DR. WOOD:
Okay.
DR. GRAHAM: There is no perfect reference
group. It turns out that this non-use
group is
really they are remote users, but it is a
question
148
and I can't answer it except to say that
when you
adjust for all the confounders you are
able to
measure, you try to remove those effects,
but there
still could be effects that you cannot
remove.
The data are what the data are,
and here
what I was trying to show is that based
on--if
these data were looked at the way most of
the other
studies were done, it gives a very
different
result.
If it turns out that all of the
NSAIDs
increased the risk a little bit, the fact
that
naproxen doesn't increase it as much,
could look
protected, and you really don't know.
The real conundrum is to get an
anchor
point to help you interpret everything,
and there
is no perfect anchor point.
DR. FLEMING: Your motivation for wanting
to know what the placebo-controlled
result is, is
clear and justified. This analysis, though, has
the same potential flaws as the Solomon
analysis.
So, the motivation for the question is
clear, as
you are just restating, but the
reliability of the
conclusions are suspect for this very
reason that
you correctly noted, due to the
selectivity in the
Solomon categorization.
149
DR. GRAHAM: You need then to sort of
generalize that to all of the
observational
studies, because all of them, you had--
DR. WOOD: Why don't we continue this
conversation later, and, Tom, you can
present
discussion on that later.
DR. FLEMING: Well, there is much more to
say, but I will defer to tomorrow.
DR. WOOD:
I am sure there is.
Dr. Hennekens will be our last
question
150
before the break. Just to encourage you, we will
be back here just after 20 to, so make it
fast.
DR. HENNEKENS: A question and a comment.
Ten years ago, a large body of basic
science,
clinical studies, case-control, and
prospective
cohort studies consistently showed that
patients
with hypertension prescribed calcium
blockers had
1.5 to 2-fold increased risk of MI even
after
controlling for a large number of
available
confounders.
I wrote a JAMA editorial asking
for
randomized evidence, but I assume, based
on what I
heard you say, that you would have asked
the agency
to withdraw the drugs. So, I would ask you to
consider whether protecting the public
from harm is
an optimal goal.
It is far more simple and
straightforward
than trying to maximize benefit and
minimize harm,
which would do the most good for the most
people,
but doing the most good for the most
people does
not, strictly speaking, protect the
public from
harm.
DR. GRAHAM: Do you want a response to
that?
Okay. I think that when you are
faced with
a large risk that affects large numbers
of people,
151
and has a large consequence, that you
don't have
the luxury of time to wait 10 years to
get
clarification on the issue, and you have
to use
what data you have available at the time.
I think that just as we have
imperfect
measures of risk, I would submit that we
have even
more imperfect measures of actual
benefit. In the
case of hypertension, I think, you know,
that has
been studied dramatically and we actually
know that
not all antihypertensives lowering blood
pressure
at the same amount, confer the same
population
benefit.
I would say that with this
class of drugs,
we really haven't even demonstrated--I
mean
yesterday, the question came up why would
a company
do a study on polyp prevention, had they
thought
about what the benefit of this was, and
nobody had
started to think, well, how many lives
are we going
to save by giving people these drugs, and
I would
152
submit that if you were to ask the agency
or ask
the company on this, if you don't have a
good
measure on benefit, so you want to make a
benefit-risk assessment.
We have measures of risk, they may
be
imperfect, but I would argue that from a
population
perspective, you don't really have nearly
as good
information as you might believe you do
from the
clinical trials, what the benefit in the
population
is, how many lives are actually saved by
the
COX-2s, for example.
DR. WOOD: On that note, I am told the
lines are building at the men's room, so
we need to
be back here at exactly quarter to.
(Recess.)
DR. WOOD: Let's get going.
Arcoxia (etoricoxib)
Merck Research
Laboratories
Sponsor Presentation
Sean P. Curtis, M.D.
DR. CURTIS: Mr. Chairman, members of the
Joint Advisory Committee, FDA, ladies and
153
gentlemen: My name is Dr. Sean Curtis, Senior
Director, Clinical Research, at Merck and
Company,
and I would like to thank you for the
opportunity
to review data from the Etoricoxib
Development
Program.
I believe the committee will
find these
data informative and contribute to the
further
evaluation of this therapeutic class, a
goal we all
share collectively.
Drs. Konstam and Loren Laine
are serving
as consultants today and are available as
a
resource to the committee.
Following an introduction,
results from
the development program will be
summarized
beginning with efficacy, followed by a
review of
the safety findings. I will first review the
gastrointestinal and renovascular safety,
followed
by thrombotic cardiovascular safety.
I will then review the design
of three
studies, which together are designed to
further
characterize and assess the
cardiovascular safety
of etoricoxib in arthritis patients.
154
Cardiovascular safety data from the first
of these
three studies, the EDGE study, will be
reviewed,
and I will conclude with a summary.
My presentation will focus on the
following points. Etoricoxib, as a selective COX-2
inhibitor, has a role among the current
treatment
options for patients with diseases and
conditions
characterized by pain and inflammation.
Supportive data will be
reviewed, namely,
efficacy that has been demonstrated to be
similar
and, in some cases, superior to NSAIDs,
specifically naproxen 1,000 mg;
gastrointestinal
safety and tolerability, favorably
differentiated
from NSAIDs; and a renovascular safety
profile,
which is dose dependent and generally
similar to
the effects observed with comparator
NSAIDs at
therapeutic doses.
With regards to thrombotic
cardiovascular
safety, cardiovascular events occurred at
a similar
rate on etoricoxib as compared to
non-naproxen
NSAIDs over the course of approximately 1
year.
Data are currently limited beyond 1 year
of
155
treatment, and events occurred at
different rates
in comparison to naproxen.
The other key point for my
presentation is
that large, randomized clinical trials
are
currently ongoing to further characterize
the
long-term cardiovascular safety of
etoricoxib as
suggested by many members of this joint
committee.
These results will provide a
full
characterization of the cardiovascular
safety
profile of etoricoxib in arthritis
patients as
compared to diclofenac.
These data are critical to the
current
scientific debate over cardiovascular
safety.
Specifically, we will address whether the
long-term
cardiovascular safety of a selective
COX-2
inhibitor is similar to, or different,
than that of
a traditional NSAID.
Let's begin reviewing the data.
Etoricoxib represents a
distinct chemical
entity. It consists of a bipyridine ring
with
methyl sulfone side chain. In the clinical dose
range, etoricoxib has demonstrated
selectivity for
156
the COX-2 enzyme using human whole blood
biochemical assays.
Its absorption is rapid with a
peak plasma
concentration achieved by approximately 1
hour and
with an effective half-life of
approximately 22
hours, it is suitable for once daily
dosing.
Etoricoxib is currently
approved in
approximately 60 countries. Core indications
include osteoarthritis at a once daily
dose of 60
mg, rheumatoid arthritis at a once daily
dose of 90
mg,
and acute gouty arthritis. The dose is
120 mg
for the acute symptomatic period only.
In the United States, the FDA
issued an
approvable action on our new drug
application.
I would now like to summarize
efficacy.
The efficacy of etoricoxib has been
demonstrated
across a range of conditions and diseases
characterized by pain and inflammation.
For these conditions, efficacy
data have
been published or accepted for
publication
including 3 diseases and conditions for
which an
indication is not currently granted in
the United
157
States for a selective COX-2
inhibitor. These
include studies in chronic low back pain,
ankylosing spondylitis, and acute gouty
arthritis.
As you will remember, the acute
gouty
arthritis data were discussed with the
Arthritis
Advisory Committee in June 2004 in the
context of a
committee meeting design to look at gout
study
designs.
Efficacy data are summarized in
your
background package, however, I would like
to draw
your attention to results obtained in
three
specific disease models.
The rheumatoid arthritis
program included
2 pivotal double-blind, placebo and
active
comparator- controlled studies in
approximately
1,700 patients. In one study, etoricoxib 90 mg
demonstrated efficacy that was
statistically
superior to naproxen 1,000 mg for all
primary
endpoints and all additional endpoints
including
the ACR20.
In the other study, etoricoxib
demonstrated efficacy that was similar to
naproxen,
158
and in patient with the ankylosing
spondylitis, we
performed a single pivotal double-blind,
placebo
and active comparator-controlled study
which
enrolled approximately 390 patients.
Over the 52-week treatment
period,
etoricoxib demonstrated efficacy that was
statistically superior to naproxen 1,000
mg for all
3 co-primary endpoints, and in patient
with acute
gouty arthritis, we performed 2
double-blind,
active comparator-controlled studies
enrolling
approximately 350 patients in total.
In those studies, etoricoxib at
a dose of
120 mg for 7 days demonstrated efficacy
that was
comparable to indomethacin.
I would now like to begin
reviewing the
safety data.
The gastrointestinal safety
program, as
summarized in your background package,
was designed
to evaluate the entire GI tract. Clinical outcomes
based on pooled data from the entire
development
program were prespecified for
analysis. These
include a combined analysis of upper GI
clinical
159
events, or PUBs, and a combined analysis
of GI
tolerability.
Here are summarized results
from the
prespecified combined analysis of upper
GI clinical
events which occurred in Phase IIB and
III studies
from the entire development program by
displaying
the cumulative incidence of confirmed
events by
treatment group over the entire duration
of the
studies involved in the analysis.
As you see, a statistically
significant
relative risk of 0.48 favoring etoricoxib
was
demonstrated. This represents a 52 percent risk
reduction. It was observed early and maintained
over the entire study duration. These results are
largely driven by comparisons to
naproxen.
For purposes of summarizing
renovascular
safety, we will focus on data from the
osteoarthritis and the rheumatoid
arthritis
studies, which represent the majority of
the data.
Presenting results by disease types
ensures the
patient characteristics are similar among
the
treatment groups.
This slide displays the
incidence of
hypertension adverse experiences by
treatment group
observed over a 12-week treatment period,
in OA
160
patients on the left, and RA patients on
the right.
In the OA population, the dose
response is
observed most clearly from 30 to 60 and
60 to 120
mg, 90 mg is outlying likely due to the
smaller
sample size, and in the RA population,
the dose
response is also observed although less
evident as
compared to osteoarthritis.
Overall, the rates observed for
etoricoxib, specifically the doses
indicated for
chronic use, that is, 60 and 90, are
within the
range observed with comparator NSAIDs,
numerically
higher than naproxen, numerically lower
than that
observed with ibuprofen, and in both
patient
populations, it was rare for patients to
discontinue from this adverse experience
with no
clear difference observed between
treatment groups.
In addition to hypertension, we
looked
closely at adverse effects related to
edema and
congestive heart failure. Tabulated here are the
161
incidence of congestive heart failure
adverse
effects as spontaneously reported by
investigators
in our placebo-controlled population of
up to 12
weeks duration.
As you see here, incidences are
low among
the active treatment groups. I would like to show
you the cumulative incidence of
congestive heart
failure adverse events which occurred
over the
entire duration of our chronic exposure
studies.
We see here that etoricoxib as
compared to
comparator NSAIDs pooled are associated
with
similar rates of congestive heart failure
adverse
events.
The grouping of terms is indicated on the
bottom of the slide.
The data provided in your
background
package and summarized thus far support
the
improved gastrointestinal safety and
tolerability
of etoricoxib compared to non-selective
NSAIDs,
with clinical outcomes data including
PUBs and GI
intolerance endpoints, as well as
endoscopic data.
These data also provide
evidence of the
renovascular profile of etoricoxib, that
is,
162
hypertension, edema, and heart failure
are dose
related as would be expected, and generally
similar
to the effects observed with comparator
NSAIDs, in
some cases numerically higher and in some
cases
numerically lower.
I would now like to move on to
cardiovascular safety data review. The process
that Merck instituted for prospectively
adjudicating all potential thrombotic
events as
described by Dr. Braunstein yesterday for
rofecoxib, was operative for etoricoxib
from the
beginning of Phase IIB.
We prespecified an analysis of
all such
events using individual patient data from
studies
of at least 4 weeks in duration across
the clinical
development studies.
In total, there were 12 studies
included
in this analysis including approximately
6,700
patients and 6,500 patient years of
exposure. For
the analysis, comparisons of etoricoxib
were made
to placebo or active comparator NSAID
using data
only from the studies that contained the
treatments
163
being compared.
The etoricoxib group and
analysis you will
be seeing shortly consists of data
combined from
doses of 60, 90, and 120 mg in order to
improve
statistical precision, and for the
comparison to
NSAIDs, naproxen was compared to
etoricoxib
separate from the other 2 NSAIDs,
diclofenac and
ibuprofen, based on the fact that
naproxen is
distinct pharmacodynamically from both
ibuprofen
and diclofenac, and because qualitative
differences
were observed in the comparison to
naproxen versus
the comparison to non-naproxen NSAIDs.
The endpoint specified as
primary for the
assessment of cardiovascular safety in
the
etoricoxib development program was a
composite
endpoint of all confirmed thrombotic
events
confirmed by the Adjudication Committee,
and
includes cardiac, cerebrovascular, and peripheral
vascular events.
The primary results for the
pooled
analysis are summarized here by
presenting the
point estimate of the relative risk and
the
164
corresponding 95 percent confidence
interval for
the comparisons of etoricoxib to placebo,
to
non-naproxen NSAIDs, and to naproxen for
the
composite primary endpoint of confirmed
thrombotic
events.
The naproxen-controlled data
set is the
largest of the 3 data sets, and the
placebo-controlled data is the smallest
of the 3.
This is indicated numerically on the
right in
patient years at risk and correspondingly
reflected
by
the size of the triangle representing the point
estimate of the relative risk.
When comparing etoricoxib to
placebo and
to non-naproxen NSAIDs, the relative risk
approximates 1.0 indicating no
discernible
difference in thrombotic cardiovascular
events
between those treatment groups.
However, it is important to
keep in mind
that the maximum duration of the
placebo-controlled
period was 12 weeks, and when comparing
etoricoxib
to naproxen, the relative risk is greater
than 1,
indicating a difference between the 2
treatment
165
groups in a trend favoring naproxen in
that
comparison.
Shown here are the cumulative
incidence of
confirmed thrombotic events in the
non-naproxen-controlled data set. The amount of
data are limited at longer term time
points
particularly for the non-naproxen NSAID
group.
In total, the event rates are
similar
between treatment groups.
All individual events were
categorized by
the Adjudication Committee as either
cardiac,
cerebrovascular, or peripheral vascular. In
reviewing the specific events in the
non-naproxen-controlled data set, using
this
categorization, cardiac and
cerebrovascular events
were observed in both treatment groups.
Numeric differences between
treatment
groups trended in both directions and
were observed
at the level of individual events.
As indicated previously, the
largest of
the 3 data sets is the data set which
compares
etoricoxib to naproxen. As you can
appreciate from
166
these cumulative incidence curves, the
etoricoxib
and naproxen groups separate early with a
lower
cumulative incidences observed on
naproxen as
compared to etoricoxib.
In the naproxen-controlled data
set, the
specific confirmed thrombotic events
occurred in
all 3 vascular events. In considering the overall
difference between the
naproxen-etoricoxib group,
no single event predominates, however, a
higher
incidence of ischemic cerebrovascular
strokes was
observed on etoricoxib in this
comparative data
set.
Analyses were performed to
explore the
relation between dose of etoricoxib and
rate of
thrombotic events. The left two panels
summarize
the results of a pair-wise analysis, an
approach
that includes data only from studies that
contained
the doses being compared.
The righthand panel represents
results
using a summary approach, which
incorporates rates
by dose from all studies in the pooled
cardiovascular analysis.
The data do not indicate
evidence of a
dose effect across the 60 to 120 mg
etoricoxib dose
range.
167
Summarized in your background
package are
results of subgroup analyses from the
naproxen-controlled data set including
patients at
increased baseline cardiovascular risk
and by
arthritis disease type particularly OA
versus
rheumatoid arthritis.
These subgroup analyses, as
well as
additional analyses including those subgroups
identified to be potentially at increased
risk
based on the rofecoxib APPROVe study
failed to
identify any specific patient subgroup at
increased
relative risk for thrombotic event.
It is important to remember,
however, that
the amount of etoricoxib cardiovascular
safety data
currently available do not allow us to
make firm
conclusions for any specific subgroup.
All-cause mortality in the
etoricoxib
development program is summarized here as
rates per
100 patient years by treatment
group. Included, as
168
well, are results from the EDGE study, a
study of
approximately 1 year's duration in over
7,000
osteoarthritis patients comparing the GI
tolerability of etoricoxib to diclofenac.
Rates for etoricoxib and
non-naproxen
NSAIDs in the left panel are similar and
numerically higher than those observed on
naproxen
and placebo, which are similar to each
other. The
rates here are represented as a point
estimate with
a corresponding 95 percent confidence
interval.
As you see, the confidence
intervals are
broad and overlapping between the
treatment groups.
Based on these data, there is no evidence
for a
true difference in all-cause mortality
between
treatment groups.
In the EDGE study, on the
right, rates
were numerically similar between treatment
groups
in all-cause mortality again with
confidence
intervals that overlap the point
estimates between
treatment groups, at this point
indicating no
evidence of a difference.
The cardiovascular safety data from the
169
original development program can thus be
summarized
as follows. There is no clear evidence of a
difference between etoricoxib and placebo
based on
limited amounts of short-term data.
There is no discernible
difference in
cardiovascular event rates between
etoricoxib and
non-naproxen NSAIDs. This comparison is limited,
however, by the amount of active
comparator-controlled data with both
diclofenac and
ibuprofen, and naproxen, at a regimented
dose of
500 mg twice daily is associated with a
lower rate
of thrombotic events as compared to
etoricoxib.
As you saw from the
Kaplan-Meier curves,
the cumulative incidences, a difference,
separates
early, and is, in fact, this is an
observation that
has been seen with the rofecoxib data and
similar
to the observations made from the
lumiracoxib
TARGET study, which we will be hearing
about later.
Recent results from long-term
placebo-controlled studies with rofecoxib
and
celecoxib have important implications for
etoricoxib. Specifically, these recent data
170
showing a difference in cardiovascular
safety in
long-term studies versus placebo do, in
fact,
suggest a class effect.
Despite the large size of the original
development program, over 10,000
patients,
approximately 5,800 of which were
receiving
etoricoxib, there are limitations on the
amount of
accrued cardiovascular safety data. Specifically,
the long-term data were limited in
quantity, and
limited primarily in comparison to
naproxen.
Because of questions raised
with respect
to naproxen, we decided we needed a
different
approach to accrue additional data, and I
would now
like to review the strategic approach we
took and
then discuss the specific studies that
resulted.
Our primary objective was to
further
establish the long term general and
cardiovascular
safety of etoricoxib in arthritis patients who
required treatment. At the time the strategy to
meet this objective was formulated, there
were
ongoing long-term placebo-controlled
studies with
other selective COX-2 inhibitors, largely
focusing
171
on exploring novel indications for
cyclooxygenase-inhibiting therapies.
Examples
include Alzheimer's disease and
chemoprevention.
For etoricoxib, rather than
explore novel
indications with placebo-controlled
studies, we
chose to further evaluate the group of
patients who
required treatment for arthritis. Therefore, the
plan we developed was to perform active
comparator-controlled studies in
osteoarthritis and
rheumatoid arthritis patients.
Studying this patient
population ethically
precluded use of a placebo for more than
a short
period of time, because these patients
require
active treatment. Diclofenac was chosen
as the
active comparator, and I will review our
rationale
for that choice shortly.
Although the recent study
results with
rofecoxib and celecoxib were not available
when we
designed the studies that I will be
describing
shortly, our studies are extremely
relevant as they
compared etoricoxib to diclofenac and
thus address
the current clinical question of
comparative
172
cardiovascular safety between a selective
COX-2
inhibitor and a traditional NSAID.
In order to choose an
appropriate
comparator NSAID, we established criteria
and
evaluated numerous agents and ultimately determined
that diclofenac was the most suitable
choice.
Diclofenac is effective in
treating both
osteoarthritis and rheumatoid arthritis
patients
and can be dosed twice daily, which
enhances
compliance and convenience for the
patient.
Secondly, it has been
established that
diclofenac does not interfere with
low-dose
aspirin's anti-platelet effects. Ibuprofen, on the
other hand, does interfere with low-dose
aspirin's
anti-platelet effects.
This interaction posed two
issues we felt
precluded use of ibuprofen as the
comparator. We
were not comfortable enrolling patients
who
required low-dose aspirin with knowledge
that its
anti-platelet effects could, in fact, be
inhibited,
and secondly, we were concerned that
interpretation
of study results, which showed comparable
173
cardiovascular safety, to an agent that
inhibits
aspirin's anti-platelet effects could be
problematic.
Diclofenac inhibits both COX-1
and COX-2
and confers partial inhibition on
platelet-mediated
COX-1 thromboxane. Since it lacks potent and
sustained anti-platelet activity, we
would not
expect confounding effect on the
interpretation of
cardiovascular safety results as would be
expected
with naproxen based on the cardiovascular
data from
the development program which I
presented.
Data from some of our clinical
trials
indicate that diclofenac's effect on
blood pressure
is generally similar and, in fact, in
some cases
more pronounced than the effect observed
with
etoricoxib.
In consideration of the
established
cardiovascular complications of
elevations in blood
pressure, a comparison of thrombotic
cardiovascular
safety between etoricoxib and diclofenac
can, in
fact, be considered conservative.
I wanted to briefly review some
174
pharmacodynamic data which supports
diclofenac
having COX-1 inhibiting effects. Represented on
this slide are the ex-vivo COX-2 and
COX-1
inhibiting effects of various agents.
Displayed on the X axis is the
percentage
of COX-2 inhibition as measured by
inhibition of
lipopolysaccharide-induced serum PGE
2. Displayed
on the Y axis is the percentage of COX-1
inhibition
as measured by serum thromboxane as a
weighted
average at steady state with 84 percent
joint
confidence regions around the point
estimate of the
mean.
Rofecoxib at 12.5 and 25 mg
inhibits COX-2
on the order of 60 to 70 percent in this
experiment. Diclofenac at a dose of 150 mg
inhibits COX-2, but also inhibits COX-1.
Endoscopic data are also
available which
support the COX-1 inhibiting effects of
diclofenac.
Shown here are results from two endoscopy
studies
performed with valdecoxib which included
a
diclofenac treatment arm. In each case, the
cumulative incidence of gastroduodenal
ulcerations
175
observed at the end of the study period
are
displayed by treatment group in these two
studies.
On the left are results of a
26-week study
of rheumatoid arthritis patients. The incidence of
gastroduodenal ulcerations observed on
diclofenac
was significantly greater than observed
on either
dose of valdecoxib in this study.
On the right are results of a
12-week
study in osteoarthritis patients. The incidence of
gastroduodenal ulcerations on diclofenac
was
significantly greater than placebo and
valdecoxib,
and, in fact, similar to the incidence
observed on
ibuprofen.
Lastly, I would like to point
to some GI
clinical outcomes data which also support
the COX-1
inhibiting effects of diclofenac. Dr. Braunstein
reviewed the cumulative incidence of
confirmed
upper GI clinical events of rofecoxib
versus
individual NSAIDs yesterday based on
final data
from the rofecoxib development program.
What I have done here is
instead of
looking at confirmed PUBs, I have also
added the
176
confirmed plus unconfirmed results, which
are very
consistent with what Dr. Braunstein
showed
yesterday.
You see here the relative risk
of
confirmed plus unconfirmed upper GI
events observed
on rofecoxib is, in fact, significantly
different
than the effect observed with diclofenac,
so again
to provide some clinical data that
support a COX-1
inhibiting effect of diclofenac.
The overall approach to further
characterize etoricoxib that I have been
describing
consists of a prospectively designed
analysis of
cardiovascular safety data will accrue
from three
studies, which I am going to briefly
review here.
All three studies compared
etoricoxib to
diclofenac. The first is the EDGE study, a study
of 7,111 osteoarthritis patients with a
primary
objective to compare the GI tolerability
of
etoricoxib to diclofenac. This study is now
complete.
Secondly, EDGE II, a study of
approximately 4,090 RA patients with a
primary
177
objective identical to that of EDGE. The dose of
etoricoxib in EDGE II is 90 mg. This study is
fully enrolled and ongoing. The predicted mean
duration of this study is expected to be
approximately 19 months.
Thirdly, MEDAL, a study of
approximately
23,450 osteoarthritis and rheumatoid
arthritis
patients with the primary objective of comparing
the cardiovascular safety of etoricoxib
to
diclofenac. This is an endpoint-driven outcome
study.
MEDAL is fully enrolled and currently
ongoing.
The predicted mean duration of therapy in
MEDAL is approximately 20 months with
some patients
expected to be on therapy an excess of 3
years.
Although EDGE and EDGE II are
designed as
primary GI tolerability studies, the
cardiovascular
safety data that will accrue from those
two studies
are being adjudicated and will be
combined with the
cardiovascular safety data from the MEDAL
study in
order to improve the precision of the
comparison.
The primary hypothesis for this
analysis
is that etoricoxib will demonstrate a
178
cardiovascular safety profile that is
non-inferior
to that of diclofenac. There are 2 key analyses
that are designed to support this hypothesis.
The primary analysis will
consider the
minimum required 635 confirmed thrombotic
events
from all 3 studies combined, and the
secondary
analysis will consider the minimum 490
confirmed
thrombotic events that are required from
the MEDAL
study alone.
As I mentioned, MEDAL was
designed as an
endpoint-driven outcome study and on its
own
represents a sufficiently powered
assessment of
cardiovascular safety. The patient
population that
has been enrolled in these studies
consists of
patients with a range of baseline
cardiovascular
risk and includes patients with
pre-existing
cardiovascular disease.
As clinically indicated, such
patients, as
well as others, are being prescribed
aspirin, so we
expect the total study cohort to include
approximately 30 percent aspirin users.
MEDAL and EDGE II will generate
a
179
tremendous amount of long-term
cardiovascular
safety data. As summarized on the previous slide,
the predicted mean duration of therapies
in EDGE II
and MEDAL are 19 and 29 months
respectively, and it
is predicted that out of the 635
confirmed
thrombotic events, approximately 200 of
those
events will occur in patients who have
been on
study therapy for at least 18 months.
In this cohort alone, the minimum
between
treatment group difference that would be
statistically significant expressed as a
relative
risk is approximately 1.3.
An external Data and Safety
Monitoring
Board was chartered to monitor emerging
data from
MEDAL, EDGE, and EDGE II. Since 2002, they have
been meeting regularly, most recently in
November
of 2004, at which time they reviewed a
large amount
of data.
At that time, in total, there were
approximately 21,000 patient years of
exposure and
approximately 300 confirmed thrombotic
events were
available at that time for their review.
In addition, there were
approximately
180
3,000 patients who had been on study
therapy for at
least 18 months at that time. Based on their
review, their recommendation was to
continue the
ongoing studies without interruption or
without
modification.
Of the 3 studies that we have been
discussing, EDGE is the first to be
completed, and
I would now like to review the
cardiovascular
safety data from the EDGE study.
In this study, the 7,111
osteoarthritis
patients were on study therapy for a mean
duration
of approximately 9 months, resulting in
approximately 5,400 patient years of
total
exposure.
The study population included
patients
with a range of baseline cardiovascular
risk. Here
are summarized some selected baseline
characteristics. As you see, approximately 38
percent of the patients in this study
were at
increased baseline cardiovascular risk
defined as
patients having 2 or more risk factors
for
cardiovascular disease or a documented
history of
181
symptomatic atherosclerotic
cardiovascular disease.
This slide summarizes the
cardiovascular
safety data from the EDGE study by
presenting again
the point estimate of the relative risk
and the
corresponding 95 confidence interval, for
confirmed
thrombotic events versus diclofenac, for
events
which occurred on therapy or within 14 days of
study therapy discontinuation, on study
therapy or
within 28 days, and importantly, an all
patients
treated analysis.
In the EDGE study, all patients
who
discontinued were followed up closely
with regular
phone contact to ascertain any events
that occurred
long term off-of-study therapy, and this
was done
for all patients until all patients had
completed
the study.
The cumulative incidence of
confirmed
thrombotic events in the EDGE study are
summarized
here, and indicate no evidence of a
difference
between the treatment groups over time.
This slide summarizes the
specific
confirmed events by type in the EDGE
study
182
beginning with events which occurred on
study
therapy or within 14 days of
discontinuing study
therapy.
As you see, there are events
reported in
all 3 vascular events with more cardiac
event
overall irrespective of treatment group.
Evaluation of individual event types
indicates that
the absolute number of any event was
small with
numeric differences between treatment
groups for
certain events with some occurring at a
higher rate
on etoricoxib and some occurring at a
lower rate.
For example, differences were
observed in
ischemic strokes numerically favoring
etoricoxib,
however, differences favoring diclofenac
were
observed for acute myocardial
infarctions. Neither
of these differences were statistically
significant.
It is important to remember
that even in a
study of this size, results at the level
of
individual events should be interpreted
cautiously.
For example, when looking at events which
occurred
on study therapy or within 28 days, as
requested by
183
the agency, the numeric differences
between
treatment groups has, in fact, narrowed
slightly
due primarily to an increase in the
number of acute
myocardial infarctions which occurred on
the
diclofenac group.
Data from ongoing randomized
clinical
trials will be critical to more precisely
assess
the comparative rates of myocardial
infarctions on
diclofenac versus etoricoxib.
Summarizing results of the EDGE
cardiovascular safety data next to the
results of
the pooled analysis that I presented
previously
indicate that the EDGE data are, in fact,
consistent with, and add precision to,
the
observations from the pooled analysis
when
comparing etoricoxib to non-naproxen
NSAIDs.
I would now like to
summarize. We have
demonstrated efficacy with etoricoxib
that is
similar and in the cases I have pointed
out, in
fact, superior to comparator NSAIDs
particular
naproxen 1,000 mg.
We have a GI safety program
that did
184
demonstrate improved GI safety and
tolerability in
relation to shift to non-selective NSAIDs
primarily
in relationship to naproxen, and the
renovascular
effects observed with etoricoxib are, as
again
would be expected based on the mechanism
of action
dose related, but at the doses
recommended for
chronic use are, in fact, generally
similar to the
effects observed for the comparator
NSAIDs.
We saw numeric differences
against
naproxen favoring naproxen, but we also
saw rates
of
hypertension that were very similar to those
observed with ibuprofen even at their
maximal
chronic dose.
Based on thorough and ongoing
reviews of
cardiovascular safety data, there is no
clear or
discernible difference between etoricoxib
and
non-naproxen NSAIDs up to a year. As I said, we
have limited amounts of data beyond 1
year at this
time.
Differences were observed
between
etoricoxib and naproxen rates of thrombotic
events.
Based on the data we have, the limited
amounts of
185
short-term placebo data, there is no
clear
difference between etoricoxib and
placebo. That
being said, emerging data from long-term
placebo-controlled studies with rofecoxib
and
celecoxib showing a difference in
cardiovascular
safety versus placebo do, in fact,
suggest a class
effect.
MEDAL, the largest NSAID trial
known, and
EDGE II are currently ongoing and based
on current
cardiovascular event rates are expected
to be
completed next year. Results from these studies
will further characterize the
cardiovascular safety
of
etoricoxib, and we will have data to address
numerous questions including
cardiovascular safety
in both osteoarthritis and rheumatoid
arthritis
patients, and cardiovascular safety in
patients
with a range of cardiovascular risk, and
will
include experience in aspirin users and
non-users.
We will be able to further
explore the
effect of dose as both 60 and 90 mg are
included in
the study, and perhaps, most importantly,
the
long-term cardiovascular safety will be
assessed as
186
we will have large amounts of data in
patients who
have been on study therapy for at least
18 months.
These studies directly address
whether the
cardiovascular safety including the
long-term
safety of a selective COX-2 inhibitor,
such as
etoricoxib, is similar to or different
than that of
a traditional NSAID.
In countries where etoricoxib
is currently
approved, Merck has consistently taken a
proactive
approach with regulatory agencies. From the time
it was first approved years ago, the
etoricoxib
product label has, in fact, contained a
precaution
for use in patients with ischemic heart
disease.
We continue to work
aggressively with
regulatory agencies and are currently
actively
engaged with European regulators, and
have
participated in a referral process in
Europe. Our
goal there is to ensure that the product
label
accurately reflects all accruing safety
information
that is relevant to prescribers based on
data that
are currently available.
In conclusion, etoricoxib has a
role among
187
the current treatment options for
patients with
conditions characterized by pain and
inflammation.
However, it is critical to ensure its
safe and
effective use, that a product labeling
continues to
be revised to ensure that all currently
available
data are incorporated to help guide
appropriate
use.
We remain committed to help
address public
health questions and currently, with
etoricoxib,
largely through the conduct of the MEDAL
and the
EDGE II studies. These questions posed yesterday
include, For patients who require chronic
anti-inflammatory therapy for established
indications, what is the risk and benefit
of a
selective COX-2 inhibitor as compared to
an NSAID?
MEDAL and EDGE II will provide
information
to this question in comparison to
diclofenac, and I
have provided you the data we currently
have
available that provides information
relative to
naproxen.
Other questions which remain at
this time
include Can patients at increased
cardiovascular
188
risk be identified, so the benefit is
maintained
and the risk minimized?
MEDAL, again due to its
unparalleled size,
and with the additional data from EDGE
II, will
provide information and data to allow
further
exploration to help answer this question.
Next, Is the increased
cardiovascular risk
a class effect of COX-2 inhibition, and
if so, how
large is the class, and what are the
long-term
cardiovascular effects of a selective
COX-2
inhibitor and traditional NSAIDs?
Again, MEDAL, with its
long-term direct
comparison to diclofenac, will provide
information
to address both of these questions.
This concludes my presentation. I would
like to thank the Chairman, members of
the Advisory
Committee, the FDA.
Thank you.
DR. WOOD: Thanks a lot.
Let's go
straight on to the FDA's presentation.
FDA Presentation
Analysis of Cardiovascular
Thromboembolic
Events with Etoricoxib
Joel Schiffenbauer,
M.D.
DR. SCHIFFENBAUER: Thank you and good
189
morning.
My name is Joel Schiffenbauer. I
am
going to be presenting an analysis of
cardiovascular thromboembolic events with
etoricoxib.
I will be presenting the results of
trials
for the following indications listed here
in the
NDA.
In addition, I will be presenting results of
the EDGE trial separately from those of
the trials
here.
I will first present briefly
exposure data
followed by mortality data and then spend
the
remainder of the time discussing the
cardiovascular
thromboembolic events data. Again, I will present
data first for the NDA and separately for
the EDGE
study.
First, exposure. This slide summarizes
the chronic exposure to etoricoxib across
the NDA.
As you can see for the 60, 90, and 120 mg
doses,
which were the proposed doses for the
drug, the
190
total number of patients is shown here
and the mean
number of days is shown here.
For the EDGE study, there was
approximately 3,500 patients in each arm,
exposed
for a mean of 9 months. Total patient years is
shown at the bottom.
Let me turn now to the
mortality data.
This is the mortality data across the
NDA. Rates
are shown as per 100 patient years, and I
have
listed the comparators here, placebo,
non-naproxen
nonsteroidals, and naproxen.
If we first look at the first
line of
total deaths, we can see that the rate of
deaths in
the placebo group is similar to naproxen,
followed
by the non-naproxen nonsteroidals, and
then
etoricoxib.
Let me next draw your attention
to the
third line, thrombotic cardiovascular
deaths.
There were no deaths in the placebo
group, followed
by
naproxen, etoricoxib, and then non-naproxen
nonsteroidals.
These 2 events I would point
out occurred
191
at greater than 36 months exposure to the
non-naproxen nonsteroidals, and I will come
back to
this point when I present the
Kaplan-Meier analysis
looking at non-naproxen nonsteroidals.
The deaths in the EDGE study,
the total
deaths are similar, 8 and 6, for cardiovascular
thrombotic related, it was 3 and 1.
Let me now move on to a
discussion of the
cardiovascular thromboembolic events.
The sponsor proposed a
composite endpoint,
which you have already heard about, which
included
events related to the cardiac,
peripheral, and
cerebrovascular system. I will present results for
both the composite, as well as the
components of
the composite, and I think this is an
important
point because we do not yet know the
effects of
COX-2 inhibitors on each of these
specific
cardiovascular events.
In addition, I will not present
data for
APTC events or investigator-reported
events.
Although the numbers vary slightly, the
trends are
always in the same direction as the
events that I
192
will show here.
These events were referred to
an
Adjudication Committee, that you have
heard about
already, and after being reviewed in that
committee, were then described as
confirmed
cardiovascular thromboembolic events.
This slide shows an analysis of
the
confirmed thrombotic cardiovascular
serious adverse
events across the NDA. This is exclusive of the
EDGE study. The sponsor performed 3 comparisons -
etoricoxib to placebo, etoricoxib to
non-naproxen
nonsteroidals, and etoricoxib to naproxen.
The number of patients, the
cases in
patient years of exposure is shown here,
rates, and
relative risk. I will show this slide over again.
First, let me start on the
first line. I
draw your attention to the rate of events
in the
etoricoxib group 1.25 versus placebo 1.19
for the
relative risk shown here, and an analysis
of those
events is shown in this slide.
These are the rates I showed
you, 1.25 and
1.19. There were a total of 7 patients in
the
193
etoricoxib group versus 4 in placebo, and
this
breaks down to 4 cardiac events, which
are listed
here - MI, fatal MI, unstable angina, and
sudden
death versus zero in placebo.
The number of events in
peripheral and
cerebrovascular are similar although the
rates do
vary slightly.
Let me point out here that in
some of
these slides, these numbers will not necessarily
add up.
That is for two reasons. One is
an
individual patient may have more than one
event,
and they would therefore be listed in
more than one
category, and, secondly, for the sake of
clarity
and brevity, I left out in some instances
all the
events.
This is the Kaplan-Meier
estimate of time
to event for the placebo comparison. Note that
this is only 3 months in duration. There are very
little differences between the two
groups.
Let me move on then to the
etoricoxib/non-naproxen comparisons. Here is the
rate, 0.79 and 0.80, and I will show you
that in
194
next slide. Here are the rates again, 0.79 and
0.80.
These are composed of 12 patients in the
etoricoxib group versus 4 in the
combined, and by
that I mean combined exposure to
diclofenac and
ibuprofen.
You can see, however, exposure to
ibuprofen is rather small and there were
no events,
so all of the events come from the
diclofenac
exposure.
If we examine the breakdown of
these 12
events, you can see there were 11 cardiac
events in
the etoricoxib group for the rate shown
here versus
2 in the combined for this rate, and that
is
further broken down to 3 MIs versus zero,
2 and 1
of fatal MIs, and then the rest you can
see here.
There are 2 and 2 events in the
cerebrovascular
system.
You have seen this previously,
but let me
make several points about this
Kaplan-Meier
analysis for the non-naproxen and
nonsteroidal
comparisons. First of all, you will note that the
length of exposure is out to 36 months
when there
are relatively few patients still present
in the
195
studies.
Secondly, there were 4 events in the
non-naproxen nonsteroidals, which is
shown by the
solid line. Three of those events occurred at
greater than 36 months exposure. Two of
those 3
events were the deaths that I described
in the
earlier slide.
In contrast, there were 12
events in the
etoricoxib group, 11 out of those 12
events
occurred at approximately 26 months or
earlier.
So, there is a difference in the time to
event as
demonstrated by this Kaplan-Meier
analysis.
Lastly, let me turn to the
etoricoxib/naproxen exposure. Here are the rates,
1.37 and 0.81. Again, here are the rates, 1.37 and
0.81.
There were 34 patients in etoricoxib versus
14 in naproxen, and that is broken down
into 21
cardiac versus 9 for the rate shown here,
10 MIs
versus 5, and you can see the remainder.
For peripheral, there was a
slight
imbalance, 5 events in naproxen versus 2
in
peripheral, however, when we come back to
the
196
cerebrovascular system, there were 12
versus 2,
which included 10 ischemic strokes versus
zero.
Again, you have seen the Kaplan-Meier
analysis,
which shows a separation of the two
curves almost
throughout the entire exposure.
Let me turn now to the analysis
of
cardiovascular events in the EDGE study,
and start
by making a few points. There were 7,100
patients.
It was designed as a GI tolerability
study in which
cardiovascular data was collected.
The sponsor defined a
non-inferiority
margin to diclofenac for cardiovascular
events as
the upper limit of the 95 percent
confidence
interval for the hazard ratio of 1.3.
In addition, there were several
concerns
that I would like to emphasize. First, it was
designed as a non-inferiority trial,
there was no
placebo.
Diclofenac was the only comparator, and
as we have heard here, and there is data
in the
literature to support the relative COX-2
selectivity of diclofenac.
Next, there were only osteoarthritis
197
patients studied. There were no rheumatoid
arthritis patients in this study. We know that
rheumatoid arthritis itself confers
cardiovascular
risk.
The next two bullets relate to
maneuvers
that could potentially, in the context of
a
non-inferiority trial, make it difficult
to
identify differences between the two
treatment
groups.
So, for example, there was 30
percent
aspirin use. If we believe that aspirin
is
cardio-protective even in the context of
COX-2
inhibitor, this could make it difficult
to discern
any differences between the two groups.
In addition, previous COX-2 use
was
allowed, and I have listed here what that
was, and
this could potentially lead to depletion
of
susceptible individuals to a
cardiovascular event.
Lastly, although it is
important to study
high-risk patients, if these high-risk
patients are
on aspirin, that may be a problem in
differentiating the two groups. In addition, if
198
there are more events in these high-risk
patients,
it could increase the background events,
and again
in the context of a non-inferiority
trial, may make
it difficult to differentiate the two
treatment
groups.
So, you have seen this
Kaplan-Meier
analysis. Again, the two groups separate
slightly,
but the two curves do finally converge at
approximately 12 months.
This is a breakdown of the
events in the
EDGE trial. There were 35 patients in the
etoricoxib group versus 30 in diclofenac
for the
rates given here. If we look at a further
breakdown of the components, we see there
were 27
cardiac-related events versus 19 for the
rates
given here. For MI, there was 19 versus 11. For
cerebrovascular events, there was 7 and 7
with a
slight imbalance in ischemic strokes of 6
in
diclofenac versus 3 in etoricoxib.
I think it is important, I
mentioned
earlier that aspirin use may be a
problem. I broke
down the number of events by aspirin and
199
non-aspirin users, and I have just
provided the
number of events, the patient years of exposure
are
fairly similar.
You can see that by aspirin
users, there
is little differences between the groups,
12 versus
9 here for cardiac events, 7 and 5. However, when
you look at the non-aspirin users, the differences
are more pronounced. There were 15 cardiac events
in etoricoxib versus 10 in diclofenac,
and 12 MIs
versus 6.
There was some concern about
hypertension.
Some issues were raised about that
yesterday. I
show some data for hypertension-related
adverse
events in the EDGE trial. These types of
adverse
events could include anything from a
hypertensive
crisis, malignant hypertension to
systolic blood
pressure increase among other events.
This is an analysis of patients
with
serious hypertension-related adverse
events. There
were 5 in etoricoxib versus 2 in
diclofenac, and
then another category,
hypertension-related AE
associated with systolic blood pressure
greater
200
than 180, or diastolic greater than 110,
and there
were 69 cases here versus 30 in
diclofenac.
Then, this is a cumulative
incidence of
new use of anti-hypertensive
medications. The
upper line is etoricoxib, the lower line
is
diclofenac. You can see that the two curves
separate almost throughout the entire
12-month
period.
Lastly, a description of congestive heart
failure-related adverse events. This is the
incidence of CHF pulmonary edema-related
or cardiac
failure adverse events. There were 14
versus 6.
In summary, in the NDA, etoricoxib
trends
worse in terms of cardiovascular
thromboembolic
events, particularly cardiac and MI. The one
common thread throughout all the
comparators does
appear to be the cardiac system.
There are differences in the
cerebrovascular or peripheral system, but
those are
inconsistent depending on the comparator.
Comparisons of etoricoxib to
naproxen for
the cardiovascular events is similar to
what you
201
have seen for rofecoxib and the naproxen
comparisons.
I have outlined some trial
design concerns
in the EDGE study, which I presented, and
as you
have already heard, there are two ongoing
trials of
similar design, which I believe have
similar
concerns.
There are trends in the EDGE
study for
cardiac events, worse for etoricoxib, and
that is
seen mainly in the non-aspirin users.
Thank you.
DR. WOOD: Thanks very much.
Let's go straight on to the
Novartis talk
and we recognize that will finish a
little late,
but we will have a shorter lunch break.
Lumiracoxib
Lumiracoxib:
Introduction
Novartis Pharmaceuticals
Corporation
Sponsor Presentation
Mathias Hukkelhoven,
Ph.D.
DR. HUKKELHOVEN: Thank you.
Dr. Wood, Dr. Gibofsky, Dr.
Gross, members
202
of the FDA Advisory Committees, FDA, and
Guests:
Good morning. My name is Mathias Hukkelhoven and I
am responsible for Global Regulatory
Affairs at
Novartis.
On behalf of Novartis, I would
like to
thank you for the opportunity to review
the
gastrointestinal and cardiovascular
safety data
that we have gathered in our clinical
development
program for lumiracoxib.
As a part of the program, we
have also
gathered one of the largest databases of
clinical
trial data with ibuprofen and naproxen.
Allow me to remind you of the reason that
the COX-2 selective NSAIDs were
developed. In the
U.S. alone, there are approximately
100,000
hospitalizations, and as we heard
yesterday, 16,000
deaths annually that are caused by GI
adverse
events.
Deaths due to NSAIDs are among the leading
causes of death in the U.S.
Our presentation will make the
following
key points. Each non-selective NSAID and COX-2
selective inhibitor has a benefit-risk
profile that
203
must be considered individually.
The Novartis development
program provides
clinically informative safety data for
lumiracoxib
as well as for ibuprofen and naproxen.
The GI and CV safety profile
for
lumiracoxib differs from non-selective
NSAIDs and
other COX-2 selective inhibitors.
We have investigated the use of
lumiracoxib for several indications, but
our
presentation today will focus on the
safety data
accumulated for chronic indications. We have
conducted 22 clinical trials of 1 week or
longer in
which 34,000 patients were enrolled.
The largest of the clinical studies was
the TARGET outcome study. This is the largest
outcome study ever conducted for an NSAID
or COX-2
selective inhibitor with 18,325 patients
enrolled.
It is important to note that this study
was
conducted at 400 mg daily dosing, which
is 4 times
the dose for which approval will be
sought.
This 1-year study compared
lumiracoxib to
two different NSAIDs - naproxen and
ibuprofen.
We will also present a
meta-analysis of
cardiovascular safety of all 22 long-term
lumiracoxib studies.
204
Our presentation will
demonstrate that
there is a definitive GI benefit with
lumiracoxib
in the non-aspirin population. In addition, the CV
meta-analysis of all lumiracoxib studies
at no
point revealed a significant CV risk.
I would like to introduce
today's
presenter Dr. Patrice Matchaba from our
Clinical
Research Department. In addition, we have a few
advisers with us who will be able to
answer
specific questions. These are Dr. Michael Farkouh,
a cardiologist from NYU; Dr. Raymond Hirschberg,
a
nephrologist from UCLA; and Dr. Thomas
Schnitzer, a
rheumatologist from Northwestern.
Drs. Farkouh and Schnitzer were
the lead
authors on the TARGET CV and GI
publications that
were published this past September in the
Lancet.
I would now like to turn the
podium to Dr.
Patrice Matchaba.
Gastrointestinal and
Cardiovascular Safety
of Lumiracoxib, Ibuprofen, and
Naproxen
Patrice Matchaba, M.D.
DR. MATCHABA: Thank you to the Chair,
thank you to the Committee, the FDA, and
the public
for inviting us. Just to state that for this
purpose, we will be discussing the
cardiovascular
205
and GI safety data from the TARGET study,
but we
are certainly willing to answer any
question
related to an end organ, and that we have
published
the data in TARGET in the Lancet, two
papers.
We have two of the key primary
authors for
the GI and other adverse events of safety
profile,
we have Dr. Schnitzer, and for the
cardiovascular
paper we have Dr. Michael Farkouh.
Before we got into the TARGET data for
cardiovascular and CV, I think it is
important
underlie that when the TARGET study was
designed,
that the VIGOR study and the CLASS study
had been
completed, and that the discussion for the
TARGET
design occurred between health
authorities
including advice sought from the
Arthritis Advisory
Committee in September 2001, because
there were
206
important public health questions that
were asked
after the CLASS study and the VIGOR
study.
Some of the key issues or
principles that
then drove the design of the study, the
first point
was that we should designing studies to
detect a
difference in ulcer complications because
that was
the COX-2 promise.
As a result, more patients were
required
because this event, as Dr. Cryer had
discussed
yesterday, is a fairly rare event, about
1 percent
of patients, that the patient numbers
required
increased to about 18,000 patients in
TARGET.
The second point was that in
this
population of patients that we studied in
terms of
osteoarthritis, that they do take
low-dose aspirin,
so we stratified patients to low-dose
aspirin, and
we managed to get a 24 percent
stratification, and
obviously, because of the impact of
low-dose
aspirin on GI outcomes, this necessitated
an
increase in the size of the study.
The other point that had been
made from
the previous two studies, that the median
duration
207
was short. If you recall, the VIGOR study had a
median duration of about 9 months, and
the data
that we saw for CLASS was 6 months,
whereas, in
TARGET, we had a fixed term design of 12
months.
The other design principle that
was
important, and we have heard this data
discussed
extensively, was that not all NSAIDs are
the same
in terms of COX-1 and COX-2 activity and
that we
will see differential GI and CV effects
because of
that.
So, we chose two NSAIDs that
should have a
different impact on the GI and the CV,
and
addressed that question as to what is the
difference between coxibs, and in this
case,
lumiracoxib between naproxen and
ibuprofen.
Finally, there was a need to
prospectively
define an adjudicate all outcomes, so we
had three
Adjudication Committees, one for the
cardiovascular
outcome, the other one for the CV, and
the other
one for the hepatic events.
In terms of the objective, it was to
compare lumiracoxib at 4 times the
proposed OA
208
dose, 400 mg. to naproxen 500 mg bid, and
the dose
is important here, because this is the
dose and the
dosing frequency that people have
discussed in
terms of a possible anti-thrombotic
effect, and
ibuprofen at 800 mg 3 times a day.
Key inclusion criteria that I
think are
important for the endpoints is that
patients who
had a previous history of a
cerebrovascular or
ischemic event in terms of cardiac events
were
allowed into the study if the event
occurred more
than 6 months before they entered the
study and if
they had been on low-dose aspirin for 3
months in
order to stabilize the patients, and this
is the
advice and the current thinking that was
there in
terms of patient safety if you are going
to conduct
a 1-year study.
From a GI perspective, a key
exclusion
criteria was that any patients who had
active GI
ulcerations 30 days previously were
excluded, and
any patients who had a GI bleed in the
previous
year were excluded because the thinking
again was
that with the availability of PPIs and
high-dose H2
209
antagonists, that enrolling these
patients that
required long-term treatment would have
been an
unethical thing to do.
So, the study design were 2
studies that
were identical, lumiracoxib compared to
naproxen in
1 study, and compared to ibuprofen in the
other
study.
You will note that the 2 studies are of
similar size, about 9,000 patients in
each study,
and that the studies went on to 52 weeks
or 1 year
with a follow-up at 56 weeks or at 1
month.
The key thing to note also is
that the
naproxen sub-study started recruitment 4
to 5
months before the ibuprofen sub-study,
and that
different centers were used for the 2
studies. So,
you may see differences in the baseline
risk for
the endpoints that we will be discussing.
For cardiovascular and for this particular
discussion, as I said, we had a
pre-defined and
prospectively adjudicated CV endpoints
that
included important coronary
cerebrovascular and
also the peripheral events.
In terms of the patient demographics, the
210
majority were female an average age of
63. We
managed the 24 percent aspirin
stratification. Of
importance is that within this cohort of
18,325
patients, about 12 percent of these
patients had a
high CV risk as defined by a previous
cerebrovascular or cardiac history or by
Framingham
risk equations.
The patients were fairly representative
of
an OA population. We had hypertensive patients,
diabetics, and patients with
dyslipidemia. Very
importantly, because it was fixed term
design, 60
percent of the patients finished the 12
months, a
total of about 11,000 taking treatment
for 12
months.
For the primary endpoint, which
was ulcer
complications, or perforations,
obstruction, and
bleeds in the non-aspirin population, it
was a
relative risk of 0.21 or a 79 percent
reduction if
you compared lumiracoxib to all NSAIDs.
If you made that comparison by
sub-study,
it was an 83 percent reduction compared
to
ibuprofen, and 76 percent reduction
compared to
211
naproxen.
So, although we have 2
different NSAIDs
with different COX-1 and COX-2
activities, this the
first GI outcome study that looked at
ulcer
complications as a primary endpoint and
shows
definitively a reduction in ulcer
complications for
lumiracoxib compared to the NSAID
studied.
A lot of the discussion,
because if you
recall, we have stratified patients to
low-dose
aspirin in the 24 percent of patients, about
4,000,
and the question was, and is, what is the
impact of
low-dose aspirin on this outcome, and we
will also
discuss the CV outcome.
Now, if you look at the ulcer
complications in the low-dose aspirin
population,
and I have tried to show you an analytic
figure
here, is that for the upper GI ulcer
complications,
there was a relative risk of 0.79 with
wide
confidence intervals crossing the line of
no
difference with a point estimate showing
a 21
percent reduction.
What we have done, however, for
this
212
discussion is to say when we consider
more events,
ulcer complications and symptomatic
ulcers, is the
point estimate still favoring lumiracoxib
and does
the confidence interval tighten in terms
of the
precision, and you can see that the
reduction
increases to 27 percent, but the
confidence
interval certainly still crosses one.
In this context, it is
important to
remember that the TARGET study wasn't
designed to
show a difference in the low-dose aspirin
population, but was designed to show a
difference
in the non-aspirin population, and it
cascaded to
the overall population if the first
result was
positive.
But what is encouraging is the
consistent
trend that we see in this population.
There was discussion yesterday
about do
coxibs, in this case lumiracoxib, does it
still
show benefit in patients who have a high
GI risk,
and this was prespecified in the TARGET
analysis,
and high GI risk, there were 5 categories
of risk
defined, age greater than 65, low-dose
aspirin use,
213
a history of ulcers or bleeds in H.
pylori-positive
patients.
When we do the analysis, taking one
risk
into consideration, you see that the
magnitude of
about a 3-fold reduction in favor of
lumiracoxib
for ulcer complications is
maintained. If we have
time later on, we can also show you the
data for
patients greater than 65, for patients
who were H.
pylori-positive, but because of the
exclusion
criteria that I outlined beforehand for
patients
who had a previous bleed, the numbers
become
smaller and smaller when we look at
further
increasing risk for these patients.
So, in summary, for the GI
data, the
TARGET study definitively shows benefit
for
patients taking lumiracoxib compared to
these 2
different NSAIDs, ulcer complications in
the
non-aspirin population. We have seen the high risk
or high GI population as defined in
TARGET, and we
see a consistent trend although it is not
significant because of the numbers in the
patients
taking low-dose aspirin.
The cardiovascular endpoint
that was
chosen at that time was the APTC
endpoint.
Certainly, all the other cardiovascular
events were
214
also adjudicated, peripheral events,
pulmonary
embolism, deep vein thrombosis.
At the time we published the
data, and as
prespecified in the protocol, the plan
was to
compare lumiracoxib to all NSAIDs for the
APTC
endpoint, but for the purpose of this
discussion,
if we do that, we fail to disaggregate
the relative
results for lumiracoxib compared to
naproxen and
compared to ibuprofen.
So, we will discuss the
separate studies,
but, first, you will see that when you
compare
lumiracoxib and NSAIDs, that there is no
difference
in the APTC endpoint throughout the
12-month
period, but there is greater data and
more insight
to be mined when you look at the 2
studies
separately. That is the current debate.
Before we look at the data,
look at the
baseline demographics. If you recall, these were 2
parallel studies, recruiting at different
centers,
215
different time points, identical in
design, but
what you can see is that for the endpoint
that may
have an impact on the rate of
cardiovascular events
in the 2 studies, that there seems to be
differences in the low-dose aspirin use,
for the
naproxen sub-study, patients who were
high CV risk
and patients with baseline hypertension.
For the high CV risk patients,
in terms of
patient numbers, this translates to about
140
patients difference. Now, this may or may
not be a
factor in terms of looking at the
differences in
rates, and there are other factors
certainly that
we may not have measured that could
impact on the
differences in rates.
So, we will look at the
ibuprofen
sub-study first and look at the APTC
endpoints,
myocardial infarcts, look at stroke, look
at the
cardiorenal complications, congestive
heart
failure, and a combined endpoint, and
just to state
that in terms of all-cause mortality,
there were 29
patients who passed away in the
lumiracoxib study
arm and 30 patients in the NSAIDs, and
when you
216
split it up between the 2, there was
essentially no
difference.
So, for the APTC endpoint,
looking at
lumiracoxib versus ibuprofen, if you
start off with
the
overall result, in other words, all patients
including those who took low-dose
aspirin, you can
see that the hazard ratio for all
populations
studied are consistently less than 1.
The other point that I want you
to see is
that in the non-aspirin population, the
number of
events are the same with a hazard ratio
of 0.94.
There is certainly a lot of discussion
and this was
thought to be part of the value of
looking at the
TARGET data to ask what happens in the
low-dose
aspirin population where you have this
possible
interaction with ibuprofen.
You see in this population that
there were
6 events in lumiracoxib and 10 in
ibuprofen. This
difference, however, was not significant,
and you
will see when we look at myocardial
infarct, that
the number of events in this population
when you
look down to myocardial infarcts, are not
enough to
217
definitively contribute to this debate
about the
interaction of low-dose aspirin and
ibuprofen, but
certainly all the data in this 8,600
patients
studied do not indicate that lumiracoxib
is any
different from ibuprofen in terms of the APTC
endpoint or cardiovascular risk.
For myocardial infarcts, going
through the
same analysis, the overall population, 5
versus 7,
again, you see the hazard ratio of consistently
less than 1. The number of events are low. In the
non-aspirin population, 4 versus 5, and
as I
pointed out, in this aspirin population,
1 versus
2, so difficult to comment and to
contribute to the
debate about myocardial infarct and
ibuprofen
interaction.
For stroke, again the number of
events
were low, 8 versus 9, no real difference,
6 versus
5, and 2 versus 4 in the aspirin
population,
lumiracoxib 2 events, and ibuprofen 4
events. So,
again from this data, 8,600 patients
treated for 1
year, no indication that lumiracoxib is
any
different from ibuprofen in this robust
data set.
I think it is important to
recall that
this study, in terms of patient exposure
and
patient numbers, is larger than the VIGOR
or the
218
CLASS study in itself in terms of
exposure.
The real differences we see in
the TARGET
study is in hypertension, and there has
been a lot
of debate yesterday about the possible
impact of
hypertension as a risk factor in
contributing to an
increase in strokes and myocardial
infarct, and
cardiovascular morbidity.
If we look at the cumulative
incidence of
new onset hypertension or de novo
hypertension, you
can see that over the study period, 360
days, that
the patients taking ibuprofen have a
significantly
higher incidence of new onset
hypertension compared
to the patients taking lumiracoxib. This
is
percentages, number of patients.
So, it is about 10 percent of
patients
with new onset hypertension with about 6
percent of
patients.
For a similar analysis looking
at
aggravated hypertension, if you recall in
our
219
demographic analysis, about 45 percent of
the
patients in the TARGET study were hypertensive. In
terms of aggravation or worsening of the
hypertension, you see exactly the same
trend
between lumiracoxib and ibuprofen, which
was
significant.
If we look at the mean
difference over the
entire study period, again comparing
lumiracoxib
and ibuprofen for blood pressure, we see
a systolic
of 2.7 for patients taking ibuprofen
compared to
0.7, and we see almost a 1 millimeter
increase in
blood pressure for patients taking
ibuprofen with a
zero mean increase for patients on
lumiracoxib, and
these differences again are statistically
significant.
There was a lot of debate
yesterday as to
the
possible cardiorenal implications of this in
terms of edema, congestive heart failure,
and
weight gain, and if you look at the data
in TARGET
for this sub-study in terms of edema, no
significant differences between the comparators,
but for edema and congestive heart
failure, you see
220
that there are more patients taking
lumiracoxib
with edema, congestive heart failure, but
no
difference for weight gain.
There was discussion previously
about how
do we assess benefit-risk. There was discussion
also yesterday that any advantage that
was shown in
terms of GI ulcer complication reduction
with
rofecoxib in VIGOR was negated by an
increase in CV
events.
We prespecified, and this is
not a
validated way of analyzing benefit or
risk, but at
least we prespecified this outcome to say
if we
combine ulcer complications as defined by
perforation, obstruction, and bleeds, and
combine
them with the primary cardiovascular
endpoint, of
the APTC endpoint, what is the trend
compared to
the lumiracoxib and ibuprofen, and this
is the
endpoint that I am showing you for the
non-aspirin
population, that patients taking
ibuprofen are
significantly worse for this combination
of the 2
endpoints of GI ulcer complications and
APTC.
Certainly, this is the first
time that
221
this has been done in an outcome study in
arthritis, but we hope that this will
contribute to
the discussion in terms of getting an
overall
assessment for benefit for the patients
with
osteoarthritis.
If you look at the overall
population,
this difference is still significant with
a 50
percent reduction, but if you look at the
aspirin
population alone, the significance
disappears as
would be expected.
So, in summary, in this patient
population
of more than 8,500 treated and randomized
to
treatment for 1 year with these doses of
lumiracoxib and ibuprofen, if we look at
the APTC
endpoints, myocardial infarcts, and
stroke, the
hazard ratios are consistently less than
1.
We see significant differences
in
hypertension, and obviously, hypertension
in the
long term, as discussed yesterday and
today, may be
an impact on CV adverse events for
patients with
osteoarthritis.
We have also seen that there
isn't an
222
increase compared to ibuprofen for
congestive heart
failure and for edema, and as for the
combined
safety endpoint, there is a significant
benefit for
patients taking lumiracoxib.
So, we will now look at the
naproxen
sub-study and go through the same
analysis, APTC
endpoint, myocardial infarct, stroke,
cardiorenal,
the combined endpoint.
What you see immediately is
that for this
sub-study, that the number of events is
much
greater than the ibuprofen
sub-study. Also, what
you see is that the hazard ratios are now
in favor
of naproxen, and there are more events
with the
lumiracoxib compared to naproxen.
You will see when we look at
the next
slide, and we look at myocardial
infarcts, you will
see that this is driven by the
differences in
myocardial infarcts particularly in the
non-aspirin
population.
So, if we look at the
non-aspirin
population, patients taking lumiracoxib,
10
myocardial infarcts, clinical and silent,
compared
223
to 4 in the naproxen population, a hazard
ratio of
2.37, but which is not significant over
the
12-month treatment period.
But the robustness and I think
the value
that TARGET adds to the debate is that
because we
stratified 24 percent of the population
to low-dose
aspirin, when you look at the aspirin
population,
you see the numeric difference or the
hazard ratio
decreases in this population.
Now, low-dose aspirin we all
agree has
COX-1 activity, irreversibly binds to the
platelet,
and it may contribute to 10 to 30
reduction in
myocardial infarcts.
The question then was asked
when we look
at this data, and this is the data that
we present
to you, is that if it's COX-1 activity of
low-dose
aspirin that is negating the differences
in terms
of myocardial infarct, the implication
therefore
that naproxen at 500 mg dose taken twice
daily in a
clinical trial situation to ensure
compliance, and
there is certainly pharmacological data
that shows
that this dose has got anti-thrombotic
and platelet
224
aggregation activity, that naproxen must
have
significant COX-1 activity.
There has been extensive debate
this
morning about observational studies, the
merits of
them, and looking at the naproxen and
non-naproxen
data, but this paper published by June in
Lancet
last year, looking at all the studies,
observational studies, and this is not
the
rofecoxib analysis, but just the
observation
studies looking at naproxen.
We can see that when you
combine all the
data, that the diamond at the end here
shows a 14
percent reduction in myocardial infarcts
with a
confidence interval that doesn't cross
the line of
no difference or 1.
The point I think was made by a
member of
the panel that in observational studies,
that the
dose that is taken could be less than the
500 mg
dose, and that the dosing interval would
not be the
regular dosing interval that you see in
clinical
trial situation.
I think Dr. Graham also made
the point
225
that in that case, you would see the
point
estimates moving closer to 1 in terms of
the real
effect that you would see if it had
anti-thrombotic
activity.
So, you go back to the
sub-study of
naproxen and look at strokes, you see
that in the
non-aspirin population, small numbers,
and the same
thing in the aspirin population, so no
significant
differences, and the confidence intervals
are
crossing 1.
Now, again when we do the
analysis for
blood pressure, we see that there is
significant
difference in favor of lumiracoxib
compared to
naproxen.
Now, if you recall in the VIGOR study,
where they compared rofecoxib to
naproxen, that the
differences in blood pressure were the
reverse, and
that rofecoxib increased systolic and
diastolic
blood pressures, systolic by about 3 to 4
millimeters of mercury, and diastolic for
this same
comparator.
The caveats are there that
these are
different patient populations. The RA population
226
is a high-risk population, this is an OA
population, but without the studies that
compared
directly COX-2s, this is the only way
that we can
make a cross-study comparison.
Again, hypertension may be
significant as
discussed in terms of long-term morbidity
and
mortality.
For the same analysis we did
with the
ibuprofen sub-study for de novo
hypertension, and
for new aggravated hypertension, no
significant
difference between lumiracoxib and
naproxen
although consistently, the lumiracoxib
patients
have less events over the 12 months.
This is again a revealing
analysis if we
look at the cardiorenal
complications. For edema,
slightly more patients having edema, 4.5
versus 4.2
percent, but we think what is encouraging
is the no
increase compared to naproxen for
congestive heart
failure.
Again, we saw in VIGOR, or if
you look at
the VIGOR data, that rofecoxib had more
patients
with congestive heart failure or
pulmonary edema
227
compared to the same comparator, and we
have not
seen this is the naproxen sub-study, and
weight
gain, 8.1 percent versus 9 in favor of
lumiracoxib.
For the same analysis we did
for ibuprofen
looking at this safety endpoint that we
introduced
and prespecified in TARGET, for ulcer
complications
and APTC in the non-aspirin population,
again we
see over time that notwithstanding the
reduction
that you get with myocardial infarcts
with naproxen
or when you add the 2 combined, that over
time for
patients with osteoarthritis, at the
doses that we
tested, that there is a significant
reduction and
benefit for patients taking lumiracoxib
in the
yellow line there.
So, in summary, these two
studies, huge
studies, 8- to 9,000 patients, randomized
to 1
year, show interesting data, and the
naproxen
sub-study shows no significant increase
compared to
naproxen/lumiracoxib for the APTC
endpoint, but we
see these differences in myocardial
infarcts with
more events in lumiracoxib, but of key
importance
that when you consider the low-dose
aspirin
228
population and you add COX-1 activity,
that the
numeric difference disappears.
From a public health
perspective, still
significant differences in blood
pressure, no
increase in cardiorenal or congestive
heart failure
with lumiracoxib as we saw in the other
study, and
the combined safety endpoint still
significantly
favor lumiracoxib.
Now, because the study included
a certain
number of high-risk CV patients, it
allows us to
look at a high-risk cohort within the
TARGET study
and follow them over the 12 months, and
asked in
this sensitive high-risk cohort of
patients, what
are the outcomes in terms of APTC and
myocardial
infarct, and we will discuss only the
myocardial
infarct for this high risk.
A total of over 2,200 patients,
and these
are patients who had a history of either
coronary
artery disease, a previous myocardial
infarct, and
other vascular events, and we added these
patients
to those who had a high Framingham, high
risk, so
over 2,200 patients treated for 1 year.
We look at the myocardial
infarct data
because we will probably glean more from
looking at
this specific endpoint than looking at
APTC, but if
229
you have questions, we will address those
questions.
But if you look at the overall
population,
these 2,200 patients, including those who
were not
taking aspirin and those who were taking
aspirin,
the naproxen sub-study, there were 7
myocardial
infarcts in the lumiracoxib population
compared to
5.
Obviously, the number of events low,
nonsignificant, and if you look at the
ibuprofen
sub-study, 1 in the lumiracoxib and 2 in
the
ibuprofen sub-study.
The question, and certainly
there has been
debate that by adding low-dose aspirin,
which I
think everybody thought it was a good
idea in the
year 2001 in terms of answering some of
these
biological questions on the impact of
low-dose
aspirin--
DR. WOOD: Hang on.
You are getting well
over time, so you can try and speed it up
a bit.
230
Thanks.
DR. MATCHABA: Thank you.
But this population is an
important
population because these 646 patients
have a high
CV risk, but are not taking low-dose
aspirin and
are treated over 1 year. So, high CV risk and not
on low-dose aspirin, and what you see is
that in
the naproxen sub-study, 2 versus zero,
and 1 versus
1.
The last cohort are patients
who had a
previous myocardial infarct, randomized
to
treatment for 1 year, and there were 288
patients
who had a previous myocardial infarct,
and if you
look at the repeat APTC events, for the
naproxen
patients, 6 events occurred versus 3 for
the
lumiracoxib, and certainly this is
chance, because
the number of events are low and the
patient
population is small.
But what we can comment is that
we are not
seeing an outstanding signal even in this
high-risk
population with all the limitations of
the size of
the analysis. So, that is the TARGET data.
Finally, we performed obviously
a
meta-analysis of all studies completed on
the 30th
of December last year. Math has already described
231
there were 22 of those studies. You can
see from
the analysis that 34,000 patients plus,
18,000
patient year exposure, that patients who
were
randomized to 1-year studies accounted
for almost
90 percent analysis, so it's a fairly
robust
analysis.
If you look at the APTC
endpoint, and
notwithstanding all the discussion and
comment that
has come forth including from Dr.
FitzGerald, that
combining all comparisons is probably not
the right
thing to do, we did a comparison against
all
comparators.
Now, this is a cumulative
meta-analysis
and I will just quickly run through
it. These are
the studies that we have done from 2001
to 2004.
These are the cumulative
patients you can
see as we have added a trial, over
34,000. These
are the events as events have occurred
for APTC,
156, and we have added the events to try
and get an
232
estimate as they have occurred, and you
can see
that the relative risk of 1.2 with a
confidence
interval crossing 1. This is against all
comparators for the APTC.
We do the same analysis and we
subtract
naproxen, and when you do the same
analysis without
naproxen, you see the relative risk
changes to
0.94, over 24,000 patients with cumulative
event of
0.88.
What you also see is that at no time in our
development program have we seen a
significant
increase in risk.
If we look at myocardial
infarct, same
analysis against all controls, a relative
risk of
1.28 crossing the line of no
difference. A similar
analysis minus naproxen again, and you
see the
relative risk goes to 1, 24,000 patients
and 34
events.
For strokes, all controls
comparison, a
relative risk of 1.02, 62 events, and when
we
remove naproxen for the analysis, a
relative risk
of 0.84.
Now, this reduction that we are
seeing in
233
the
more robust data set with the meta-analysis is
certainly within the bounds of the 10 to
30 percent
benefit that you would expect from
aspirin in the
idea situation.
A specific question and the
comparison has
been made for the 2 studies, the naproxen
sub-study
versus the VIGOR, but just to point out,
and we can
have discussion if time permits, that the
half-lives are different of this compound
and the
structure. Lumiracoxib has got a short half-life,
and if the hypothesis that continuous
prostacyclin
inhibitor is important, this may be an
important
factor.
A median 9-month versus 12
months, seeing
a significant difference with the caveats
of the
different populations, but not seeing it
in a
similar population not taking low-dose
aspirin, and
we have commented about the differences
in the
congestive heart failure and the
hypertension,
which we think plays a significant role
with time.
The final slide I think has
been discussed
before in terms of prostacyclin, and if
the Chair
234
and the committee decides we can discuss
that more
in detail, but the fact that the other
NSAIDs also
show a prostacyclin inhibition compared
to the
COX-2s.
So, in summary, we have seen
that the
meta-analysis is supportive of the data
that we are
seeing in TARGET. It's a robust meta-analysis,
34,000 patient. We are seeing that each time you
removed naproxen from the comparison, you
are
getting your 10 to 30 percent difference
and that
at no time point during our development
program
have we seen a significant increase for
the APTC
endpoint.
Importantly, we are seeing no
increase
with lumiracoxib with congestive heart
failure and
hypertension.
The question was asked, and
this obviously
is the subject to further debate as to
what do we
think as a company going forward.
Thank you, Mr. Chair, and thank
you,
committee.
DR. WOOD: Great.
Thanks very much.
We are going to break for lunch
and I have
to remind the members to turn in their
dinner
reservation form I guess to Kimberly, and
we have a
235
table reserved for the committee members in
the
restaurant. We will be back here and start at 1
o'clock, so you had better grab it and
eat.
(Lunch recess.)
236
A F T E R N O O N P R O C E E D I N G S
(1:04 p.m.)
Open Public Hearing
DR. WOOD: Let me begin by reading the
conflict of interest statement.
Both the Food and Drug
Administration and
the public believe in a transparent
process for
information gathering and
decisionmaking. To
ensure such transparency at the open
public hearing
session of the Advisory Committee
meeting, FDA
believes that it is important to
understand the
context of an individual's presentation.
For this reason, the FDA
encourages you,
the open public hearing speaker, at the
beginning
of your written or oral statement to
advise the
committee of any financial relationship
that you
may have with the sponsors of any
products in the
pharmaceutical category under discussion
at today's
meeting.
For example, this financial information
may include the sponsor's payment of your
travel,
lodging, or other expenses in connection
with your
237
attendance at the meeting.
Likewise, the FDA encourages
you at the
beginning of your statement to advise the
committee
if you do not have any such financial
relationships. If you choose not to address this
issue of financial relationships at the
beginning
of your statement, it will not preclude
you from
speaking.
We are ready to go and let me
give you the
ground rules before we start, so that
everybody
understands. You get two minutes to talk. We have
a light there that will go on. At 1.5 minutes it
will be green, and then yellow, and then
at zero,
the microphone will go dead and only your
lips will
keep moving.
So, it is important at that point
to sit
down because the next guy is coming up to
take that
microphone.
Let's get started. I will be impolite
enough to call you by number rather than
by name
because that is what I have here. If there are
people who have registered to speak and
have not
238
yet checked in, they need to go to the
check-in
desk outside and check in rapidly or
someone else
will get their spot.
Let's begin with Speaker No. 1.
MS. JOAN JOHNSON: Hello.
I am Joan
Brierton Johnson and this is my
7-year-old daughter
Sabrina.
She writes:
"Dear FDA:
When I was 6 years old, I had
fun visiting
my friends, playing computer games, and
drawing
lots of pictures. All of that ended when I came
home from the first grade, not feeling
very well.
My parents gave me Children's
Motrin, but
instead of getting better, I got
Stevens-Johnson
Syndrome.
Taking Children's Motrin is why
I am blind
today.
Now I wear a hat that covers my
entire
face - even indoors - because the light hurts
my
eyes.
When I go outside, I get teased because of
my hat.
People say mean things to me about it and
that really hurts my feelings.
I liked going to school, but my
immune
system is now so weak because of SJS that
it is not
safe for me to go anymore. I miss my friends.
239
Millions of kids all over the
world are
given Children's Motrin when they get
sick. But it
doesn't have a warning label on it about
SJS.
I would like to ask the FDA to
require a
warning label about SJS on Children's
Motrin and on
any other drugs that can cause this
horrible
disease.
Thank you for considering my request.
Sabrina Brierton Johnson, age
7, Topanga,
California."
Now, Sabrina would like to say
a few
words.
MS. SABRINA JOHNSON: Please do something
so other children don't get hurt by
Stevens-Johnson
Syndrome like me. People really need to
know about
it.
Thank you.
MS. JOAN JOHNSON: Thank you.
DR. WOOD: Thank you very much.
No. 2.
(No response.)
DR. WOOD: No No. 2.
All right. Let's
move on to No. 3, I know he will be here.
DR. WOLFE: Before the clock starts, I
have no conflict of interest.
Four years ago, I testified
before this
240
committee that FDA should require a black
box
warning on Vioxx and Celebrex because of
significant evidence from the VIGOR study
and
trends in CLASS of increased cardiovascular
risk.
What the FDA, the Advisory
Committee, nor
I knew then was that in the year 2000
Pfizer had
finished a study, a placebo-controlled
trial using
Celebrex to prevent Alzheimer's disease
progression
and that the study had found increased
cardiovascular risks for the drug.
What I did not know several
weeks ago,
when I made the results of this yet
unpublished
study public, was that the FDA had been
provided
the results of this study in June of
2001, even
though they held back, Pfizer held back
the study
so that it wasn't discussed at the
Advisory
241
Committee meeting four years ago, which
would have
presented a class effect for Vioxx and
this drug.
FDA was concerned enough about
this study
that it presented it internally at a
meeting in
2001, but never revealed the results to
the public
until yesterday in Dr. Witter's
presentation, which
acknowledged that in almost every type of
adverse
cardiovascular outcome, the cases
occurred mainly
in those using Celebrex, 3 cardiovascular
deaths,
non-fatal heart attacks, strokes, heart
failure or
angina out of 140 in the placebo group,
20 out of
285 in the Celebrex group.
Because of much prevarication,
to put it
mildly, by Pfizer yesterday, Pfizer
testified under
oath they might have been found to have
committed
perjury.
I recommended today that Pfizer be
criminally prosecuted for fraud to the
U.S.
Attorney's Office if they aren't already
conducting
such an investigation, and it appears
that Senator
Grassley's office will take up the
investigation as
to why FDA withheld this information for
so long.
I sent this testimony to them.
Given that Celebrex and Bextra
are making
an
important contribution of the estimated 100,000
deaths and 2 million serious injuries a
year from
242
adverse drug reactions, I hope you will
recommend a
ban of these drugs, not a don't use for
more than
10 days.
DR. WOOD: Thank you.
No. 4.
MS. SUYDAM: Thank you for the opportunity
to present an over-the-counter or OTC
perspective
on the safety of nonsteroidal
anti-inflammatory
drugs.
The Consumer Healthcare Products
Association is a national trade
association
representing manufacturers and
distributors of OTC
medicines and has a long history of
working with
FDA
on important safety issues.
In considering the safety of
NSAIDs, I ask
the Advisory Committee to consider three
important
points.
First, the use of OTC NSAIDs
clearly
should be distinguished from long-term or
chronic
prescription use. OTC NSAIDs have a different
243
overall benefit-to-risk equation and a
wider margin
of safety because they are used at lower
doses and
are not intended to be used on a chronic
basis
unless directed by a physician and are
used for
mild, self-limiting conditions.
Second, OTC medicines differ
from
prescription drugs because the OTC label
contains
all of the information that consumers
need to
decide if the medicine is right for them,
how to
take the product, and when to see their
doctor if
needed.
OTC NSAIDs are not intended to
be used for
long durations unless directed by a
physician, and
this is clearly stated on the label.
Third, OTC NSAIDs are safe for
consumer
use when used according to label
conditions. Every
OTC NSAID has been extensively reviewed
by FDA and
FDA Advisory Committees. This review has confirmed
that OTC NSAIDs are safe and effective
and that the
benefits of OTC use outweigh the risks.
In closing, it is important to
clearly
distinguish the benefit-to-risk equation
for
244
prescription NSAIDs from that of OTC
NSAIDs. The
millions of consumers who rely on OTC
NSAIDs for
temporary pain relief should continue to
feel
confident that these medicines are safe
and
effective when used according to the
label.
DR. WOOD: Thank you.
Jennifer Lo.
DR. LO: To facilitate the benefit and
risk assessment of COX-2 inhibitor in
each
individual, we propose to the Committee a
new test
under development, iHAD test, used to
assess the
cardiovascular disease risk in patients
taking
COX-2 inhibitors.
Our test reveals the
pathobiological
effect of inflammatory
mediators/inflammation
related agents (IRAs) on each
individual's vascular
system ex vivo. Individuals found to be at high
risk because they are likely to suffer
the same
pathobiological effect of IRSs if present
under
desirable conditions in vivo.
The ex vivo pathobiological
effect may be
quantified in the form of cytotoxicity
which can be
245
revealed in 2 general categories:
cytolysis and
cyto-aggregation. The severity of cytotoxicity is
used to determine the level of CVD risk
of
asymptomatic individuals. Individuals tested with
a high risk may choose not to use COX-2
inhibitors.
Others tested with a low risk may benefit
from the
use of COX-2 inhibitors with periodic
retesting.
This picture depicts the
cytolysis of
cultured fibroblast induced by the basic
nature of
a protein like many inflammatory
mediators.
The next picture depicts the
cyto-aggregation of human blood cells
induced by
multiple IRAs. Phospholipase A2 is one of the many
significant inflammatory mediators used
in our
assessment test.
This simplified proposed
mechanism for
Acute Coronary Syndromes (ACS) forms the
basis of
our new iHAD test, including the
involvement of
COX-2 inhibitors. Inflammation produces many
IRAs
and some of them are prothrombotic. PLA2 and other
IRAs act on blood components to cause
cell damage
in the form of cytotoxicity.
Cytolosis may be responsible
for rupturing
atherosclerotic plaques, leading to
thromboembolism, predisposing ACS.
246
Cyto-aggregation may lead to thrombosis,
predisposing ACS.
COX-2 inhibitors prevent the
synthesis of
Prostaglandin (PGE2) that is responsible
for
triggering the pain, but they have no
inhibitory
effect on arachidonic acid (AA) a
byproduct of
phospholipase A2, which is also
prothrombotic.
Our new iHAD test is intended
to evaluate
the response of individual blood cells to
IRAs in
assessing the baseline CVD risk based on
the
severity of cytotoxicity.
We urge all individuals taking
the COX-2
inhibitors or considering taking the drug
to take
the iHAD test.
DR. WOOD: Thanks.
No. 6, Jim Tozzi.
MR. TOZZI: Thank you, Mr. Chairman,
Distinguished members of the
Committee. Having
been a resident of New Orleans, I cannot
speak that
247
fast, and I have burned up 10 minutes or
10 seconds
I am Jim Tozzi. I am the member of the
Board of Advisors of the Center for Regulatory
Effectiveness. The Center receives no funding from
the pharmaceutical industry although a
number of
years ago we did receive grants from the
industry.
The Center is a regulatory
watchdog. To
this end, we have a particular interest
in the FDA
compliance with the requirements of the
recently
passed Data Quality Act. When the agency makes
determinations regarding the benefits and
risks
associated with the use of non-steroidal
anti-inflationary drugs--sorry, I am an
economist--anti-inflammatory drugs. They may be
anti-inflationary, too.
The Data Quality Act required
OMB and FDA
to issue guidelines which would maximize
the
quality, the objectivity, the integrity,
and the
information FDA disseminates to the
public.
So, you may be asking why am I
here.
Well, the guidelines require certain
analytical
results to be reproductive and
248
unbiased--reproducible and unbiased. The Data
Quality Act places no requirements on the
distinguished members of this committee,
however,
the FDA cannot rely upon the information
it
receives from the advisory committee
unless the
advisory committee information meets the
requirements of the Data Quality Act.
Furthermore, any third party,
such as CRA,
can petition under this act for FDA not
to use the
results if they do not comply with the
Data Quality
Act, and I thank FDA for allowing--.
DR. WOOD: No. 7.
Dianna Zuckerman.
MS. ZUCKERMAN: The National Research
Center for Women and Families is an
independent
nonprofit organization with no conflicts
of
interest on this issue.
We focus on research, but we
know that
when Americans take medication, they
don't expect
to have to read the studies that have
been
conducted on the product, and their
physicians
don't expect to have to read them either,
and the
patients don't expect to have to
carefully
249
scrutinize the fine print and personally weigh
the
risks and benefits.
They expect that medications
that are
FDA-approved are safe and effective for
almost
everyone and therefore safe for them.
So, please, when you vote
tomorrow, please
treat your votes as if they are the most
important
ones you will ever make, because there
are a lot of
people depending on you.
There is plenty to be concerned
about
regarding the medications that you are
considering,
but unfortunately, we don't have access
to all the
data that you have access to, so I am
going to
focus on the broader issue, which is the
failure of
the FDA to scrutinize long-term safety
data.
This is a systemic problem and
it will not
be fixed by wishful thinking or by
advisory panel
instructions.
Unfortunately, drugs that are
studied on a
few hundred or even a few thousand people,
for a
few weeks or months, are then taken, as
you know,
by millions of people for many
years. The FDA
250
really doesn't always know what the
long-term risks
are especially if the companies involved
don't
reveal all the information that they
have.
The FDA should be requiring and
carefully
monitoring long-term studies of medical
products
that patients will rely on for a long
time. Our
Government needs to strengthen the FDA
and other
security checkpoints designed to protect
us from
those very real dangers.
In the meantime, please don't
assume that
the companies can be trusted to carefully
conduct
postmarket studies or that the FDA will
enforce
requirements to conduct such studies and
act on
their results in a--.
DR. WOOD: Thank you very much.
The next speaker is No. 8,
Elizabeth
Tindall.
DR. TINDALL: Good afternoon. I am Dr.
Elizabeth Tindall and I am speaking today
as a
practicing rheumatologist from Portland,
Oregon,
and as President of the American College
of
Rheumatology. I have no consulting or financial
251
relationships with the companies or
products being
discussed at this meeting.
The ACR represents more than
6,000
physicians, scientists, and health care
professionals who care for people with
arthritis
and other musculoskeletal diseases. Our
members are
actively involved in treating the
estimated 70
million Americans who are affected by
osteoarthritis, rheumatoid arthritis, and
other
musculoskeletal diseases for which
traditional
NSAIDs and COX-2 selective NSAIDs are
used.
Limited and emerging data about
the
cardiovascular toxicity of COX-2 and non-selective
NSAIDs, which has received widespread
media
coverage, has caused anxiety among the
patients and
the physicians who treat them. We are concerned
that this controversy has damaged public
confidence
and trust in drug safety, and we believe
the
following points are central to the
continued
discussion of this issue.
First, the FDA should lead the
effort to
ensure that patients and the public are
made much
252
more aware of the most common and serious
toxicities of all medications including
those of
the traditional and COX-2 selective
NSAIDs.
This information should be
given to the
public with information about what groups
of
patients may be at greatest risks
including age and
underlying comorbidities. That allows
physicians
and patients to make the best decision
about their
health care.
The American College of
Rheumatology
supports the FDA's efforts to ensure
clarification
of the most important drug toxicities in
all
direct-to-consumer advertising in print
and
broadcast media, and we also applaud the
full
disclosure of any advertising presented
to the
public as promotional educational
material.
We also support the full
disclosure of the
test results of all industry-related
trials for
drugs that are FDA approved, so that
public and
scientific scrutiny may occur. We applaud the FDA
in forming a new independent drug safety
oversight
board this week. This board must ensure that all--.
DR. WOOD: Thank you very much.
The next speaker is No. 9. Dimitra
Poulos.
253
MS. POULOS: Good afternoon and I am here
at my own expense.
Every time you take a drug,
there is a
risk factor to be considered. I believe it's
important for the government to keep us
informed on
all drug findings and potential risks, so
we are
able to make informed decisions.
Cigarettes come with a warning
label,
there is no prescription needed for
alcohol, yet
taken by the wrong person, we are all at
risk.
Liver is damaged from Lamasil
and Lipitor,
Coumadin is a risk of bleeding to death.
When I was diagnosed with rheumatoid
arthritis in 1998, my life changed
dramatically.
Professionally, it had an impact on the
quality of
my work.
Socially, I could no longer sit in a
movie theater, take a walk, car trips to
visit
out-of-town family members was out of the
question.
Personally, arthritis attacked
my husband,
254
too.
He had to assume most of my responsibilities
for running the house. As daily functions became
impossible for me, I needed his help to
get
dressed.
On day he found me in the bathroom, on
the commode, crying, unable to get off of
it.
But that was before Vioxx. I have taken
Vioxx for over 5 years with absolutely no
side
effects.
Vioxx gave me my life back. We
have no
idea of the risks involved with any of
the new
drugs, but a known risk can be dealt
with.
As I speak, I have 40 Vioxx
left. I have
40 days before my life and my abilities
will be
severely altered.
I will assume all
responsibility and sign
any waiver. Please give me that option and thank
you for allowing me this time.
DR. WOOD: Thank you very much.
The next speaker is No. 10,
John Pippin.
DR. PIPPIN: Before the clock starts, may
I mention my affiliations? I am here representing
myself and the Physician's Committee for
Responsible Medicine, a nonprofit. I have no
255
commercial affiliations.
While the primary focus of
these meetings
concerns whether the COX-2 inhibitors
should be
withdrawn from clinical use, we also must
address
the more fundamental problem regarding
drugs
developed and approved in the U.S., and
that
problem is how to identify safe and
effective drugs
before they are approved for human use.
The greatest obstacle to
accomplish this
goal is the continued use of animal
testing to
evaluate drug safety and efficacy. For reasons
which are genetically based and
immutable, drug
testing in rodents, rabbits, dogs, and
monkeys
produces widely different results, none
of which
correlates with human results.
For example, 9 of 11 studies of
vascular
disease in mice and rats showed that
COX-2
inhibitors, the very drugs we are talking
about
today, were beneficial for heart disease,
and, in
fact, some of the investigators suggested
they
would be useful drugs for heart
disease. We know
from the clinical trials that all three
COX-2
256
inhibitors are dangerous for heart
disease.
What I have just told you is no
secret.
Everyone involved, the pharmaceutical
companies,
their researchers, the FDA, we all know
that animal
testing is unreliable. However, we have
been
unreasonably slow to replace animal
testing with
newer and better tests for drug safety
and
efficacy.
First of all, we must eliminate
animal
testing from this process since this
flawed method
costs billions of dollars and tens of
thousands of
human lives annually in the U.S. In-vitro testing
using human cells and tissues,
computer-based
modeling, microdosing studies in humans,
stem cell
technology to allow testing of human
cells and
tissues, and the burgeoning field of
pharmacogenomics, which allows us to
compare DNA
and predict toxicity and efficacy of the
drugs.
They are all superior to animal
testing.
We should be promoting these
methods. As a group,
these methods are light years ahead of
our crude
animal tests, they are safe, accurate,
and cost
257
effective, and we must move toward these
methods if
we are to have safe and effective
medicines in
America.
DR. WOOD: Thank you.
No. 11.
Major Grubb.
DR. GRUBB: I am Christopher Grubb, M.D.
I am in the Army Medical Corps at Fort
Bragg, North
Carolina.
I am supported by the Department of
Defense and I have no financial
interests. As a
military physician, I have no other
interests at
heart but the health and safety of our
men and
women in uniform.
As a pain specialist, my
mission is to
conserve the fighting strength by
treating acute
and chronic pain in our active duty
soldiers and
returning them to the battlefield.
However, we don't like to send
soldiers
into harm's way on non-selective NSAIDs
due to
their anticoagulant effects and the
potential for
worsening bleeding after battlefield trauma.
Instead, they go to war with COX-2
selective
inhibitors or coxibs.
Consequently, the 82nd Airborne
Paratroopers are required to carry a
coxib drug to
be taken in the event of a battlefield
injury, one
258
of three drugs in what is called the
soldier's pill
pack.
Many soldiers are fearful of
the bleeding
risk with NSAIDs, so they ask
specifically for
coxibs.
Since service members are young and very
physically fit, the armed forces
constitutes one of
the lowest cardiovascular risk
populations in our
society, so the recent COX-2 risk data
was of very
little concern to the military.
So, in this meeting, we warn
against using
a broad brush when painting the portrait
of risk.
Military personnel suffer frequent
injuries and
have a higher incidence of chronic pain
than
civilians, further increasing our need
for coxibs.
Coxibs have allowed the
worldwide
deployment of many previously disabled
soldiers.
Many are now in Iraq on daily regimens of
coxibs.
Without these products, we can't keep as
many
soldiers functional on the battlefield.
The study of coxibs for chronic
pain is in
its infancy. Although efficacy data for coxibs may
be equivocal for arthritic conditions
versus
NSAIDs, the same can't be assumed for
other types
of pain.
Indeed, most military personnel use
coxibs for non-arthritic pain, such as
low back
259
pain.
We have found coxibs to be superior to
NSAIDs for spine pain, so we are planning
controlled trials of our own to compare
these drugs
head to head.
In summary, our bravest
Americans are
reaping benefits from coxibs without drug
adverse
events.
This large population should not be
disenfranchised here. Consider our military in
this particular drug decision. Coxibs are
essential in the global war on terrorism.
Thank you.
DR. WOOD: Thank you.
Dr. Arrowsmith Lowe, No.
12. Not here?
Okay, we will go on to No. 13, Mark
Einstein.
DR. EINSTEIN: My name is Dr. Mark
Einstein and I am an Assistant Professor
of
260
Gynecologic Oncology at the Albert
Einstein College
of Medicine, Montefiore Medical Center at
Bronx,
New York.
My academic department has
supported my
expenses to attend this meeting. I have not been
asked to speak to you by any
pharmaceutical
company, however, one of my clinical
trials is
partially supported by an unrestricted
grant from
Pfizer.
As a gynecologic oncologist, I
am
committed to finding new therapies to
prevent and
treat women's cancers. Recent trend data
suggest
cancer is overtaking cardiovascular
disease as the
leading cause of death in the U.S.
COX-2 inhibitors are one of the
promising
class of agents used in cancer therapy,
however,
many current and planned cancer clinical
trials
using COX-2 inhibitors are on hold
pending the
results of these hearings.
Expression of COX-2 has been identified in
many human cancers including gynecologic
cancers.
One of the COX-2-expressing cancers is
endometrial
261
cancer, which is the second most common
gynecologic
malignancy in the U.S. after another
COX-2-expressing cancer, breast cancer.
The number of deaths from
endometrial
cancer has risen 128 percent since
1987. Responses
to toxic chemotherapy in women with
recurrent
endometrial cancer are dismal. These generally
elderly women have comorbidities that
also limit
their tolerability of chemotherapy.
We identified high rates of
COX-2
expression in the most chemo-refractory
endometrial
cancers.
These data led us to begin a pilot trial
using Celebrex in women with endometrial
cancer
that is grant supported by the American
College of
Ob-Gyn.
This trial has been suspended.
Cervical cancer, the number 1
cancer
killer of women in many countries also
strongly
expresses COX-2. Currently, two
cooperative group
trials that were designed to observe the
effects of
Celebrex in pre-invasive cervical cancer
have also
been suspended.
COX-2 inhibitors are one of the
targeted
262
agents that are being used for
prophylaxis in women
at risk for ovarian cancer where survival
using
toxic chemotherapy regimens has not
changed in over
15 years.
In summary, gynecologic cancers
remain a
critical issue in women's health and
standard
therapy are not very effective at
limiting the
death rate and are not well
tolerated. The thought
of using target agents, such as COX-2
inhibitors
that have less toxicities than most
chemotherapies
have many--
DR. WOOD: We found No. 12.
DR. LOWE: My name is Janet Arrowsmith
Lowe.
I am a physician and epidemiologist and the
president of a small consulting firm in a
tiny town
in New Mexico. I do want to state that some of my
clients, my pharmaceutical clients
include Bayer,
Glaxo-Smith-Kline, Merck, Pfizer, and
Wyeth, but
today I am just representing myself and
my firm.
It has been refreshing to hear
discussion
of risk and benefit, because I think too
often in
the press, concerning safety of marketed
drugs only
263
risk is discussed, and I think as we all
know, that
when a product is approved, FDA weighs
risk and
benefit before approval.
Now, the calculus may change
over time as
new drugs or new information is
available, but in
my several years of experience at FDA,
and since
leaving, I am assured that the agency is
still
functioning, and I don't believe that FDA
is
broken.
It is not perfect. Is there a perfect
institution? But it probably can be improved, but
I think the proposals for a separate
agency for the
review of safety are not rational. I think that
the premarket review really provides
appropriate
balance in deciding whether a product
should stay
on the market.
Now, I would like to see
greater access to
some drug development data including more
user-friendly public access to the safety
databases
at FDA modeled along the lines of the MOD
database
in the Center for Devices.
So, in my opinion, the public
health is
264
best served by a careful study of risks
and
benefits, and FDA, with the proper
funding balance
and authority, an engaged industry, and
an educated
public.
Thank you very much.
DR. WOOD: Thank you.
Next, we will go to No. 14, who
is Dr.
Abramson.
DR. ABRAMSON: Thank you for having me
here.
I do serve as an expert on cases involving
Vioxx and Celebrex. I want to say that in order to
get to the bottom of what went wrong with
Vioxx, I
think it is important to address first
what went
right.
At the February 2001 Advisory
Committee
meeting, the reports of the FDA reviewer
showed
conclusively that Vioxx caused
significantly more
cardiovascular complications in people
with and
without cardiovascular history, and
overall, the
people who took Vioxx developed 21
percent more
serious complications.
So, the question before us is
why do
265
American physicians prescribe $7 billion
worth of
Vioxx after Merck and the FDA knew that
Vioxx was
significantly more dangerous, no more
effective,
and far more expensive than naproxen.
In order to answer that
question, we need
to look at the sources of information
that
physicians trust most. That data was
reported in
the New England Journal of Medicine in
2000. The
article acknowledged that there was a
cardiovascular risk in theory and
measured
cardiovascular events, but the article
did not
report those cardiovascular events, nor
did the
article report serious adverse events
overall.
It did report heart
attacks. The heart
attacks were reported as not
statistically
significant in people without a cardiac
history,
and therefore, the issue was not brought
to
physicians' attention. All 13 authors had
financial ties to Merck.
We look at the clinical
practice
guidelines from the American College of
Rheumatology. We see that first is Tylenol, and
266
next recommended is Vioxx and
Celebrex. All four
authors have financial ties to the
manufacturers of
both drugs.
The problem here is that the
information
that docs are getting is so heavily
filtered
through commercial sources that no matter
what the
FDA does with drug safety, unless the
integrity or
doctors' information is not improved and
doctors
and patients don't take good information
into the
exam rooms, this exercise is going to be
for
naught, and the quality of American
medicine will
not improve.
DR. WOOD: Thank you.
We will go to No. 15, Dr. Baraf.
DR. BARAF: I have consulted to and
performed clinical trials for many of the
companies
whose drugs are being discussed today.
As a busy practicing
rheumatologist, I
have asked to be here to speak for my
patients with
arthritis. For four and a half months, their needs
have been ignored in virtually every news
report
and medical journal editorial discussing
NSAID
267
therapy.
Indeed, we have all learned
that we must
be more mindful of each patient's risk
factors for
cardiovascular disease in selecting
COX-2s or other
NSAID treatment, but data regarding this
risk for
COX-2 inhibitors is incomplete, sometimes
contradictory, and begs further
investigation.
The risk for cardiovascular
disease with
non-selective NSAIDs is unknown and
untested. I
urge this panel to give careful thought
to the
considerable benefits COX-2 inhibitors
offer
patients with arthritis especially those
with GI
risks.
For large numbers of my
patients, COX-2
inhibitor diminish the threat of serious
drug-induced gastrointestinal injury,
thereby
eliminating a major barrier to their
treatment.
How are we to balance the competing risks
of
cardiovascular and GI toxicity against
real
therapeutic need for patients with
debilitating
pain?
We must heed the advice that we
give to
268
our patients. There are no completely safe drugs
in any treatment category. It is my responsibility
to weigh and risks and benefits of drugs
with my
patients, to make individualized
decisions.
Sensationalizing and
highlighting only the
risks of these drugs based on scanty and
incomplete
information, as many of our colleagues
have chosen
to do, have created an atmosphere in
which an
informed discussion with patients is
difficult, if
not impossible.
For many patients with
arthritis, these
drugs are not superfluous as some have
suggested,
but greatly impact their quality of
life. To
withdraw one drug might put us on a
slippery slope,
leading to withdrawal of all NSAIDs. My patients
must not be denied access to the widest
variety of
therapeutic options.
Thank you.
DR. WOOD: Thank you.
No. 16. Dr. Hamburger.
DR. HAMBURGER: I am a practicing
rheumatologist and the President of the
New York
269
State Rheumatology Society. I have been a speaker
for several of pharmaceutical companies
mentioned
today.
I polled New York
rheumatologists, State
rheumatology society leaders, and I spoke
to my
patients, and we have remarkably
consistent views.
Events have reminded everyone of what
rheumatologists and our patients already
know.
NSAIDs are important because of their
role in the
treatment of the pain of arthritis and
because of
the numbers of people who suffer from
this pain.
We have seen recently far too
many
patients who have experienced the
recurrence of
their pain and their suffering because
they stopped
their medications out of fear or because
of changes
in managed care formularies.
None of us can emphasize enough
the
importance to these patients of reducing
their pain
and preserving their mobility. So, our consensus
opinions are, number one, that access to
anti-inflammatories needs to be
preserved.
Physicians and patients need to be
provided with
270
the important information about these
medications
in a more rational and timely fashion,
and the
process for disseminating this
information should
be improved.
The coxibs, we have learned
today, and we
have known, have less GI toxicity, but
their own
side effects. Everyone wants an NSAID free of
toxicity, but no one can say today to any
patient
that this NSAID has been tested and found
to have
no CV, GI, or renal toxicity.
So, we need to maintain access
while
deciding the best next research.
Patients act on what they read
and hear,
and they believe the information that
appears in
the media. The evidence on NSAIDs presented to the
public has focused on only a small number
of
published studies, and the public is
making its
judgments without knowing all the
information.
Juries in this country do not
deliberate
and reach a verdict based on the last three
pieces
of evidence.
DR. WOOD: Thank you.
The next speaker will be Dr.
Qureshi, No.
17.
DR. QURESHI: Good afternoon. Before I
271
start I should let you know that I am
being paid by
Given Imaging to be here, but not enough
to
influence my results.
I am going to talk about NSAIDs
and the
small intestine injury they cause. The occasional
findings of intestinal blood loss or
anemia in the
setting of normal upper and lower
endoscopy led to
the realization that NSAIDs cause
significant
disease in the small intestine.
We performed the first
controlled study to
look at NSAIDs using new technology that
is a
camera pill that takes a video wirelessly
of the
small bowel. We looked at 41 patients, half of
them on NSAIDs for at least three months
and half
that took Tylenol or nothing.
This is a camera that you
swallow.
Much to our surprise, we found
small
ulcers in the small bowel, large ulcers,
and
bleeding in the small intestine.
We found that 71 percent of
NSAIDs takers
had some form of injury in their small
intestine,
20 percent had severe injury compared to
none in
the controls.
So, symptoms and signs of ill
health among
chronic NSAIDs users is often attributed
to the
272
underlying disease, but we think that
dyspepsia and
not responding to acid suppression, vague
abdominal
symptoms, iron deficiency anemia, or
hypoalbuminemia may result from small
intestinal
injury.
We have a new technology now
that enables
us to look at the small intestine. Video capsule
endoscopy is very useful for diagnosing
and for
comparing the damage that different
NSAIDs might
cause on the small bowel, and in a subset
of
patients where we suspect small bowel
injury, this
technology is useful and shows promise.
Thank you.
DR. WOOD: Thank you.
We will go on to No. 18, Mr.
Matthews.
MR. MATTHEWS: Thank you.
My name is
273
David Matthews. I am a lawyer and I represent
individuals who have been harmed by the
drugs being
discussed here today.
The fact that these hearings
have become
necessary to address the safety of COX-2
drugs is
yet another tragic example of the
continuing
failure of the pharmaceutical industry to
disclose
the truth, the whole truth, and nothing
but the
truth to the FDA, prescribing physicians,
and the
citizens of this country.
Why is the whole truth not
forthcoming?
Simple. Billions and billions of profit
dollars and
absolutely zero individual accountability
by
company officers who submit drug safety
data both
before and after a drug is approved.
With the coxibs, the FDA has
had to
negotiate with the drug sponsors to change
labels,
conduct patient and physician education,
limit
advertising, modify approved indications,
and to
even complete studies.
The time for these negotiations
should
end.
In response to a rash of corporate scandals
274
involving the likes of Tyco, WorldComm,
Enron, and
others, Congress passed the
Sarbanes-Oxley Act of
2002.
It provides criminal penalties of up to $5
million and 20 years in prison for
knowingly
submitting false finance information to
the SEC.
These penalties are for lying
about a
company's financial status, not for
causing injury
or death to an individual. Because everyone
deserves nothing less than the whole
truth from
pharmaceutical companies and complete
disclosure
about clinical trial data, there must be
personal
accountability for any individual who
fails to do
so.
I urge Congress, and I hope
these hearings
can be a springboard, to enact
legislation which
follows the Sarbanes-Oxley Act, but with
more
severe penalties for any drug company,
officer, or
employee who submits false, misleading,
or
deceptively modified drug safety data to
the FDA, a
physician, or to the public.
If someone who submits false
financial
information to the SEC can be filed $5
million and
275
sentenced to 20 years in prison, there is
no
compelling reason that the penalties for
submitting
false, misleading, or deceptively
modified data to
the FDA.
DR. WOOD: Next speaker will be No. 19,
Dr. Wilson, and as you start, Dr.
Wilson--we are
not counting your time yet--try and step
back a
little bit from the microphone. Apparently, there
is a lot of distortion from people being
too close
to the microphone, and that goes to the
other
speakers, as well. Thanks.
DR. WILSON: First of all, I have no
sponsorship, I am here on my own
recognizance. I
am a practicing rheumatologist in
Atlanta, Georgia,
and my life is dedicated to alleviating
the pain of
arthritis.
Almost 2 million Georgians
suffer from
arthritis. In fact, the latest figures from the
CDC are that 1 in every 4 Georgians has a
chronic
joint symptom, and arthritis is the
number one
cause of disability in America.
Pain matters. It may not kill you, but
276
you may wish that you were dead.
My patients are not concerned
about living
forever, they want to live well without
arthritis
pain.
It is not surprise that the more experience
we gain using medications, the more we
learn when
to use it and when not to use it. Patients do not
take medications if they don't work, and
millions
of patients taking COX-2 selective
medications
evidence that they are effective. Indeed, this has
been my experience.
I am concerned about
safety. We should
try to figure out what is unique about
the 1 to 2
percent of patients with very serious
side effects
rather than depriving the 98 to 99
percent of
patients with significant relief from
their
arthritis pain who have not experienced a
serious
side effect.
In a perfect world, I would
have endless
choices because all patients are not
created equal.
I believe that the choice to choose COX-2
selective
medications is too important to answer
for the
patient.
To limit choices based on evolving
277
knowledge is unfair to tens of millions
of
Americans with arthritis pain.
On average, 29 people a week
die in a car
in Georgia. I suspect that all of us came in a
motor vehicle today and accepted a risk.
We must consider both sides of
the
equation when we decide how to treat
patients and
what to treat them with. Ideally, it
should be a
patient's decision to decide based on the
information provided by their personal
physician.
Most of my patients would take
some
significant risk for a better quality of
life with
relief from arthritis pain. Please thoughtfully
consider our patients' pain when you make
your
decision.
Thank you for your time.
DR. WOOD: Thank you.
The next speaker is No.
20. Dr. Williams.
DR. WILLIAMS: I am Dr. Gary Williams. I
am here on my own time and at my own
expense.
It is generally accepted that
COX-2
inhibitors are a safer alternative to
patients with
278
arthritis. Cost containment has been a competing
force.
Those among us who feel these drugs are
expensive or overused may be pleased with
the
recent changes in the market share of
COX-2
specific drugs.
This shift has been caused
largely by
prolonged concerns regarding Vioxx,
culminating in
the decision by its manufacturer to withdraw
the
drug from the market.
Our current attention is
directed to
possible cardiovascular risks for two
currently
marketed drugs, celecoxib and
valdecoxib. The data
that concerns us is to date in
non-arthritis trials
designed to explore possible additional
uses of
these drugs beyond their current
indications.
The largest effort to date to
assess the
impact of these drugs on cardiovascular
risk in
patients using them for their current
indications
is the FDA-sponsored Kaiser trial. This trial
reinforces the cardiovascular risk for
users of
Vioxx and raises additional concerns for
possible
increases in cardiovascular risk in users
of
279
nonsteroidal anti-inflammatory drugs
including
Naprosyn.
In this trial, Celebrex was not
associated
with increased risk compared to any other
treatment
option or even when compared to non-users or
remote
users of any of the treatment options.
On this background, we should
be cautious
in recommending that thousands, or even
millions,
of current users of COX-2 specific
inhibitors move
to other, older non-selective NSAID
options.
We should be realistic and
assume that
they will continue to use
anti-inflammatory drugs
obtained either over the counter or by
prescription. Since they would be moving away from
the GI safety advantage demonstrated with
the COX-2
selective drugs toward the options
included in the
Kaiser trial, they would be moving toward
increasing GI risk.
Unfortunately, as it relates to
the
decisions facing this Advisory Committee,
the same
FDA Kaiser data suggests that the
recommended
movement--
DR. WOOD: Thank you.
The next speaker is Rebecca
Burkholder,
No. 21.
280
MS. BURKHOLDER: I am Rebecca Burkholder
from the National Consumers League. In the
interest of full disclosure, NCL
occasionally
receives unrestricted financial support
from
pharmaceutical companies for consumer
education and
research projects. The research cited below is one
of those projects. My expenses for this meeting
were not paid by an external organization
and my
statement reflects the interests of those
NCL
represents, consumers.
NCL urges the FDA to carefully
weigh the
risk and benefits of COX-2 inhibitors as
it decides
how best to protect the public. Whatever action
this committee takes, NCL believes it is
important
to anticipate consumer response in the
wake of the
publicity surrounding COX-2 drugs.
Although COX-2 drugs were
originally
intended for use by those patients who
had GI side
effects with traditional NSAIDs, a much
broader
281
population actually took the
medications. Given
recent events, some patients taking COX-2
drugs for
arthritis for other pain will now likely
turn back
to traditional over-the-counter NSAIDs
for relief,
but consumers likely do not understand
how to
safely use these OTC NSAIDs.
A 2003 survey of over 4,000
adults
commissioned by NCL on consumer use and
attitudes
towards OTC pain relievers found that 47
percent of
those who take OTC NSAIDs take more than
the
recommended dose. Nearly half would not consult a
doctor when taking for more than 10
days. Nearly
half thought it was more important to
control pain
regardless of risk, and the survey
revealed the
following about arthritis sufferers - 85
percent
take OTC for pain relief with 60 percent
choosing
OTC NSAIDs, 30 percent take pain
relievers on a
daily basis, and 70 percent do not
discuss the
risks.
Based on these findings, we
believe
consumers must be educated about the
relative risks
and benefits of all medications, OTC or
282
prescription. We call upon the FDA to engage with
relevant partners in a broad-based
educational
campaign that would cover relative risks
and
benefits of various pain medications,
appropriate
pain management strategies, the
importance of
talking with a health care professional,
and the
role--.
DR. WOOD: Thank you.
The next speaker is
No.
22. Amye Leong.
MS. LEONG: My name is Amy Leong. Before
I begin, I would like to say that while
my funding
here was as a result of the Foundation
for Better
Health Care, a nonprofit health education
firm, I
have had a role as a motivational speaker
in
previous years with several of the
pharmaceutical
companies mentioned today. However, my presence
here today is as a concerned patient and
a citizen.
As President and CEO of Healthy
Motivation, a consulting firm in health
education,
and as spokesperson of the United
Nation's endorsed
Bone and Joint Decade, I am very
concerned about
the issues that you all are addressing
today. I am
283
very pleased that you are addressing
them, but I
think that we need to look at the
benefit-risk that
you are all so diligently doing today.
I am that patient that you are
addressing.
I have got rheumatoid arthritis, I have
had it for
over 25 years. Within 8 years of
diagnosis I ended
up in a wheelchair, unable to feed
myself. As a
teenager, not being able to walk or feed
herself,
it
is one of those frightening scenarios that we
know should not ever happen.
Because of arthritis
medications that did
not work in my years, I ended up going
through 16
surgeries, 12 of those were joint
replacements. I
have been hospitalized for over 312 days,
and have
indeed taken over 35 arthritis
medications
including every single nonsteroidal
anti-inflammatory and the celecoxibs.
So, I am here today to just
tell you and
to share with you that while we look at
risk, we
really do have to consider the
benefit. I am a
standing benefit in front of you. It is my choice
to work with my physician to determine
what is at
284
higher risk for me and what is not.
Every single arthritis
medication I have
taken has come with some serious adverse
effect -
abdominal pain, fluid retention, gastric
ulcers,
upset stomach, nausea, vomiting,
heartburn,
indigestion, ringing in the ears,
reduction in
kidney function, increasing liver
enzymes, rash,
weakness, unusual tiredness,
sleeplessness,
sleepiness, respiratory infections,
infections,
sepsis, and it goes on and on.
This is what I deal with.
DR. WOOD: Thank you.
No. 23. Donna Fox-Keidel.
MS. ZUCKERMAN: My name is Diane
Zuckerman. I am here on my own to read for Donna.
She was unable to attend because her son
is a
juvenile RA patient, and he had a serious
flare.
She writes:
"I am 39 years old and
have lived with
scleroderma and juvenile rheumatoid
arthritis for
35 of those years. I began taking Celebrex in 2001
as part of my treatment plan. Prior to
2001, I had
285
been on almost every medication known to
treat
juvenile arthritis. I had endured many corrective
and replacement surgeries. I have suffered
setbacks and side effects too many to
mention.
When my doctor spoke of this
new
medication called Celebrex, I was indeed
skeptical,
what would the side effects of this new
medication
bring to me, headaches, fatigue, and the
dreaded
gastrointestinal problems I had learned
to despise,
would it alter organ function, or, better
yet,
would it really even work, because so many
medications I had experience had not
shown any
benefit, and my drug cocktails were never
less than
two medications and that is not counting
the
injections I received.
With my skepticism aside, I tried the new
drug and within weeks saw a remarkable
difference.
I was able to attend school full time
versus part
time, I was able to manage my home
better, and,
most importantly, I was able to be a mom
I wanted
to be.
I was able to spend quality
time with my
286
boys, maintain my home, and continue my
work with a
volunteer group I started for children
with
arthritis. My life was full for once and I was
able to enjoy every moment of it.
For once, taking medication
didn't mean
chasing the pills with a bottle
antacid. I could
eat without fear of feeling
nauseated. My then
90-pound frame was able to gain 15
pounds. For a
brief period of time, I was taken off
Celebrex due
to insurance issues. I was borderline depressed
because I was afraid my new-found life
would
disappear. Fortunately, this did not happen
because my rheumatologist and I fought
for my--."
DR. WOOD: Thank you.
The next speaker, Erika
Umberger, is she
here?
No? All right.
Let's go to No. 25, Theresa
Ray.
MS. SARAFIN: Hi. I
am Judy Sarafin. I
am here on my own and speaking for
Theresa, who was
unable to attend due to a last-minute
emergency and
she asked me to read her story.
"I am 35 with a history of
osteoarthritis
287
starting in college. After the birth of my second
child, my arthritis worsened. Advil wasn't
working, my GP gave me Celebrex, which
worked for
about four months. When that was no longer
sufficient, he sent me to Dr.
Fleishman. Together,
we worked through Mobic and Bextra before
settling
on Vioxx.
With the combination of Vioxx,
multivitamins, glucosamine, and avoidance of
caffeine, I became stable. For the first time in
about five years, I could honestly say
that I had
periods of time where something didn't
hurt. I
could always feel pain somewhere prior to
this
point.
I reached stability with the
Vioxx
combination in August of 2004. When the FDA pulled
Vioxx, I had no choice but to go back to
the Bextra
at least temporarily. Once again, Bextra failed to
give me a sufficient quality of
life. I hurt so
badly I could feel it in my toes.
We are now trying to find
something that
will return me to my Vioxx quality of
life. My
288
family has no history of heart disease or
stroke,
my blood pressure is perfect, and my
cholesterol is
ideal.
I understand and do not wish to dispute
that Vioxx can cause some serious
complications in
a
certain portion of the population, however, what
about someone with my medical history?
I completely agree that all new
information, whether good or bad, should
be
disseminated to patients and physicians,
but I
believe the withdrawal of Vioxx was
premature.
Each patient and physician should be
allowed to
perform the risk-benefit assessment and
further
studies should be performed to fully
understand the
interaction before removing this drug
from the
marketplace."
DR. WOOD: Thank you.
The next speaker is No. 26,
Judith
Whitmire.
MS. WHITMIRE: Pfizer has paid my travel
expenses. I came from Reno, Nevada. I contacted
Pfizer, though, because I wanted to try
to keep my
drug of choice, Celebrex, on the market,
so that is
289
why I am here today.
When I was a young teenager, I
helped my
grandfather in his home printing
business. It was
difficult for him to set type since his
hands were
even worse than mine are now. Certainly, I never
did think that my hands would resemble
his one day.
Now I face a similar
challenge. When I
retired at the end of 2002 from a 40-year
career in
public health microbiology, which was a
problem
with my hands, my husband introduced me
to the
wonderful world of woodturning. It seems I have a
natural talent and my wooden bowls are in
much
demand if I can only keep my
osteoarthritis under
control, and this is what I do and love.
I will be 65 years old next
week.
Subsequent to a severe whiplash when I
was 16, I
developed osteoarthritis in my neck at
the age of
30.
It was then that I embarked on the search for
an effective anti-inflammatory.
I started with Cliniril and
have spent the
next 30 years trying all of the new drugs as
they
became available. They either provided limited
290
relief or caused me gastritis, or
both. I had a
three-day run on Naprosyn before my
stomach said
no.
My new rheumatologist
prescribed Celebrex
last fall for the osteoarthritis in my
hands, neck,
and right knee. It gives me far better relief than
all of the other anti-inflammatories, and
no
gastritis.
I do not have any risk factors
for
cardiovascular disease. Interestingly enough, most
of my family has died of cancer. My rheumatologist
is comfortable with my low dose regime of
200 mg
per day.
I urge you to keep this drug available
for the clinicians to judge if it is
appropriate
for their patients like me.
Thank you.
DR. WOOD: Thank you.
The next is No. 27, Judy Fogel.
MS. FOGEL: My name is Judy Fogel. I
drove myself here from my home in Ithaca,
New York,
to talk to you today. I found out about
this
hearing from inputting in Google the word
Celebrex,
291
a drug I have been taking with great
success for
three years.
I feel like Celebrex was
created for me.
My OA started when I was in my early
20s. It
started with pain and stiffness in my
fingers. The
symptoms continued to worsen. In the early '70s, a
rheumatologist had me take increasing
doses of
aspirin, which led to gastric upset and
ringing in
my ears.
Since there was no other drug available,
I would sometimes take an aspirin and
just pay the
consequences.
We raised three children and
being a
soccer, football, and ice hockey mom,
cold weather
environments was especially
difficult. In the '80s
and early '90s, I tried about 10 of the
NSAID
drugs.
As each new one came on the market saying
it was better than the preceding one, I
would take
one pill and have gastric upset,
bruising, and
ringing in my ears.
Three years ago I went to my
rheumatologist with an inflamed right arm
and hand.
He prescribed a new drug that would be
easier on my
292
stomach, he said. It was called Celebrex. He gave
me samples and a prescription form.
After taking the samples with
no adverse
aftereffects, I had the prescription
filled and
have taken 200 mg of Celebrex each day
ever since.
It took several months to have the
pain
and swelling in my right hand and arm
subside, so I
could use them again, and gradually, the
morning
stiffness and pain in the rest of my body
was
remarkably better.
Most days I feel better than I did 30
years ago. I downhill ski, play golf,
shuffle cards
at bridge, sit through days of lectures
and take
notes, dig and clip in my gardens. I have regained
the manual dexterity--.
DR. WOOD: Thank you very much.
The next one is Dr. Preston
Mason, No. 28.
DR. MASON: Thank you.
I would also like
to acknowledge the contribution of my
colleague,
Professor Corey, Nobel laureate in
Chemistry.
Both the studies I will discuss
were
conducted without interference from the
293
pharmaceutical industry. We both purchased the
drugs used in our studies. I have received
unrestricted grants from the
manufacturers of these
drugs.
Dr. John Vane, also a Nobel
laureate,
suggested as early as 2002 that
differences in CV
risk observed among COX-2 inhibitors may
be
attributed to their physico-chemical
properties.
Confirmation of this hypothesis
was
provided by Professor Corey. He reported that
rofecoxib readily formed potentially
cardiotoxic
metabolites under physiologic
conditions. One of
these metabolites would promote LDL
oxidation, a
well-known contributor to
inflammation. Such toxic
metabolites were not observed in the
other agents
he tested.
The findings of Professor Corey
corroborate our own findings submitted
before Vioxx
was removed from the market. We showed that this
drug dramatically damaged LDL and
membrane lipids
through oxidative modification. We saw this at
pharmacologic levels.
In this figure, we also show an
increase
in isoprostanes, a mediator of
inflammation, and we
again report that this change in LDL
oxidation was
294
not seen among other agents tested.
In the next slide, we contrast
the
pro-oxidant effects of Vioxx against a
potent
antioxidant. Remarkably, the combination only
partially attenuated the effects of the
rofecoxib.
We also saw that rofecoxib
reduced the
capacity of human plasma to defend
against free
radicals.
We have seen, and others have reported,
similar changes in patients with diabetes
and a
recent MI.
The next slide is a further
explanation
for the cardiotoxicity. We evaluated its molecular
effects on lipid structure. Vioxx indeed altered
lipid structure in a manner that we have
seen
consistent with increasing rates of
oxidative
damage.
We also saw adverse effects on
lipid
structure and oxidative damage with
etoricoxib,
another sulfone-type agent.
So, in summary, the last slide,
we have
seen increased reactive oxygen species
with
rofecoxib that contribute to mechanisms
that lead
to cardiotoxicity.
Thank you.
DR. WOOD: Thank you.
295
Is No. 29, Dr. Ross, here?
No. All
right.
Let's move on to No. 30, Dr. Singh.
DR. SINGH: I am Gurkiepal Singh and I am
here on my own. This morning you heard data from
the collaborative study that David Graham
and I
did.
I am also the lead author of the Estimate of
NSAID GI Bleeds in the Country that Dr.
Cryer
referred to, and as a handout, I provided
you our
latest study on the hospitalizations
because of
complicated gastric and duodenal ulcers
in the
United States from 1988 to 2001 that I
presented in
a plenary session last year.
In the next 30 seconds,
reviewing it very,
very quickly, if you go on to page 3, the
top slide
on
the right side shows you what we found, that
there were a total of 493 million
hospitalizations
296
in the U.S. and 3.6 billion patient
years, and over
the years, there has been a decline in
the amount
of gastric and duodenal ulcer
complication
hospitalizations in the country with two
periods of
remarkable decline, the first one '94 to
'95,
perhaps coinciding with the introduction
of H.
pylori guidelines by the NIH, and the
second one in
1999, coinciding with, not necessarily
caused by,
the introduction of COX-2 inhibitors.
The last slide also shows you
the same
rate expressed for 100,000 NSAID
prescriptions, and
you would see that the 1999 decline was
of 22
percent.
We do not know what causes it, but here
are the numbers.
One last point I would like to
make on our
Medi-Cal study, is that we did look at
the recent
exposures and current exposures and
remote
exposures. I know that issue came up, and the
study was internally consistent and that
the
current exposure was always the highest
followed by
the recent exposure and then the remote
exposure.
So, internally, we were consistent in
defining that
297
exposure.
Thank you very much, ladies and
gentlemen,
and I will be here to answer any
questions that you
want.
DR. WOOD: Thank you very much.
The next speaker is No. 39, Dr.
Allan
Fields.
DR. FIELDS: Good afternoon. My name is
Dr. Allan Fields. I have been a physician
practicing general and pelvic surgery and
sports
medicine for over 30 years. Presently, I
am also
the medical spokesperson for Swiss
Medica, the
maker of 024, Essential Oil Pain
Neutralizer.
This is a potent, safe, and effective
topical analgesic. It contains only natural
ingredients that have been clinically
studied and
tested in the U.S. and around the world
including
double-blind studies. It carries a U.S. process
patent.
As physicians, we have taken an
oath to
provide the most effective care while not
knowingly
harming the patient. To that end, I would like to
298
share some of my experiences with you.
I personally am asked on a
daily basis
what can patients do or take to control
pain for a
variety of medical conditions. I have been
advising my patients to minimize the use
of oral
prescription OTC medications and instead
to use the
024, which due to its purity, does no
harm to the
human body.
We also recommend that 024 be
applied with
massage therapies. This has provided pain relief
that has often lasted 6 to 8 hours. These results
have been very exciting. The patients
have been
using less of the aforementioned drugs
and saving
money in the process.
No serious adverse effects,
such as GI
bleed, hypertension, or cardiovascular
problems
have ever been reported. There is no interference
with other medications that are necessary
to
maintain the patient's health because 024
is all
natural.
It contains no binders, preservatives, or
additives. Diet, exercise and work control are
299
also stressed, but in the future, we must
strive to
enhance our body's natural responses to
pain and
healing by safe and effective methods.
I am also a patient with
diabetic
neuropathy. I use it on a twice daily basis. I
have had no pain since.
Thank you very much.
DR. WOOD: Thank you.
The next speaker is No. 32,
Grant Johnson.
MR. JOHNSON: Thank you.
I would like to
start off by saying there is a little bit
of a
logistical mistake. The presentation
packages will
be circulated at the end of the public
presentations.
My name is Grant Johnson. I am the
present Chief Operations Officer at Swiss
Medica,
the manufacturer of 024. It's a topical
pain relief
medication that competes against the
NSAIDs and the
COX-2 inhibitor class of drugs.
We are all very aware of the
huge
potential negative side effects when
certain
high-risk patients take NSAIDs and COX-2
medicine
300
for any length of time. At Swiss Medica, we have
compiled scientific evidence that
powerful topical
pain relievers, such as the 024, are as
effective
as many oral medications, but without the
side
effects, such as the bleeding ulcers,
high blood
pressure, or increased risks to the
heart.
These claims are supported by
three
European medical studies, one
American-based open
trial, and a recently completed Canadian
randomized, double-blind clinical study
over an
extended period of time.
Every one of these studies
demonstrates
that there was a 60 percent or greater
quantifiable
reduction in pain for those who suffer
from chronic
pain conditions.
The first study was conducted
five years
ago, the latest was concluded last
month. In your
presentation packet folders I have
included the
appropriate summaries and the five pages of
professional endorsements, and you will
also find
anecdotal feedback from pain sufferers
who switched
to the 024 after failing to find relief
from a wide
301
variety of pain medicine and magic
solutions,
particularly the NSAIDs and COX-2
inhibitors.
Consumers need to be better
advised by the
FDA, healthier eating choices, regular
exercise.
These are things that have worked for
centuries on
this planet. Does it make sense to allow
multibillion dollar companies to spend
tens of
millions of dollars to persuade consumers
to pop a
pill instead of making a healthy
lifestyle
decision?
I propose the FDA consider a
moratorium on
all direct-to-consumer advertising until
these
drugs have been properly studied, and as
of today,
no one has a straight and honest answer
to the
question how many have really died from
using these
pain pills.
Thank you.
DR. WOOD: Thank you.
The next speaker is No. 33,
Necole Kelly.
MS. KELLY: Hi. I
am here speaking for
the American Chronic Pain
Association. We want to
make sure that everyone here understands
that
302
chronic pain also destroys lives.
People who have chronic pain
fight to get
their pain validated, to keep their jobs,
to keep
their health insurance, to maintain their
homes and
their families.
For 25 years, the ACPA has
offered support
and taught pain management skills to
people with
pain, to help them live more normal
lives. Yet, in
spite of their best efforts, many of
these people
still need medications including COX-2
inhibitors
that come with both benefits and risks.
Imagine learning that one of the
tools you
need to live a normal life is not longer
available.
In recent weeks, we have received
hundreds of
letters and e-mails from people who have
told us
they have stopped taking their
medications because
they are afraid of heart attacks.
Others also have told us that
they would
rather live 10 years with manageable pain
than live
20 in agony. Some people are getting
their
medications from Canada because they
can't function
without it.
The ACPA is not a research
facility. We
can't speak to the science behind these
studies.
We can speak for people with pain. What these
303
people want and need is to share with
their doctors
the medical decisions that affect their
lives.
They need to know the risks of taking any
medications and weigh them against the
benefits, to
make intelligent personal treatment
decisions.
They need to retain the right to make
these
decisions for themselves.
People with pain need the FDA
to continue
helping the public to get the accurate
science-based information they need to
make good
decisions, but we ask you to look beyond
the
science and see the human face of pain.
Imagine just one person who
woke up today,
as every day, with intractable pain,
unable to
function, and ask yourself what is best
for that
individual. We hope your decision will make a
positive difference for that person.
DR. WOOD: Thank you.
The next speaker is No. 34,
Karen Kaiser.
304
Is she here? No.
All right. Then, let's go on to No. 35,
Robert Thibadeau.
DR. THIBADEAU: I am an experimental
research scientist in a nonmedical field
with no
financial interests in the medical
industry
whatsoever.
I have had rheumatoid arthritis
and
ankylosing spondylitis since 1973,
diagnosed by
blood tests in 1983. Vioxx saved my life. It acts
in an hour with no high or other
perceptual side
effects.
It is like aspirin for headaches, it just
makes the arthritis pain and stiffness go
away.
I am here solely to reinforce
the
probability of an experimental
confounding and ask
for public analysis and full disclosure.
The confounding. You don't exercise for
25 years and now you have no pain and
stiffness.
You run upstairs because you are amazed
you can.
Risk of heart attack or stroke goes through
the
roof, not for bad reasons, but for good
reasons.
Control. Since these are brief,
305
unpredictable episodes, electronic
monitor all
waking hours to see if patients show
brief,
spontaneous increases in aerobic physical
activity
over placebo controls. I have not seen this done
or even mentioned for control by any
study
available to be read by the public.
I predict mentally incompetent
people,
Alzheimer's, much more likely to show
this exertion
side effect. People physically debilitated by
joint damage should show less effect due
to
physically restricted mobility. Other predictions
are in my longer paper.
I ask the advisory group to
review for
this confounding and ask the FDA to
report the
findings and justifications out publicly.
Thank you.
DR. WOOD: Thank you.
The next speaker will be Lois
Humphrey,
No. 53.
No? Not here.
I beg your pardon, Glenn Eisen,
No. 36,
was 52.
MR. EISEN: Close enough.
I would like in
306
the interests of full disclosure to
acknowledge
that I have done research and consulted
with
Pfizer, AstraZeneca, and Given, and like
Dr.
Qureshi, they are barely covering my
expenses
today.
Next slide, please.
I would like to discuss the
fact that
there has been accumulating data over the
last
decade as far as gastrointestinal
toxicity that has
gone beyond the ligament of trique (ph)
to both the
small and large bowel.
This is an autopsy study from
the New
England Journal approximately 10 years
ago, which
showed a greater than 10-fold incidence
of
nonspecific ulcers in an autopsy study.
Next slide.
A case-control study of
hospitalized
patients who presented with upper and
lower GI
bleeding found that patients within a
week of
admission had equal use of NSAIDs whether
it was an
upper GI bleed or a lower GI bleed, and
this was
twice of the control population.
Next.
As a secondary analysis in the
VIGOR
trial, you can see from these bars that
there was
307
twice the risk of lower gastrointestinal
bleeding
for naproxen as compared to rofecoxib.
Next slide.
In another analysis from the
CLASS study,
showed that in an FDA-mandated outcome,
having a
greater than 10 percent drop or a drop in
hemoglobin of greater than 2 grams per
deciliter,
there was double the risk of dropping the
blood
count in both diclofenac and ibuprofen as
compared
to celecoxib.
If we remove patients who have
had overt
bleeding, the trend continues.
Next slide.
So, because of this, we
developed a study
to show proof of principle for small
bowel damage,
and the combination of a nonspecific
NSAID with a
proton pump inhibitor should be
associated with a
rate of small bowel mucosal break that is
significantly higher than the rate for placebo
or
308
COX-2 selective agent.
Next slide, please.
We have already shown this.
Next slide. Dr. Qureshi showed some nice
pictures.
Next slide.
This was a double-blind,
randomized trial
where healthy volunteers had a two-week
run-in
period, were randomized after a baseline
capsule,
which was normal, and then were given 1
of 3
treatment arms.
Next slide.
The primary endpoint was the
mean number
of small bowel mucosal breaks, and as you
can see,
naproxen with a PPI had 10 times the
number of
mucosal breaks as compared to celecoxib.
Next slide.
The secondary endpoint showed
that there
was 55 percent incidence of small bowel
mucosal
breaks for combination therapy as
compared to 16
percent for celecoxib.
Next slide.
So, in conclusion, as in the
upper GI
tract, inhibition of COX-1 by naproxen,
and not
celecoxib, translated into significantly
different
309
rates of mucosal injury in the small
bowel, and
these findings extend the original
COX-1-sparing
hypothesis beyond the upper GI tract and
into the
small bowel.
Next slide.
You can read it because I am
out of time.
Thank you.
DR. WOOD: Thanks.
The next speaker is Susan
Winckler. Is
she here?
Yes, No. 37.
MS. WINCKLER: I am here representing the
American Pharmacist Association, and we
did not
receive funding to participate in today's
meeting.
The views I am presenting are solely
those of the
Association and its membership.
We are here because the safety
profile of
COX-2 selective NSAIDs has recently come
into
question.
Some have suggested that these drugs are
too risky to be marketed, but a
consideration often
310
lost in comments and debates, such as
this, is the
reality that no drug has zero risk.
Every medication has benefits
and risks,
and those risks increase exponentially
when the
products are used inappropriately.
Unfortunately, patients have
lost access
to several medications because the health
care
system failed to appropriately manage
risk.
Patients should not lose access to these
products
because of the health care system's
failure to
reduce risk.
If the agency determines that
the
benefit-risk profile is insufficient for
these
products to remain on the market, that
assessment
must consider the responsibility of
health care
professionals and patients in making
medications
work.
By collaborating, pharmacists,
physicians,
and patients can mitigate some level of
risk if we
focus on identifying potential risks and
determining systematically how best to manage
those
risks.
There are a few things that can
help us
with that risk management. First, is to increase
the reporting of adverse events by
pharmacists and
311
other health care professionals and to
continue to
encourage that reporting.
Providing pharmacists with
complete
information about the patients would also
improve
our ability to manage potential risks.
When products are identified as
having a
risk or requiring more attention, access
to a more
complete medical history would allow
pharmacists to
help assure that at-risk patients do not
take
medications that could exacerbate such a
condition.
If the agency determines that
there is a
need for special oversight of COX-2
inhibitors or
other NSAIDs, we urge the FDA and product
sponsors
to involve pharmacists in both the
development and
implementation of any risk management
program.
Please avoid the misperception
that only
these products present a risk to patients
when, in
reality, every medication has benefits
and risks.
Thank you.
DR. WOOD: Thank you.
The next speaker is No. 38,
Virginia Ladd.
MS. LADD: My association is paying for my
travel.
Good afternoon. My name is Virginia Ladd.
I am President of the American Autoimmune
Related
312
Diseases Association. We are a nonprofit health
organization representing patients living
with
autoimmune diseases, which include
rheumatoid
arthritis, lupus, scleroderma, and over
80 other
disorders sharing similar complications
as the
result of the body's attack on itself.
Autoimmune disorders are
serious chronic
and disabling conditions that often
present with
constitutional symptoms of joint and
muscle pain,
widespread inflammation, and fatigue.
We ask that the agency and its
advisory
committee respectfully consider the
critical role
of
patient and physician dialogue in conducting
risk-benefit analysis of any therapy at
the level
where it belongs - with the individual
patient
rather than a diverse clinical population
as a
313
whole.
We believe that patients should
have
access to as broad an array of
essentially safe and
effective therapies as possible, with
informed
labeling, providing the means by which the
provider
and the patient can consider treatment
options.
For many patients, the remote
and even
more common risk of a serious acute
adverse event
is, and would be, overweighed by the
benefit of
maintaining or regaining freedom from
pain,
mobility, and independence.
Since there has not been a new
drug
approved specifically for the use of most
autoimmune disorders in the last 40
years, it is
necessary that clinical reliance on
off-label use
of existing anti-inflammatories and
immune-modulating drugs.
In particular, the COX-2
inhibitors have
contributed to the improved life quality
of many
autoimmune patients to which I have personally
spoken.
Without COX-2 inhibitors, many autoimmune
patients with sensitivities to other
NSAIDs would
314
be relegated to the use of low-dose
corticosteroids
with therapy for the treatment of their
debilitating symptoms, and as you are
aware, such
therapies carry--.
DR. WOOD: Thank you very much.
The next speaker is No. 39,
Paola
Patrignani.
MS. PATRIGNANI: I am Paola Patrignani,
University of (inaudible) Italy. I am Professor of
Pharmacology. I am in the field for 20 years.
This slide compares the
therapeutic plasma
concentrations of cyclooxygenase
inhibitors,
reported in pink, with the concentrations
of the
different drugs inhibiting by 80 percent
the
activity of platelet COX-1, a biomarker
of
gastrointestinal toxicity, shown in panel
A, and
monocyte COX-2, a biomarker of efficacy,
shown in
panel B, as determined in the whole blood
assay
that I developed. This is in vitro,
reported in
blue.
It should be pointed out that
80 percent
inhibition of COX-2 is associated with
clinical
315
efficacy.
Ibuprofen and naproxen
therapeutic
concentrations are proper to inhibit more
than 80
percent platelet COX-1 and monocyte COX-2. Thus,
these two drugs have similar
pharmacodynamic traits
and they should be placed in the same
box.
Differently, therapeutic
concentrations of
COX-2 inhibitors are from 4 to 200-fold
lower than
those inhibiting platelet COX-1 by 80
percent, thus
demonstrating a variable impact on COX-1
depending
on the dose and selectivity.
The impact of COX-2 inhibitors
on monocyte
COX-2 is shown in panel B.
The therapeutic plasma
concentrations of
nemesulide, rofecoxib and etoricoxib are
proper to
inhibit more than 80 percent COX-2.
Diclofenac and lumiracoxib
plasma
concentrations are several fold higher
than those
inhibiting by 80 percent COX-2 while
celecoxib and
valdecoxib plasma concentrations are 2-
to 4-fold
lower.
In summary, ibuprofen and
naproxen have
316
similar pharmacodynamic features towards
COX
isoforms, so they have to be in the same
class.
Diclofenac and celecoxib have
superimposable pharmacodynamic traits,
but they are
given at not comparable doses.
Lumiracoxib 440 mg is an
overshooting
dose.
Next slide, please.
This slide is very interesting
because I
compared, I gave different drugs,
lumiracoxib,
rofecoxib, celecoxib, ibuprofen, naproxen
to
healthy subjects or patients, and I
compared the
inhibitory effect on COX-1 and COX-2 and
the
synthesis of prostacyclin.
The most interesting part of
the slide is
that all the other coxibs gave a similar
inhibitory
effect of prostacyclin. Also, the other
important--.
DR. WOOD: Thank you very much.
The next speaker will be No.
40, Betsy
Chaney.
MS. CHANEY: Good afternoon. I am Betsy
317
Chaney.
I am a Celebrex user. I took
Vioxx
before.
I am here to say would you all
pick up
your elbow and whack your funny bone and
feel that
pain that stops you in your tracks from
doing what
you are doing. All you want to do is say a bad
word.
Well, I have cracked vertebras
in my neck,
and without Celebrex, I start to lose the
feeling
in my hand, and I can't grasp a paper, I
can't hold
onto something, I can't do things around
my house.
I am concerned that you all
will take my
ability away to make a decision with my
physician,
my family, and my friends, to make an
advised
decision to take COX-2 inhibitors.
There is a lot of people here
for profit,
for many things, whether it be the drug
companies
or the lawyers, or whoever, but my issue
is please
don't take this medication that works so
well for
me.
I can't take other medication
because I am
taking two Nexium and a Xantac
today. That is
318
maxed out on the stomach medication. They looked
inside and said it looks like Barrett's
esophagus.
I have GERD and you all know I have
NSAID, NSAID,
NSAID.
I could name 100 of them, but those names
don't matter.
What matters is that I retain
the right to
make a decision, with my doctors and my
family, to
continue taking this medication even if
there are
risks.
I am willing for my quality of
life to
take those risks, and I thank you all
very much for
watching over us.
Thank you.
DR. WOOD: Thank you very much.
The next speaker is No. 41,
David
Peterson.
Is he here? No.
Then, the let's go on to No.
42, Jack
Klippel.
DR. KLIPPEL: Thank you, Mr.
Chairman.
The Arthritis Foundation
represents and is
the voice of millions of Americans with
arthritis.
Our constituency is keenly interested and
is a
319
major stakeholder in the discussions
being held
today.
They seek clear answers from
us, you,
their doctors, industry leaders, and
regulatory
authorities about the role of COX-2
inhibitors and
other NSAIDs in the treatment of their
arthritis.
The Arthritis Foundation
believes there
are two main factors that must be
considered in
these discussions of these drugs and
similar
discussions about other medications in
the future.
First, there must be a more
balanced
discussion about the benefits, as well as
the risks
for these medications. Recent attention of COX-2
inhibitors and NSAIDs have focused almost
exclusively on one particular risk,
cardiovascular
disease, with little mention of other
risks
associated with these drugs, or more
importantly,
the benefits of this class of drugs.
Numerous studies have documented
that
COX-2 inhibitors and other NSAIDs relieve
pain and
inflammation which has benefited millions
of people
with arthritis. Many have found COX-2 inhibitors
320
to provide greater pain relief than other
medications. For some, COX-2 inhibitors have
controlled pain when nothing else has
worked.
They would ask you the
question, whether
their public health was made better or
worse by the
decision to withdraw Vioxx. Their greatest concern
and risk is not about side effects of
drugs, but
that they live with arthritis.
Second, is the central role of
informed
patient choice in allowing patients with
arthritis
to make their own decisions about
treatment. We
believe that patients should be able to
choose for
themselves whether or not the benefits of
a
particular medication or treatment
outweigh the
risks.
Full disclosure of these
benefits, side
effects, and risks, and discussion with
the
patient's doctor--
DR. WOOD: Thank you.
The next speaker is labeled as
No. 43.
Kathy Pinkert. Is she here?
No.
Then, No. 44, Carol Spitz.
MS. SPITZ: Hi. My
name is Carol Spitz
and my travel expenses have been paid
for.
I have severe
osteoarthritis. I have had
321
a knee replacement, shoulder replacement,
and three
back surgeries.
Bextra has allowed me to be
able to
function, some of the normal things that
people
take for granted like walking and
dressing. I
couldn't even do that before.
I am unable to take Motrin and
Naprosyn
and aspirin due to anaphylactic
reactions. Other
NSAIDs have given me adverse reaction of
my
stomach, and that's it.
Thank you.
DR. WOOD: Thank you.
The next speaker is No. 45,
Eileen
Lacijan.
MS. LACIJAN: Good afternoon. My name is
Eileen Lacijan. I am grateful for the opportunity
to be here today to speak to you about my
experience with COX-2 inhibitors.
I would like to advise the
committee that
322
I do not have any financial relationship
with the
sponsor, product, or competitors. I am here today
representing myself on the advice of my
cardiologist.
I am 57 years old and reside in
Arnold,
Maryland.
I am a registered nurse and the
Executive Director of a Hospice Program
in
Maryland.
I have osteoarthritis of the
basal thumb
joints of my hands. I was first prescribed Vioxx
in March of 2000. My rheumatologist changed my
prescription to Celebrex in June of the
same year.
I then took Celebrex for the next four
years until
July of 2004. Following a flare-up, the Celebrex
was no longer effective and I was
prescribed Bextra
in July of 2004.
I have never smoked. I don't drink
alcohol.
I don't have diabetes or any family
history of heart disease. I have never
had high
blood pressure. I exercise regularly. I am not
overweight, and I have always maintained
a health
diet.
On the evening of August 12,
2004, I
survived a myocardial infarction. A cardiac
catheterization, which was performed the
following
323
day
at GW Hospital, revealed no blockages.
My
heart attack was thought to be caused by
a coronary
vasospasm, which affected the left
anterior
descending coronary artery and initially
resulted
in a moderately large amount of heart
damage.
I received excellent cardiac
care and was
able to return to work full time a month
after my
heart attack. I continue to work out at
cardiac
rehab several mornings a week before
work. I thank
God every day that I am alive and have
the love and
support of my family and friends. However, I still
have many unanswered questions about the
cause of
my heart attack as does my cardiologist.
DR. WOOD: Thank you very much.
The next speaker is No. 46,
Gloria
Barthelnes.
MS. BARTHELNES: I am Gloria Barthelnes
and I am from South Grafton,
Massachusetts.
When I was in my 30s, I was
having
324
problems with my legs and my neck, didn't
figure
what it was. I figured it would go away. Finally,
in 1984, I was living on a second floor
apartment
and I was carrying my grandson who was 10
months
old up the stairs, got halfway up the
stairs and
couldn't finish, I had to sit down the
pain was so
bad.
I had contacted the doctor and
had me go
through several tests. Finally, he recommended a
rheumatologist. They tried several
medications on
me, it didn't work. Then, finally, he had given me
Vioxx.
It was such a relief that I was able to go
to work without any pain, without any
problems.
To go to work, I had to travel
like 37
miles one way, and sometimes there was a
lot of
traffic, and just to sit in the traffic
was the
hardest thing to do.
I have the arthritis in my
neck, lower
back, and in my legs. When they took the Vioxx
away, I panicked and I tried using just
the
over-the-counter medication. It didn't work. So,
finally, I had called the rheumatologist
and I
325
said, "Can you help me?"
So, he put me on Bextra. I am hoping that
you people can help me, and not take
these
medications away.
Thank you for your time.
DR. WOOD: Thank you very much.
The next scheduled speaker is
No. 47,
Rebecca Dachman. Is she here?
DR. DACHMAN: Hello.
My name is Dr.
Rebecca Dachman. I am an occupational medical
physician, and I also have significant
experience
in clinical trial design.
There were a number of thoughts
that came
to my mind as I read in the papers about
what was
going on with the COX-2 inhibitors. One of them,
as an occupational medicine physician,
there are
many people who only respond to COX-2
inhibitors,
and that makes a difference between
working and not
working for them, which has significant
effects
both on disability and ultimately on
their health,
because nonworking, sedentary people are
a setup
for cardiovascular disease, as well.
As a clinical trialist, looking
at the
data, I know I was surprised that I
didn't get more
subgroup analysis of those who ended up
having the
326
cardiovascular events, whether there were
more
diabetics in that group or whether there
were any
other ancillary factors that one could
tell that
would identify them, and I think that is
important.
I also think that vis-a-vis
drugs, we have
to put it all in context, all drugs do
have ADRs.
Birth control pills are as extensively
used as
nonsteroidals and anti-arthritic drugs,
and they
all do cause increase in thrombotic
events, yet, we
haven't taken them off the market either.
I think we have to remind
ourselves of
that and what it means is not that they
won't have
events, but knowing about them and
knowing how to
subgroup the people in who those events
occur.
I think from the FDA stance,
they have to
develop registries post licensure, so
that for the
first two years, you get all the adverse
events
that occur, and that is what is being
done in
Britain.
DR. WOOD: Thank you very much.
The next speaker is No. 48,
Barrett
Collins.
Not here.
No. 49, Cynthia Lee. Not here.
No. 50, Robert Humphrey.
No. 51, Michael Paranzino.
327
MR. PARANZINO: I am Mike Paranzino. I am
here on behalf of Psoriasis Cure Now, a
patient
advocacy group. We have no financial conflict. We
receive no pharmaceutical industry
funding or
funding from their trial lawyer
opponents.
I am here to represent the 6
1/2 million
Americans with psoriasis, more than a
million of
those who have psoriatic arthritis, and
many of
those psoriatic arthritis patients take
NSAIDs
and/or the COX-2s.
Our written statement is on the
FDA
website.
It is available at psoriasiscurenow.org,
and there are some copies in the press
room. Our
central point there was that absent a
scientific
consensus against these drugs, that they
continue
to be available so that patients can
decide, with
328
their physicians, if in their own
particular set of
circumstances, the benefits outweigh the
possible
risks.
But in the remaining time, I
want to make
a
different point and I am amazed that in the last
50 people no one has made, and that is,
that in
some of the rhetoric surrounding some of
the
critics of FDA, some of the critics of
the
pharmaceutical industry, we are hearing
even some
buzz in Congress, that somehow the drug
approval
process is broken, and we think that is
false.
Patients need expeditious
approval of
medications, and there are many still in
clinical
trials that need to get approved, and we
are
concerned that the FDA may become timid
or gun-shy
and flinch about approving those drugs
that are
coming down the pipeline that millions of
Americans
with disease desperately need.
Where it does appear--and I am just a lay
guy, I am lay person, liberal arts
guy--but where
it does appear we need work is in
postmarketing
monitoring, post-FDA approval, that is
where we
329
need long-term monitoring.
We can't wait 20 years to get
long-term
studies before drugs are approved, but
when that
data does become available, it does
appear that the
ball is being dropped on a lot of sides
in adding
that information to the mix.
So, please, keep approving the
drugs. We
need new treatment options, and I thank
you.
DR. WOOD: Thank you.
The next speaker is Dr.
Lawrence Goldkind.
DR. GOLDKIND: I would ask that I go over
20 or 30 seconds, I could use some of the
time that
some others didn't use.
DR. WOOD: No, you get two minutes. Good
try.
DR. GOLDKIND: That's the Chair's
prerogative, I understand.
DR. WOOD: Good try.
DR. GOLDKIND: From 2001 to 2003, I was
the Deputy Division Director of the
Anti-Inflammatory and Analgesic Drug Products
Section at the FDA.
Over the past decade, there has
been an
evolution of what is considered feasible
in the
realm of clinical trials. Drugs, such as the
330
statins, beta blockers, ACE inhibitors
have been
developed to reduce mortality from
cardiovascular
disease.
Demonstration of these benefits requires
large and multi-year study.
Risk-benefit analyses are not
so hard when
there is superiority in an outcome of
death, and
placebo control, which is really add-on
to standard
care, is ethical and feasible.
What is unique about the COX-2
story is
that the indication is pain relief,
chronic in the
case of arthritis, but the perceived
value was a
safety advantage compared to NSAIDs,
which were
known to have substantial risks that were
reflected
in the labeling.
In fact, everybody here knows
that NSAIDs
have been the poster child for problem
drugs for
over a decade. So, it seemed obvious that
large
outcome studies would adequately test the
hypothesis of superiority of safety.
The concept of a large simple
trial sounds
simple, but, in fact, is not. We are now learning
the limits of outcome studies. At the time that
VIGOR and CLASS were done, they were the
longest
and largest trials by an order of magnitude
of
NSAIDs.
331
We can now say they were
imperfect and
lessons can be learned. One, therapeutic,
super-therapeutic doses are not the best
choice.
They promote off-label usage, and you
cannot
extrapolate well back to the therapeutic
dose
levels.
Single comparator trials, when
there are
many standards of care available,
likewise is hard
to interpret and put into a context of
therapies.
Allowing the duration and size
to be
driven by a single prespecified safety
endpoint
does not provide robust evidence
necessarily of
overall superiority, and yet it is
impossible to
power a study for unexpected or as yet
uncharacterized safety problems.
Even today, the term
"cardiovascular
332
outcome study" is bantered about as
if it were
cookbook simple. Well-known cardiologists
have
stated that the obvious population for
study is the
high risk patient--
DR. WOOD: Thank you very much.
The next speaker is Louis
Humphrey, No.
53. Is she here? No.
Then, we will go through the
ones that
didn't respond to our call earlier.
Rakesh Wahi, No. 2. Erika Umberger, No.
24. Gilbert Ross, No. 29. David Peterson, No. 41.
Barrett Collins, No. 48. Cynthia Lee, No. 49.
Robert Humphrey, No. 50. Lois Humphrey, No. 53.
In the absence of them, we will
take
somebody off the wait list, who is Yvonne
Shira.
Is she here? Yes.
DR. SHIRA: Hi. My
name is Yvonne Shira.
I am a practicing rheumatologist, and
while I have
worked with all of the companies
mentioned here,
and many others, doing clinical trials
and as a
consultant, I am here today representing
myself. I
paid for this trip myself.
I am representing my patients
and I hope
most of the rheumatologists who are
seeing patients
day by day.
333
I ask the committee to consider that
quality of life issues are as important
as length
of issues to many of our patients. When Vioxx was
removed from the market, a number of my
patients
refused to discontinue the drug despite
its risk,
because they deemed the quality of life
benefit to
be greater than the risk.
One patient said to me
regarding its
removal, "Dr. Shira, they just don't
understand how
much we suffer."
So, I ask that you do not take away
choices unless there is compelling
evidence that
the coxibs are substantially less safe
than the
available alternative NSAIDs.
The data you have presented so
far does
not suggest this, but that rather the
traditional
NSAIDs have not been sufficiently
scrutinized in
long-term trials.
Remember that real life data is
more
334
consequential than theories no matter how good they
sound.
Rheumatologists have always been aware of
the cardiovascular effects of all NSAIDs.
That is
why most of us monitor patients at high
risk by
having them come back within a week or so
for blood
pressure monitoring.
The problem has been that we
have accepted
blood pressure increases that we thought
were
insufficient, that in light of new
cardiovascular
information, may actually have hit
long-term
consequences.
It is likely, I believe, that
all NSAIDs
have cardiovascular risk, but they have
not all
been studied equally.
Please don't away our patients'
choices
without compelling evidence that the
alternatives
are truly safer.
Thank you.
DR. WOOD: Thanks very much. That was the
last speaker in the public hearing. I am grateful
to all of you for sharing your views with
us. I am
sure they will be helpful to the
committee.
We are going to go straight
back to the
program, and Dr. Villalba.
MS. MALONE: Excuse me, Dr. Wood.
335
DR. WOOD: Sure, yes.
MS. MALONE: I am Leona Malone. I am the
patient representative on the
program. I just
wanted to tell the people who did give
testimony
that--and this is not facetious at
all--that I
literally do feel your pain, and I think
that
everyone here is here because we are
aware of the
pain and the situation that you are in,
and no one
here is taking it lightly.
I know how much trouble
especially for the
patients it was to get here, to sit here,
to
listen, and to get up to speak, and I
applaud you
for that.
I just want you to be assured and to be
confident that all of us here will take
it
seriously and give a voice to everything
that you
have said.
Thank you.
DR. WOOD: Thank you, Leona. That was
helpful.
Are we ready, Dr. Villalba?
DR. VILLALBA: I am ready.
DR. WOOD: Okay, let's go.
Lumiracoxib
FDA Presentation
Lourdes Villalba, M.D.
336
DR. VILLALBA: I am going to talk about
the cardiovascular safety of
lumiracoxib. I want
to make some points before I am going to
show the
data, and this is that my talk is
restricted to
cardiovascular safety.
I will start again. I am going to talk
about cardiovascular safety only in
TARGET. So,
this is a very focused presentation, and
I am not
going to discuss any other aspects of
safety, such
as hepatotoxicity, I am not going to
discuss
efficacy, so I would urge you not to jump
into
conclusions regarding the risk-benefits
of
lumiracoxib without having all the data
on hand.
I am going to TARGET. I hope you remember
everything that was presented before
lunch, because
I don't want to repeat everything. We know TARGET
337
was a large study, 52 weeks, 18,000
patients with
osteoarthritis, that had two sub-studies,
one
comparing lumiracoxib and naproxen, the
other,
lumiracoxib and ibuprofen.
About 25 percent of patients
were on
low-dose aspirin, and they were two
identically
designed studies although there was a
little less
exposure in the second study, in the
lumiracoxib
and ibuprofen study, and there was some
imbalance,
slight imbalance in the cardiovascular
risk factors
between these two studies.
I want to point out that the
dose of
lumiracoxib that was used was 400 mg
daily and that
this dose has been mentioned before, that
it is 4
times the recommended dose, however, that
the
effectiveness of this dose has not been
demonstrated to the FDA's satisfaction
yet.
So, we don't know exactly what
this dose
means. Initially, it was thought to be
twice the
recommended dose, now the sponsor is
pursuing the
100 mg dose. So, again, this is hard to draw
conclusion from this dose into what is
going to be
338
in the final dose.
Regarding the cardiovascular
safety, I
want to point out that the primary
endpoint here
was confirmed and probable APTC
endpoint. For
example, Merck used only confirmed
events. So, you
cannot really cross-compare the numbers
here to the
other trials at Merck.
So, it includes cardiovascular
and unknown
cause of death, myocardial infarction,
clinical or
silent, and stroke, hemorrhagic or
ischemic.
Again, this specifically includes silent
myocardial
infarction, which was not particularly
specified in
the Merck definition. And then there were other
variables, they were looking at
everything.
So, here we have the same
disposition of
the slides that I showed yesterday. Here, you have
the name of the study, the drugs used
lumiracoxib,
naproxen, ibuprofen, the number of
patients
randomized in this row. Before, I didn't have it
up here, but now I have the patient years
of
exposure.
As you see, there is a little less
exposure of this study, but not that
different.
Here we go to the APTC
events. We have 40
events with lumiracoxib and 27 on
naproxen as
compared to 19 on lumiracoxib and 23 on
ibuprofen.
339
So, the first thing that I think stands
out is that
there is a different number in the total
number of
events, and particularly the number of
events on
lumiracoxib is half in the study 2332
than in study
0117.
If you go through the different
rows, the
difference is driven by the non-fatal MI
here in
the lumiracoxib as compared to
naproxen. This
number, as I mentioned, includes silent
MI.
Here, we have in this column
the number of
events and the rate expressed in 100
years of
exposure, 100 patient years of
exposure. This is a
different way of presenting the data than
the
sponsor presented.
Here, in this column, we have
the relative
risk, which is the overall risk of
lumiracoxib
versus naproxen, and I did not include
the
confidence intervals here basically
because it
would make the slide so busy, but also
there were
340
not statistically significant difference
in any way
you looked here.
So, again, if you look at the
relative
risk here of all the events to be
increased. For
lumiracoxib, it is increased, lumiracoxib
compared
to naproxen particularly driven by the
number of
non-fatal myocardial infarctions.
This is the Kaplan-Meier plot
with the
time to events information, with the percentage
of
patients with events, and here time and
date. As
you see, there is a separation between
lumiracoxib
and naproxen, that it starts early,
before day 50,
and seems to have a constant overall risk
here.
However, if you remember, for example, in
VIGOR we have the separation after a
month, but if
you think about APPROVe, the separation
wasn't
until after 18 months. So, this is only a year, so
we didn't get into what we saw with
APPROVe yet
here.
These are the numbers for 2332,
the number
of confirmed and probable APTC events,
and here you
see that the numbers look pretty much the
same.
341
The relative risks are all around 1 or
below 1 for
all categories.
Here, we have the Kaplan-Meier
plot and
they look pretty close here. This is lumiracoxib
in red and this is ibuprofen.
Now, if we put the two lines together,
we
see that lumiracoxib in study 0117 was up
here, and
lumiracoxib in study 2332 was down here
with
ibuprofen and naproxen in the middle.
So, I think that is very
difficult to
interpret anything from this study,
because
lumiracoxib look like two different
products in two
sub-studies within the same study.
This slide shows the difference
in the
number of events by aspirin use. I am not going to
give the relative rates, et cetera, but
it is just
to show you the numbers, how if you look
in the
lumiracoxib/naproxen sub-study, again,
the number
we said was driven by the number of
non-fatal MIs,
the non-fatal MIs among the non-aspirin
users,
because if you look at the number of
aspirin users,
the number is the same, 6 and 6.
This has to do with the size,
because only
25 percent of the patients were on
low-dose
aspirin, so this may have something to do
with
342
power, but again I think it is unclear
what the
role of aspirin is here, may be
protecting, that is
possible, but what I am concerned about
is that the
use of aspirin, if you have a substantial
number of
patients on aspirin in a trial that is
evaluating
cardiovascular safety, actually, that may
blur a
little bit the results.
Here, in 2332, we see that in the
non-aspirin users, there is no
difference, and if
you look at the aspirin users, actually,
there is a
trend that goes, that the situation was
on
ibuprofen users who also use aspirin, and
this
trend is consistent with that hypothesis
that
actually ibuprofen depleted the
anti-platelet
effects of aspirin.
This is just to show the number
of
non-fatal myocardial infarctions in the
first
study, the lumiracoxib/naproxen. Here, we have all
patients.
The number was 18 versus 10. In
the
343
non-aspirin population 10 versus 4, in
the low-dose
aspirin population 8 versus 6, and here
you have
the relative risks.
This is taken from the paper in
the Lancet
by Farkouh, et al.
Again, we see a signal here of
lumiracoxib
and naproxen, but this signal seems to
start
earlier than what we have seen before.
So, in conclusion, we cannot
draw
definitive conclusions regarding the
COX-2
selective class effect. If anything, I think that
this is consistent with what we have been
discussing during the last two days, and
that this
seems to be a class effect.
We don't know that selectivity
is a
continuous variable, so different NSAIDs
have
different degrees of selectivity, and
they are
associated with different cardiovascular
risks, and
the same with the different so-called
coxibs, but I
never like that name, because to me they
always
were NSAIDs.
But anyway, I think that this
adds some
344
information to the puzzle that we need to
put
together and decide what to do with this
class of
agents.
I know that this was only one year.
Now,
we are expecting to see longer studies
than one
year now, this is up to a year, which at
that time
seemed to be a long time, but now that we
look at
it, we think, okay, we would like to see
what
happened in the next two years.
These included patients, some
of the
patients had increased cardiovascular
risk as they
were using low-dose aspirin, however,
this was a
study only in patients with
osteoarthritis, it did
not include patients with rheumatoid
arthritis, and
we know that rheumatoid arthritis is
associated
with higher cardiovascular risk than
osteoarthritis.
So, that may have something to
do with the
findings, although we did see the
findings really
in the naproxen sub-study.
Again, I am not clear as to the
role of
aspirin here. Regarding blood pressure, for
rofecoxib I think that blood pressure is
an
345
important factor. I am not saying it's the whole
explanation, but I think that is an
important role.
However, here, I am not showing
any data,
but if you remember the data presented by
the
sponsor, ibuprofen affected blood
pressure more
than what lumiracoxib did. Actually,
ibuprofen
affected blood pressure more than what
naproxen
did.
It was like a 2.7 change in mean blood
pressure for ibuprofen. It was a 1.4 change in
mean blood pressure for naproxen.
So, here, we see the
association. There
is not a big increase in blood pressure,
but we are
still seeing the signal. Again, we didn't have
placebo here, so we don't know how these
were
compared to placebo.
Another thing that I want to
mention is
that lumiracoxib is structurally related
to
diclofenac, and we don't know how
diclofenac would
compare to lumiracoxib in this case.
This is it.
DR. WOOD: No back-up slides, good.
We are going to take a break
and we are
346
going to be back here and start at five
past 3:00,
and then we will start with the
discussion of the
presentations of the two previous drugs,
and then
we will go to the general questions after
we have
dealt with that.
So, we will come back at five
past 3:00
and start with the discussion of the
Merck
presentation and go on to this one
second.
(Recess.)
Committee Questions to the
Speakers
DR. WOOD: We have three tasks that we
need to get through this afternoon, so
pace
yourselves as you think about that,
colleagues.
We have got to deal with the
questions and
the issues that came up from the last two
sets of
presentations. We need to have Dr.
Furberg address
the Pfizer issues that he raised
yesterday and give
Pfizer the chance to respond to that, and
we will
come back to that in a second.
The third we need to do is
start to
address the questions that the FDA
prepared for us.
So,
there are three tasks we need to get through.
347
It is just after five past 3:00, and we
need to get
started on that.
Let's begin with the questions
for the
speakers on etoricoxib.
Oh, Dr. Hennekens first.
DR. HENNEKENS: In the 1970s, I was in
Oxford with Richard Peto. I had the privilege to
help him put together the APT
Collaboration. We
prespecified non-fatal MI, non-fatal
stroke, and
all vascular deaths as the combined
endpoint. We
specifically excluded silent MIs in the
first cycle
in '88 and the second with Rory Collins
leading in
'93, and the third with Colin Baigent,
now called
the ATT.
So, Merck, in my view, has used
the
correct APT now ATT definition. It is Novartis and
the FDA that are at variance with what
the APT
definition.
I had a question for the FDA
presenter.
One of the things Peto told me is if you
torture
the data enough, they certainly will
confess, but
with that as a background, the
lumiracoxib
348
comparison versus ibuprofen is 0.76,
against
naproxen it's 1.46, and the conclusion is
that the
drug is behaving differently in the two
studies.
Well, the alternative
hypothesis based on
the evidence we have seen so far is that
there may
be a protective effect of naproxen and
perhaps some
harm of the shorter acting NSAIDs, a
hypothesis
supported by the basic science showing
some
deleterious actions of all the NSAIDs,
but this
potential beneficial effect on platelets
of the
longer acting NSAIDs.
So, I think it may not be
necessarily true
that we need to conclude that this drug
is behaving
differently in two studies with two very
different
comparators.
DR. VILLALBA: My conclusion was that I
really don't know what to make of it, and
that is
why I need the opinion of other people
here.
The conclusion really was that
this
probably a class effect, this is a very
heterogeneous class, and you have all the
degrees
of selectivity there. So, that is what we need to
349
determine.
DR. WOOD: We have got Dr. Stephanie
Crawford.
DR. CRAWFORD: Thank you.
I would like to
ask Dr. Sean Curtis to please come to the
microphone if you are in the room.
Dr. Curtis, this morning you
stated that
in
global markets, Merck is currently revising its
labeling for etoricoxib to address new
safety
information relative to the safety of
selective
COX-2 inhibitors, so I am intrigued. In what
manner, specifically, what is the sponsor
stating
in its revised labeling worldwide on the
safety of
this product?
DR. CURTIS: We participated in the
European referral. It has been basically a
referral process for all the COX-2
inhibitors, and
that is actually just wrapping up, as you
know. I,
of course, have been here, but I am aware
of now
that there has now been wording for the
label that
talks--and this is basically class
labeling in
terms of contraindications--but I think
really what
350
it boils down to, you know, we have been
informed
from the CHMP that there will now be a
classwide
contraindication for all coxibs related
to
congestive heart failure.
It was previously classed as 3
and 4, it
has been extended to Classes 2 through
4. In
addition, there will be contraindications
in
patients with established ischemic heart
disease
and/or cerebrovascular disease, so that
will be
class contraindication, class labeling.
In addition, for Arcoxia or
etoricoxib,
there will be contraindication in
patients with
hypertension whose blood pressure has not
been
adequately controlled.
So, that is obviously new
information as
of today, and that is, in essence, what I
mean by
working with the regulators, based on new
and
evolving information, to come up with
product
labeling that accurately and adequately
reflects
current knowledge.
DR. WOOD: I think she was asking
you--which I suspect is going to be the
committee's
351
focus the rest of the afternoon for both
the
sponsors, for the committee at least to
decide what
the committee would need to see before
they approve
new drugs like this--I think what Dr.
Crawford was
asking was what were the studies you were
proposing
to do to do that. Is that right, Dr. Crawford?
DR. CURTIS: Could you restate the
question?
I couldn't hear you.
DR. WOOD: I think the question was what
studies were you proposing to do, that you thought
would help get this drug approved in the
future.
DR. CURTIS: As I reviewed through my
presentation, we feel the underlying
safety
information that is most relevant to
ensure that we
are all comfortable with the safe and
effective use
of the drug, is to proceed with the
studies that I
outlined this morning, namely, EDGE II
and MEDAL,
which are, as I reviewed, opportunity to
assess the
long-term safety of the compound in contrast to
traditional care, namely, diclofenac.
I reviewed the reasons why we
chose
diclofenac. There is pluses and minuses of the
352
comparators, but that is our primary
method to
further assess the compound at this point
in time.
DR. WOOD: Put on slide 31 again, would
you.
That was the slide that showed the relative
potency on the COX-1 and COX-2.
Basically, I think Dr.
FitzGerald said
earlier that he saw this as rofecoxib
lite or
something. So, given that you presumably wouldn't
have expected to see a difference between
your new
drug and rofecoxib, it seems like you picked the
next best thing to do as your comparator.
Naproxen is up there higher up,
and you
picked the one that was closest to
rofecoxib to
make your comparator, so the chances of
seeing a
difference seemed to me extraordinarily
small, and
I am not sure what that will teach us.
DR. CURTIS: Could we go to slide 1115,
please. The slide that I just showed as
part of the
core presentation was the weighted mean
average. I
did also want to point out that
diclofenac here,
what is plotted here is again at steady
state and a
percent inhibition from baseline again of
a COX-1
353
assay looking at platelet, thromboxane,
B2.
This is a plot of inhibition
both at peak
and at trough of the exposure in the
blood. You
see diclofenac at trough has about 60
percent
inhibition of thromboxane, but at peak,
achieves
levels that are close to 90 percent, so
there is
some variability in the degree of
thromboxane
inhibition throughout the dosing
interval.
I went through the reasons
why. I showed
some clinical data, too, that did suggest
that at
least from a GI tract perspective, which,
of
course, is ultimately one of the key
safety
endpoints, that there is a way to
differentiate
diclofenac from other NSAIDs--excuse
me--from what
we consider COX-2 selective inhibitors.
I showed you data with
valdecoxib and
rofecoxib. In thinking about other
comparator
choices, there are limitations to the use
of the
other NSAIDs that I reviewed, and I think
fundamentally one needs to keep in mind
that
diclofenac at this point is, in essence,
probably
the NSAID used most worldwide currently.
So, you know, in acknowledgment
of the
limitations of choosing any single
individual
comparator, and in acknowledging some of
the
354
limitations that were reviewed perhaps in
the
TARGET study even, where if you do start
to do
sub-studies, you do run the risk of
showing
different estimates even with one
comparator, even
with the same compound.
We felt that doing a large
study of the
magnitude that I described for MEDAL
against one
comparator, and I reviewed the reasons
why we chose
diclofenac, was as reasonable a choice
given all
the alternatives.
DR. WOOD: Garret, are you still here?
Maybe the question to him is supposing
that study
turns out with no difference, are you
going to hear
from him that he doesn't believe that
tells you
anything because it is just another COX-2
selective
drug, is that what we are going to hear,
Garret?
DR. FITZGERALD: I would take a slight
different tack. We have heard the words
"continuous variables" used
quite a lot, and I
355
think it is a continuum from as one
extreme, very
selective, very long-lived drugs, going
through
shorter lived, less selective drugs
through to very
non-selective drugs.
I would guess that the ease of
detection
and the size of signal would move across
that
spectrum from being very large to being
very small
or undetectable.
So, I won't reiterate the
reasons why. I
think diclofenac resembles remarkably
Celebrex with
respect to selectivity, and I would view
this trial
as actually a very useful trial,
beginning to
address for us information that we need
to know. I
would cast it as a within COX-2 selective
trial in
that respect.
It is like we have a surrogate
for
Celebrex.
We saw a lot of little trials with many
flaws in the blood pressure arena
yesterday,
setting up Celebrex against rofecoxib
with
arguments about timing of dosing, and so
on.
Well, here the rubber meets the
road. We
actually addressed the question of
whether a
356
commonly used, relatively selective drug,
diclofenac, stacks up in a way that
segregates from
a longer lived, much more selective drug,
etoricoxib, so I think it does provide
useful
information in that regard, although I
might cast
the reasons for why I think it is useful
in a
slightly different way.
DR. WOOD: Any other questions? Dr.
D'Agostino.
DR. D'AGOSTINO: This is both for Joel and
Sean.
You raised the question, Sean,
about doing
a non-inferiority study, and I am
wondering--that
certainly will be a discussion that we
will
have--and I am wondering if you realized
the
implications of that.
When you look at, for example,
slide 44,
in your presentation, and you look at the
EDGE
study, was the EDGE study a non-inferiority
trial?
DR. CURTIS: I actually wanted to clarify
something that Dr. Schiffenbauer
mentioned. So,
the answer is no. The non-inferiority criteria
357
that I identified in the presentation is
based on
cardiovascular safety data accrued from
three
studies:
EDGE, EDGE II, and MEDAL. So, the
cardiovascular non-inferiority criteria
is to be
applied to the minimum 635 confirmed
thrombotic
events that will accrue from three
studies.
DR. D'AGOSTINO: From the three studies,
not one at a time.
DR. CURTIS: That's correct, but I am
providing you data that is coming available,
and
EDGE had finished, and it is an important
piece of
information.
DR. D'AGOSTINO: That is comforting in
terms of what is possible, but just to
point out
that on that result, that would not be
very
positive for you if you did the 1.3. You
would
actually, in that case, say that the
comparator
could be better. I mean that would be a conclusion
in that study.
I don't want to go into the
details of
that, but one has to be very careful when
they go
the non-inferiority route, and we will
talk about
358
that more. This slide frightened me a bit.
The other is if you do go the
non-inferiority route, what about the
inclusion of
the aspirin individuals, it probably
won't be a
constant hazard in the sub-groups, but
what will
happen then with your
non-inferiority. This was
raised by Joel, and I would like an
answer. I
would love to hear what your answer is.
DR. CURTIS: Aspirin, of course, it is
hard to win with that, and I will tell
you why. On
the one hand, you want to include
patients with a
range of baseline risk, and certainly one
criticism
of some of the studies is that patients
with
cardiovascular risk have not been
included in these
studies.
Both us and the FDA felt it was
important,
as the data provided to included patients
with
baseline cardiovascular risk, but, of
course, those
patients should be on aspirin.
So, we, of course, allow
patients to be on
aspirin as per clinical guidelines. As I
mentioned, we expect about 30 percent of
the total
359
patient cohort in the cardiovascular
analysis will
be on aspirin.
But I want to be clear, the
primary
analysis will be based on all patients whether they
are on aspirin or not.
DR. D'AGOSTINO: But are you going to be
assuming in the 1.3 that the hazard ratio
will be
the same within that sub-group, but just
that it
will be a different level of absolute
risk? We
will talk about those things, but those
are serious
implications.
I would have to have a study
design where
the very first thing you do is say, well,
gee, I
couldn't do what I set out to do, I have to
look at
subsets, namely, I have to get rid of the
aspirin
users because they are confounding
things.
Was that the concern that the
FDA is
having?
DR. SCHIFFENBAUER: Yes, as I expressed,
in the non-inferiority design where we
don't have
the placebo background, this would be a
maneuver to
make the two groups look more
similar. I mean if
360
you extrapolate it to 60 percent or 80
percent
aspirin use, I think the two groups would
look
almost identical, so you would end up
having to
look at subsets, that is true.
DR. WOOD: Dr. Abramson.
DR. ABRAMSON: Yes, I have a question for
Dr. Villalba
DR. WOOD: Can we just deal with the first
presentation first.
DR. ABRAMSON: I am sorry.
Then, I will
wait.
DR. WOOD:
Dr. Gibofsky.
DR. GIBOFSKY: Dr. Curtis, I have a
concern about the selective emphasis of
data being
presented in seeming replicate
trials. If we go to
slide 10, for example, and again in slide
46, you
commented that etoricoxib was superior to
naproxen
in one of two pivotal studies, but
similar in the
other study, and based on that one study,
you have
used the term "superiority" at
least twice in your
presentation.
I guess I am kind of wondering,
if you did
361
a back of the envelope calculation, like
Dr.
Fleming did yesterday afternoon when we
were
discussing two polyp trials, one of which
we gave
more focus to I think than the other,
would you
still be able to make this claim of
superiority
based on the meta-analysis with both
trials?
DR. CURTIS: My point in highlighting the
efficacy data was, of course, not to talk
about a
claim of superiority. The purpose was to provide
data that provides you and all of us an
opportunity
to look at both the risks and the
benefits of the
compounds, and the data in RA were
compelling, and
I fully disclosed results from both
studies.
Furthermore, the data, these
really were
the first studies that we are aware of
that showed
a statistically significant
difference. So, my
point was again in the context of an overall
risk-benefit assessment, to claim--to not
claim,
but to show the data for this compound at
the doses
that were studied provide a level of
efficacy that
certainly should be part of the
consideration.
I certainly would not be
claiming any sort
362
of label claim or anything like that,
because we
are not here to talk about such things.
DR. GIBOFSKY: I take your point, but
specifically, if you combine the second
study with
the first, would you use the word
"superior" to
naproxen, or would you use the word
"equivalent" to
naproxen?
DR. CURTIS:
I can only talk about a
clinical study within the context of that
clinical
study where patients were randomized
evenly between
treatment arms. I think it would be speculative to
talk about combining the results.
DR. WOOD: Dr. Shafer.
DR. SHAFER: If you can go to slide 19,
and we see here that once again the
confidence
bounds around the three groups do not
really
justify the breaking out of naproxen, it
would
appear to me, as a separate group.
Now, go to slide 44. Once again you have
broken out naproxen as a separate group
although it
is not clear that the confidence bounds
would
support that either.
So, we have a pattern where you are
constantly seeing a worse outcome
compared to
naproxen, and similar to rofecoxib, where
the same
363
signal came up, you asked, I think, or
you mean to
imply to us that naproxen is
intrinsically
different, but we have heard multiple
experts over
the course of the last day and a half
tell us that
they don't believe that naproxen is
intrinsically
different.
We have seen observational
trials in which
there may be a modest effect of naproxen,
but
certainly nothing of the magnitude to
explain a
1.5, 1.7 risk relative to naproxen that
you have
seen in your data, and even the sponsors
themselves, Roche and Bayer, in their
presentations, felt that naproxen did not
have the
cardio-protective effects that you have
attributed
to it.
So, first, I am disturbed that
your
primary analysis isn't versus NSAID
comparisons,
all NSAIDs, and then as a subgroup, you
compare
naproxen out. Instead, you pull naproxen out and
364
ask us, I mean the implication almost is
that we
should dismiss it, because it's naproxen,
and then
look at everything else. It concerns me that we
aren't primarily looking at all NSAIDs as
the
comparison group.
Secondly, at this point in
time, do you
truly believe that naproxen and the
postulated
cardio-protective benefits of naproxen
truly
explain the difference that you are
seeing, and
that we are not actually seeing a very
solid signal
for intrinsic increased cardiovascular
toxicity
with the COX-2 antagonists?
DR. WOOD: And while you are answering
that question, tell us why the right
study wouldn't
be to do a naproxen with omeprazole
versus your
drug.
I mean you obviously believe naproxen beats
the drug, right? And the only advantage of the
drug over naproxen is a GI benefit.
Supposing omeprazole gave you
the GI
benefit and you still had the
cardiovascular
benefit, wouldn't that be the optimal
therapy? And
why, given your data here, did you choose
to go
365
with the drug that has less benefit than
naproxen?
I still don't understand that.
DR. CURTIS: I am going to answer your
second question first. Naproxen clearly is a very
effective drug, however, as we heard
repeatedly
today, patients have different responses
to
therapies. Again, the reason people with arthritis
take drugs is so they can have some
relief. Not
everybody responds to naproxen.
So, I think naproxen clearly is
a very
logical choice for many patients, but
there are
going to be patients who do not respond to
naproxen, and when you factor in GI risk,
adding a
PPI certainly would appear to likely to
mitigate
some of the risk, but you are still going
to be
left with patients who don't respond to
naproxen,
who still are going to have a residual GI
risk, and
we have seen data that suggests even when
you add a
coxib or a PPI to an NSAID, there is
still room to
improve from a GI safety perspective.
So, I think that as a
therapeutic option,
selective COX-2 inhibitors, including
etoricoxib,
366
still have a role. As to why we chose not to use
naproxen as the comparator in our outcome
study, I
reviewed the reasons. We have now seen qualitative
differences in cardiovascular outcomes
against
naproxen with three different COX-2
selective
inhibitors: rofecoxib, etoricoxib, and
lumiracoxib.
We felt that doing an outcome
study
against naproxen, we would likely
replicate that
observation again. We felt it was
important to
accrue additional data against another
traditional
NSAID that was used widely around the
world to get
a more firm estimate of what the
cardiovascular
risk looked like against another NSAID.
DR. WOOD: You looked at that data. You
saw that naproxen beats your drug. So, you decided
to pick one that didn't look like it
would--because
it is as selective as your drug is--and
you are
going to come back with that data and say
wow, it
doesn't produce any cardiovascular signal
because
it's the same as diclofenac. That doesn't make any
sense.
DR. CURTIS: Again, I think it is
important to remember that the
qualitative
differences that were observed against
naproxen
367
were being seen at the same time that no
differences were being observed with
non-naproxen
NSAIDs, and in a time frame like a year
for which a
difference from placebo with COX-2
inhibitors has
not been appreciated.
So, I think all that data, to me,
continues to say that there is something
different
about naproxen. I can't quantify that, I don't
think the data allow that, but there
clearly
appears to be something different about
comparisons
to naproxen to the other NSAIDs.
DR. WOOD: I understand that, but the
issue that has changed since hour initial
studies
with naproxen is that we now have three
randomized
trials against placebo in which placebo
beat the
drug.
So, using an active comparator that you have
chosen to match in terms of
cardiovascular adverse
events, etoricoxib, isn't acceptable in
terms of
showing that the drug doesn't have an
effect on
368
cardiovascular mortality or morbidity.
It might have been acceptable
in the days
when you believed that naproxen was
beneficial and
that that was the total explanation, but
by your
own admission, you don't believe that
anymore.
DR. CURTIS: So, if I understand the
question, you are asking why we are not
doing a
large outcome study against naproxen?
DR. WOOD: I guess I am asking you what
you are going to learn from the
diclofenac study.
You are certainly not going to be able to
say that
this drug does not produce cardiovascular
problems
given that you have deliberately chosen a
drug that
looks as similar to etoricoxib as you can
get, and
from your earlier studies, namely, this
one, you
have seen that it does produce a
difference with
naproxen, and it doesn't appear to
produce a
difference with this, and it has got a
very similar
pharmacology.
So, if you can imagine an
imputed placebo
arm here, and given what we know about
placebo, you
would predict that this drug would do
worse than
369
placebo, and you won't be able to exclude
that from
the study you are designing.
DR. CURTIS: The data that are emerging,
that we have all seen the APPROVe data,
we have all
seen the difference against celecoxib in
the APC
study, to us, that suggests a class
effect. I have
showed you our placebo-controlled data
for
etoricoxib, it's very limited.
With that being said, the class
effect
related to COX-2 inhibition, we would
presume
extends to etoricoxib, and, to us, the
real
clinical question is in patients who
require
chronic treatment, what is the
cardiovascular
safety against a standard of care, and
for the
reasons I reviewed, we chose diclofenac.
DR. WOOD: So, let me be sure I
understand. So, we are going into this study
saying that we know and believe that the
drug will
produce a cardiovascular signal, we are
just trying
to work out if it's better or worse than
diclofenac.
DR. CURTIS: No, I think what we are
370
asking is--
DR. WOOD: Well, that is what you just
said, isn't it?
DR. CURTIS: If I could rephrase what I
said, I think what we are saying is we
are
suggesting there is a class effect, and
we are not
sure how big the class is, and we feel
that the
MEDAL study will help provide information
to
address that specific question, whether
cardiovascular safety for selective COX-2
inhibitor
is the same or different than that of a
traditional
NSAID, one that is the most widely used
NSAID
around the world currently.
DR. WOOD: Okay.
Dr. Bathon.
DR. BATHON: I was going to say much the
same thing. I have the same concerns about this
especially since naproxen is the most
widely
prescribed NSAID in the U.S. and the most
relevant
to our practice, whereas, diclofenac has
much more
hepatotoxicity especially in RA patients
where
methotrexate is co-administered.
So, I think it would have added a lot more
371
to our clinical practice management to
see another
big trial against naproxen rather than
diclofenac,
plus you could have added these results
to your
prior trials and had more power to assess
the
effect of naproxen versus etoricoxib with
all of
your trials combined, but now, since you
are using
diclofenac, you don't have that extra
power.
DR. WOOD: Dr. Reicin.
DR. REICIN: Let me just make one comment,
and as all you start to talk about
designing
clinical trials, I think you will see, as
many of
you know, it is quite difficult and you
cannot
answer every question in every study.
MEDAL was started over two
years ago, and
at that time there was no
placebo-controlled data
to suggest that COX-2 inhibitor was
different than
placebo.
Obviously, that has changed. The
studies
are fully enrolled and ongoing.
I can't disagree with you that
the idea of
doing a naproxen plus PPI study versus a
COX-2
inhibitor isn't a good idea and isn't an
important
question.
Unfortunately, we didn't design that
372
study, we designed this one, and I think,
as Garret
said, it will provide information about
how big the
class is.
While some of you may not be
using
diclofenac, it is the most widely used
NSAID in the
world, and therefore, I think it will
provide
beneficial safety data to see what a
selective
COX-2 inhibitor looks like versus a
non-selective
inhibitor albeit not as non-selective as
naproxen.
DR. WOOD: Thanks.
Dr. Dworkin.
DR. DWORKIN: Yes, a simple question. You
said that the CPMP had come up with class
labeling,
but you neglected to tell us CPMP defined
the
class.
Is it all NSAIDs, is it COX-2 inhibitors,
and if the latter, what drugs were
included in that
subclass?
DR. CURTIS: I am going to give my
understanding as a clinician who has been
here for
the last 48 hours, but my understanding
it is
specific to what we consider the
selective COX-2
inhibitors - celecoxib and etoricoxib,
and that
373
that is how the class is being defined
currently.
DR. DWORKIN: So, those two drugs, but
not, for example, Meloxicam.
DR. CURTIS: Dr. Erb, would you like to
comment on any additional agents?
DR. ERB: Yes, Dennis Erb from Regulatory
Affairs.
The CHMP is included in the
class, what we
have been referring to today as the
coxibs,
lumiracoxib, celecoxib, and etoricoxib,
and
valdecoxib.
DR. WOOD: Dr. Platt.
DR. PLATT: More on the history of the
choice of comparators. Dr. Schiffenbauer, could
you tell us more about the conversations
between
the agency and the sponsor around the
choice of
comparators?
Your comments and the materials
you
presented to us suggested that you had
reservations
about that choice.
DR. SCHIFFENBAUER: Yes, we had extensive
discussions with the sponsor. At the time we
374
appreciated the difficulties doing a
placebo-controlled trial, but we had
requested--and
I can't quote you whether it was
additional
comparators or comparator--but we had
recommended
strongly that additional agents be
studied to get a
better handle on the true cardiovascular
risk.
DR. PLATT: Was there discussion about
naproxen as a comparator?
DR. SCHIFFENBAUER: Not specifically other
than to mention that we recommended
additional
comparators.
DR. WOOD: Dr. Farrar.
DR. FARRAR: One of the things that
strikes me about all of the studies that
we have
been looking at, and perhaps most in the
comparison
of studies that we are still waiting for
some data
on, namely, APC and CPAC, is the
difference in the
underlying risks between some of these
different
comparisons.
I noticed that in your
particular study,
the cardiovascular risk, you felt that 38
percent--I think that was the
number--that in your
375
slide you had 38 percent at an increased
risk of
cardiovascular disease with 24 percent on
aspirin
and 10 percent of them as being diabetic.
I just wondered if you could
comment on
what the mix of the MEDAL study is likely
to be or
is.
I mean you certainly would have the data at
this point.
DR. CURTIS: Yes.
1103, please. The
MEDAL study population is, as I
mentioned, both OA
and RA patients, so approximately 75
percent of the
patients have OA and about a quarter have
RA. What
is represented here are the risk factors
for the
cohort, the entire cohort, and it is not
dissimilar
to what I highlighted for the EDGE study.
These are basically baseline
medical
diagnoses at the time of entry into the
study, so
about half have hypertension, which is a
little
higher than the EDGE study, which was
about 40
percent, as you see here, the individual
cardiac
risk factors, and this 12 percent of
history, that
is documented atherosclerotic
cardiovascular
disease.
The 38 percent that I quoted for the EDGE
376
study was patients with this or to
primary risk
factors.
So, that percentage, if I were
to
calculate that percentage for this study,
it would
probably be a little higher than EDGE,
probably
about 40, 42 percent. So, these are the patients
in MEDAL.
DR. FARRAR: If I could just follow up and
ask actually Garret FitzGerald, whether
he has any
comments on the relative risk of patients
who have
either high or low cardiovascular risk
factors.
I mean we know from the study,
the CABG
study, that patients with very high risk
clearly
have a marked increased response to these
drugs,
and whether people who have cardiac risk
factors
are also in that category, or whether it
really is
restricted to sort of the release of
active agents
from the surgical procedure.
DR. FITZGERALD: Well, obviously, the
actual information we have relevant to
your very
important question is conjecture. What we know
mechanistically is that what we would
expect would
377
be the response to thrombogenic stimuli
would be
enhanced, as would the predisposition to
the other
cardiovascular adverse manifestations of
this
mechanism, namely, hypertension and
atherogenesis.
So, for example, if a
population was
enriched in patients with secondary
hyperaldosteronism, they would be more
prone, on
average, to exhibit hypertension in
response to an
NSAID or particularly a selective COX-2
inhibitor.
Similarly, if they were at
advanced risk
of hemostatic activation, they would be
prone to
the
thrombogenic complications, and I think with
the CABG patients, we had an extreme
phenotype of
excessive hemostatic activation.
Now, as we move away from that
extreme
through what we call
"heightened" cardiovascular
risk, there is probably a continuum of
predisposition that is a mix of
predisposition to
the various types of manifestation of
this
mechanism that could occur.
So, we have only crude
indicators
obviously, and to some extent, as I
talked about
378
yesterday, it's in the eye of the
beholder as to
what defines heightened cardiovascular
risk, but on
average, the group defined as having
higher
cardiovascular risk, for example, RA
compared to
OA, on average would be expected to show
a signal
easier than in a group with low
cardiovascular
risk.
I mean I would think with this
type of
study, we may have had a premonition of
the outcome
from the EDGE result. For example, if we think of
these two drugs as defining the limits of
a class,
just for fun, one could say like in the
EDGE
results, you wouldn't see a distinction
in the hard
GI endpoints or the hard cardiovascular
endpoints,
but what you might see a distinction in
is their
fringe surrogates, which might be easier
to pick
up, such as discontinuations because of
hypertension or discontinuations because of
GI side
effects, and that is actually what was
seen at the
two ends of the spectrum in the EDGE
result.
DR. WOOD: But we do know from the APPROVe
study that the point estimate, even in
the people
379
with no history of cardiovascular
disease, which
would be the only clinical measure we
could
reasonably use to distinguish that, it is
still
substantially greater than 1.
DR. FITZGERALD: Yes, I mean I did try to
raise the issue yesterday that how we
define
underlying clinical substrate is an
inexact
science, on the one hand, and on the
other, that
many other factors that we discussed
yesterday
could play into the likelihood of
manifestation of
risk at the individual level.
DR. WOOD: Steve.
DR. NISSEN: I want to maybe bring us back
to earth a minute and talk about the time
horizon
for such a trial. I feel compelled to point out
that we have got a lot of history in
cardiovascular
medicine of studying drugs for
atheroprotective
effects.
Those trials are typically not
one year or
two years or even three years, they are
typically
five-year studies, and in many of them,
let's take
a blockbuster class of drugs like the
statins.
Look at the CARE trial. The CARE trial,
the Kaplan-Meier curves didn't diverge at
all for
two years, and so now we have got a drug
here that
380
may be promoting atherogenesis, and so we
are going
to
say, well, we are going to have a 20-month mean
exposure, and if it doesn't produce a
problem,
then, there must not be a problem, and I
am not
sure that's right.
The problem we have is that
what has been
done here is the sample size has been
increased to
a large sample size in order to shorten
the
duration, but that may not be the same as
studying
a more modest size group of patient for
three or
four years.
It is assuming that the hazard
is constant
over time, and I am not so sure that it
is here.
If, in fact, Garret is right, and he has
been right
about a lot of things, that these drugs
are
potentially atherogenic, then, an
atherogenic
intervention may not produce an effect
for several
years.
So, how can you reassure us
here that a
381
20-month mean exposure is enough to allow
us to
move forward with a drug like this?
DR. CURTIS: I think what you are touching
on is--I am not going to disagree--what I
am going
to point out is the fact that I think
running an
arthritis study is perhaps different, and
I have
not designed outcome studies,
cardiovascular, other
than this--but to keep arthritis patients
in
studies is difficult, and that has to do
with the
treatment of the disease.
As the rheumatologists here can
speak to,
a traditional trial has 40 percent of the
patients
discontinuing after one year, and another
10 to 20
percent dropout rate every year
subsequent, so
there are significant practical
limitations to
keeping patients on study therapy into
the time
frame that you proposed, Dr. Nissen.
So, that is a practical
limitation to
running arthritis studies.
DR. NISSEN: I just would also point,
however, that the patients that we
studied
initially with these atheroprotective
drugs were
382
very high risk secondary prevention
patients.
These were not low risk people.
So, you are going to take a lower risk
population and you are going to look for
a signal
at a 20-month mean duration, and that
signal may
actually take longer to show up in a
lower risk
population.
So, I am troubled by how long
we have to
look for with a drug like this before we
really can
say there isn't a problem. People may take these
drugs for a decade. We heard that from people at
the microphone here.
So, these are some of the
things that
trouble me about the whole question.
DR. WOOD: I have got a whole list of
questions here, but I want to keep us
moving here.
So, are there any people who
have burning
questions that they want to torture Dr.
Curtis with
before we let him off? It has to be specific. We
will take Tom, we have not heard from you
yet.
DR. FLEMING: Burning?
DR. WOOD: Burning.
DR. FLEMING: There are two or three
issues I want to quickly review. You didn't
mention in EDGE the new ischemic heart
disease or
383
the heart failure, pulmonary edema,
cardiac
failure.
I think the FDA indicated in their
review, there was a 25-19, and a 14-6, so
basically
about a 30 percent relative increase and
a doubling
in those two, is that your understanding?
DR. CURTIS: The numbers, yes, Dr.
Schiffenbauer quoted, those are the
correct
results, and that information was in your
background package.
DR. FLEMING: And then very quickly, your
slide 19 and then your slide 25. On your slide 19,
do you have the analogous slide for the
APTC
results?
If you don't, my understanding is the
relative risks are less favorable than
this or more
unfavorable, depending on your
perspective.
They are 1.8, 0.87, and 2.72?
DR. CURTIS: That is correct, yes.
DR. FLEMING: So, essentially, we are
looking at with roughly a 3 to 2
randomization in
384
the aggregate, and the aggregation of
these events
here, we are looking at 43 versus 12, so
a pretty
substantial excess in the critical APTC
measures.
DR. CURTIS: Well, again, as you know, the
APT events in total are smaller than
these numbers,
so your confidence intervals around those
point
estimates are, in fact, much broader.
DR. FLEMING: But at 43 to 12, they are
certainly well outside of unity.
The last thing is slide
25. You give the
mortality results, but it is difficult to
really
see in this scale, but it appears that
the relative
risks are roughly in the range of 1.6
against
placebo, also 1.6 against naproxen, and
1.2, and
then in addition to that, it is also 1.33
in the
EDGE trial.
So, it looks as though when you
look in
terms of relative risks, that you are
looking at
about a 1.5 relative risk on mortality
across the
aggregate of these data.
DR. CURTIS: Yes, this slide shows the
rate with the confidence interval. I don't have
385
the relative risk.
DR. FLEMING: But those aren't relative
rates is my point.
DR. CURTIS: That's correct, these are
absolute rates here.
DR. WOOD: So, you are saying this stuff
doesn't look it's good for you. Anyone else who
has a burning question? Go ahead.
MS. MALONE: It's burning.
I would like a
simple answer. How much different--now, I heard
him call this like Vioxx lite, I believe
I heard
him say that--how different is this from
Vioxx, you
know, chemically, and do you see it as a
substitute
for people who are perhaps taking Vioxx?
DR. WOOD: I think we are talking about
diclofenac. It was the comparison to diclofenac
which had been referred to.
MS. MALONE: He also did a presentation on
etoricoxib. So, can he answer that?
DR. WOOD: You are asking me?
MS. MALONE: No, him.
Okay, I am sorry, I
thought you had said that about
etoricoxib.
DR. CURTIS: Can you clarify the question,
please?
MS. MALONE: I am just wondering how the
386
compound in etoricoxib compares to Vioxx.
DR. WOOD: You mean chemically?
MS. MALONE: Yes, but in simple terms.
DR. CURTIS: The human whole blood assay,
if that is your specific question, the
human whole
blood, which is sort of the gold
standard, that
shows a degree of COX-2 selectivity that
is greater
for this drug, but in the clinical dose
range,
etoricoxib, just like rofecoxib, just
celecoxib,
just valdecoxib, are selective for the COX-2
enzyme
in the clinical dose range, so in that
regard, they
are similar.
Does that answer your specific
question?
MS. MALONE: I am just wondering, you
know, I have heard people say that
Celebrex or
Vioxx was much more selective than
Celebrex and
Bextra, and where does this fit in, in
that scheme?
DR. CURTIS: Again using the human whole
blood biochemical assay, this drug would
be
387
considered more selective, but I think
the key
concept, at least for me as a clinician,
is that in
the dose range that these drugs are used,
they all
selectively inhibit the COX-2 enzyme and
do not
inhibit COX-1.
DR. WOOD: Let's move on to the next set
of presenters and let Dr. Curtis off the
hook.
Thank you very much.
Are there questions for the
Novartis
presenters from the committee? Some of the people
who are still waiting for the questions,
we will
begin with them if they want to start
with the
other ones. Dr. Abramson had one, I know, and we
punted.
DR. ABRAMSON: That was the TARGET
presentation by Dr. Villalba. I would like to just
throw slide 9 up, if we could, and follow
up on a
point that Dr. Hennekens made when we
started this
session.
In that slide, you combined the
two
component studies of TARGET and again
said that
lumiracoxib behaved differently in the
two studies,
388
but I think that is probably incorrect to
put up a
slide like that. It is like putting up a CLASS and
a VIGOR slide together, because these
were, as I
understand it, separate studies and
separate
populations.
That is the comment, but the
other
interpretation, as we heard, is that
lumiracoxib
performed less well than naproxen, maybe
because it
has a risk and maybe the naproxen has
some
protective effect, but was comparable
ibuprofen,
which again raises a question whether
ibuprofen has
some risk attached to it.
But my question is that you
then said that
you attributed these findings to a class
effect,
and since definitions are going to become
very
important for us going forward, I was
wondering if
you could tell us what you meant by a
class effect
and what you were referring to, is it the
class of
NSAIDs?
DR. VILLALBA: Yes, yes.
First of all,
this slide was made by the sponsor, we
didn't make
the Kaplan-Meier curve, so this was just
a
389
different way of presenting the
data. I don't
think it was in the background package
for you, and
I thought it was an interesting way of
looking at
it, raising the issue that precisely you
cannot
just combine the two studies, because the
sponsor
also has presented the data of the two
studies
combined, lumiracoxib with NSAID, and you
cannot
just combine these two studies, because
they are
different studies.
I agree with you, you cannot
cross-compare
even within the same study that had two
sub-studies, so we cannot compare to
other studies
that were done with different designs and
different
entry criteria, different endpoints, so
that was
the point of the slide.
Regarding the class effect I
mentioned, I
referred to the NSAID class effect. I think that
if there is an effect, it is for the
entire class,
and that is a very heterogeneous class
with
different degrees of selectivity within
the NSAID
class.
That is what I meant.
Actually, let me clarify. We also thought
390
that naproxen could be protective. I was seeing
these data at the same time that I was
reviewing
all the other rofecoxib studies, so I
guess you can
understand what our position was at this
time.
DR. WOOD: Dr. Furberg?
No? All right.
Dr. Bathon?
DR. BATHON: This was a question for Dr.
Matchaba.
I think there is an interesting
observation about the TARGET trial. Before we even
consider comparing lumiracoxib to the
NSAID
comparators, but just looking at the
baseline APTC
events in the two sub-studies of TARGET,
there is
an event rate of 0.43 percent in one
trial and 0.84
percent in the other trial, in the
lumiracoxib-treated individuals.
That is a 2-fold difference
although the
numbers are small. I am wondering if that could
have been contributing also to the
ultimate
difference between lumiracoxib and the
comparator
drugs.
Even though you used the same
inclusion
391
and exclusion criteria, could you tease
out any
differences in the two study populations that
were
enrolled into the studies that could have
explained
the baseline difference in events in the
lumiracoxib groups? I don't mean baseline, I mean
the accumulated events.
DR. MATCHABA: Thank you very much for the
question.
If I could have No. 8 and then could I
have CV No. 67.
As we discussed, the TARGET
study was a
combination of two studies. The only thing
identical about the studies is the design
of the
studies, but as I mentioned in the
discussion
today, that this study against naproxen
started
about 4 to 5 months before this study
against
ibuprofen, and that the centers that were
used for
this study were different centers even
within the
same country, and the staggering of the
recruitment
was to ensure that centers were not
recruiting for
the same study.
In some cases, countries that
participated
in one study did not participate in
another study.
Can I see the CV67, please.
We have also asked this
question to say
why are we seeing differences in the
rates of
392
cardiovascular events for the 1-1 study
versus the
ibuprofen sub-study. What we have done here is to
look is it a center effect, and there is
obviously
a lot of reasons, we don't have all the
answers or
explanations, but if you see for the
major
recruiting countries, Argentina, Germany,
and the
U.S., that the naproxen sub-study, in
terms of
rates of APTC events, were always higher
than for
the ibuprofen sub-study even in the same
country.
So, if you look at the
demographic data
that we also presented to you today,
where 25
percent of the patients in this study
were taking
low-dose aspirin, where we had a
difference of 14
percent versus 10 percent in high CV risk,
that
these populations are different in terms
of
baseline risk, and certainly that might
be an
explanation, it could be chance because
the
confidence intervals cross, but we don't
have all
the answers, but we think we have
different study
393
populations.
I might ask Dr. Michael Farkouh
to
elaborate on that because he was involved
in the
design of the study and he was the
primary author
for the TARGET cardiovascular paper.
DR. WOOD: Have we got the question
answered?
I think we have. Let's move
on. Dr.
Abramson, did we answer your question
already?
Okay.
Dr. Cryer.
DR. CRYER: Thank you.
I have been trying
to understand the differences in the
results
between the TARGET trial and previous
outcome
studies of COX-2 specific
inhibitors. One very
clear difference is in how the
definitions were
rendered.
One thing that concerns me is
that in the
lumiracoxib experience, both your CV and
GI events
are defined to include people that not
only had
definite MIs and definite GI events, but
also
included those people who had probable
events.
Typically, I am more used to
seeing trials
in which we are looking at fully
adjudicated
394
definite events. When I looked here, for
example,
at your CV events, and eliminate what you
call
silent MIs and look at just what would be
considered clinical MIs, there is an
apparent
3-fold increase with lumiracoxib for
clinical MIs
compared to NSAIDs, which dramatically
differs from
your other conclusion.
With respect to the GI events,
I think
that you actually studied a low GI risk
population.
We know that the relative risk of COX-2
specific
inhibitors to have a GI benefit is
greater in a
population that has low GI risk.
Specifically, you didn't
include anyone
who had had a previous history of a GI
bleed in the
last year, and greater than 50 percent of
your
patients were less than 64 years of ago.
So, my question to you then is,
have you
re-evaluated your data using more
conventionally
accepted criteria, for example, fully adjudicated
clinical events rather than include their
probable
events?
DR. MATCHABA: All the cardiovascular
395
endpoints, APTC, including silent MI, peripheral
events, deep vein thrombosis, pulmonary
embolism,
TIAs were all predefined and
prospectively
adjudicated blindly by an adjudication
committee
before the study started.
This includes the GI or ulcer
complications and PUBs with the different
definitions that have been used,
including
clinically evident bleeds, were also
predefined by
a gastrointestinal committee.
DR. CRYER: I understand it may be
predefined, but I am asking do you have
data if you
excluded the probable?
DR. MATCHABA: Yes.
Perhaps Dr. Farkouh
would like to comment.
DR. FARKOUH: Michael Farkouh from New
York University. Our blinded adjudication
committee, the definitions of probable or
definite
were purely on the basis of if we had
all-source
documentation versus our clinical
judgment of the
committee, which is many years of
experience. I
happen to be the most junior member. A probable
396
cardiovascular event really, in our mind,
was a
definite, that we just may not have had
all the
source documentation we needed, so it
really was
adjudicated as--probable was an element
or a degree
of definite is how I would put it.
DR. CRYER: With all due respect, I will
ask the question a third time. Do you have data
eliminating the subset of people who were
classified as probable, and looking only
specifically at those who you felt were
definite
events?
DR. FARKOUH: From our clinical
cardiovascular committee, we did not feel
there was
any distinction between the two of them,
so we did
not mandate that. To be a probable event, I think
any cardiologist that would be on this
committee or
anywhere else would have documented this
as an
event.
So, it is a degree of definitiveness.
We
did not mandate that.
DR. MATCHABA: If I can just add to that,
the answer is yes, and if you just look
at
confirmed cardiovascular events, the
analysis is
397
the same, and just to add, that for
silent MIs
besides what Dr. Farkouh has added in
terms of
prospective definition, there was a total
of 32
clinical MIs in TARGET, and there were 8
silent
MIs.
Of those 8 silent MIs, 5 of
them were in
NSAIDs and 3 on the lumiracoxib. When we look at
silent clinical MIs, we still see the
same trends
whether you compare the naproxen versus
lumiracoxib
with ibuprofen versus lumiracoxib.
DR. FARKOUH: There is a moving target
here.
The definition of MI has changed over the
last five to six years. We have a much more
enzymatic definition of MI which we have
adopted,
and
also the definition of silent MI has been
adopted into this modified anti-platelet
trial.
I agree with Dr. Hennekens that
it is not
part of the sharp definition, but rather
we were
encouraged due to the signal of MI that
has been
seen in this class of drugs that we
document silent
MIs, and this was adjudicated through a
blinded ECG
core laboratory run at the University of
398
Pennsylvania.
DR. WOOD: Dr. Fleming.
DR. FLEMING: Could we go to slide 33.
There, I think what you have tried to do
is capture
the aggregation of the favorable effects
on
reducing upper GI ulcer complication and
the
unfavorable effects on the APTC.
I guess my first thought is
that since you
didn't present the global data, I would
assume the
global data is your primary analysis, and
by my
crude calculation, the relative risk
reduction is
probably more towards 25 percent or so
rather than
the 41 percent that you are showing.
But I guess more to the point,
is it not
apples and oranges here as you are trying
to look
at
the aggregation of evidence?
The ulcerative complication
rate has been
reduced from 1 percent to 0.4 percent, so
we can
think of it in terms of per 1,000 people,
there are
about 7 cases that are prevented, and the
APTC is
increased from 0.57 percent to 0.84
percent, so for
1,000 people, there are 3 of those cases.
Isn't it a little fairer to
think of it in
that context? We have got per 1,000 people, 7 of
these ulcerative complications prevented,
and while
399
those are substantial events, is it not
true that
predominantly patients recover and don't
have
long-term sequelae, while you are
inducing 3 APTC
events that are CV-strokes or MIs that
have much
more long-term effects?
So, isn't that a fairer
question, and
while this picture makes it look like it
is a clear
positive, I would have thought the answer
is much
less clear, if not clearly negative.
DR. MATCHABA: Thank you.
It's a fair
question.
If we look at this combination of safety
data for the overall lumiracoxib compared
to
NSAIDs, the reduction in the overall
population is
35 percent. It was 25 percent in the naproxen
population overall, and it was not
significant.
DR. FLEMING: I am focusing on just the
slide you are giving, which is the slide
against
naproxen, so just to keep it simple in
the
comparison against naproxen.
DR. MATCHABA: Yes. I
think the first
comment we will make is that the comment
was made
in the VIGOR study that any events that
do occur in
terms of ulcer reduction and
complications are
negated just quantitatively by the
increase in
cardiovascular events.
400
I also made the comment that this
is
certainly not validated, but it is an
attempt on
our part that using this unvalidated
method for the
first time and prespecifying it and
stacking up the
primary endpoints, what does the picture
look like
relative to comparators in the same
study.
DR. WOOD: What Dr. Fleming is asking you,
that there is a qualitative difference--
DR. FLEMING: Apples and oranges, yes.
DR. WOOD: And a GI bleed is not the same
necessarily as a stroke. They don't compensate for
one another. That is not a criticism, it is just a
fact.
DR. MATCHABA: Yes, that is a valid point.
DR. WOOD: And I think that is what he is
saying, am I right?
DR. FLEMING: Correct.
DR. WOOD: Any other questions for the
sponsors?
Before anyone thinks of any, let's move
along.
DR. MATCHABA: Thank you very much.
DR. WOOD:
One of the things that we left
undone from yesterday was that Dr.
Furberg raised
some issues that he was unclear of some
differences
that he thought he saw in the Pfizer
briefing book
401
and from his calculations.
I charged him with meeting with
Pfizer and
trying to resolve these. Dr. Furberg, did that get
resolved?
DR. FURBERG: We met and I got some
clarification, but I continue to be
troubled.
DR. WOOD: So, the answer is no I guess.
Why don't we do this then.
DR. FURBERG: I think there are five
issues.
DR. WOOD: Why don't you tell us about the
issues and let's give Pfizer an
opportunity to
respond.
Curt, why don't you go through
the issues
as you see these.
DR. FURBERG: The first one related to the
number of trials included in the
integrated safety
analysis for the acute pain studies. There was in
one place mentioned that there were 18
trials, in
another place there were 20, and the
explanation
that was given was that the 18 trial
analyses
excluded 2 trials, the one using the
highest dose
of the drug, 60 mg--more than 60 mg a
day.
That doesn't satisfy me. If you are
looking at safety, the trials with the
highest dose
402
are the ones that I am primarily
interested in. I
think the company did the proper thing,
they
included information about that, but they
should
have included that in the pooled
analyses, as well,
and that would have changed the message
that you
take away from that summary table. So, that was
one issue.
DR. WOOD: Let me ask Pfizer, do you want
to respond to each one in turn, is that
the easiest
way?
DR. HARRIGAN: That would be fine with me.
DR. WOOD: Let's do that, then, we can see
what the issues are.
DR. HARRIGAN: Just one slide, slide D114,
please.
Ed Harrigan from Regulatory Affairs,
Pfizer.
What we have done with this
slide is
basically pulled the two paragraphs from
the
briefing document that Dr. Furberg was
describing.
In Section 3.3, anybody who has the
briefing
document and who downloaded it from the
web would
be able to find these on pages 55 and 76.
In Section 3.3, as Dr. Furberg
points out,
we integrated safety data from acute pain
studies,
18 of these studies, and as it says in
the
403
paragraph, they represented 4,087
patients treated
at a dose range of 20 to 60 mg total
daily dose.
Later, in Section 3.6, we
described 20
completed studies representing a larger
number of
patients treated with valdecoxib at a
dose range
greater than 20 mg total daily dose.
Now, the difference between
these two
404
paragraphs is largely due to the CABG
Study 035,
which is described in great detail, in
fact, six
pages devoted to the CABG studies in the
briefing
document.
It is a matter of opinion as to
whether
one should have pooled this data. If one had
pooled all the 80 mg data, then, one
might have
been accused of diluting the 80 mg
treatment effect
that was seen in the CABG 035 study. On the other
hand, the 035 study was presented by
itself with
full representation of the safety issues
in that
study, which have been discussed in great
detail
here in the committee, appropriately so.
I think that is probably the
end of
response to that point.
DR. FURBERG: The second issue is to the
mean consistency in the reported event
data.
Again, we are back to the same integrated
safety
analysis of the 18 studies, and Tables 19
and 20
indicate that there was a total of 4 to 6
MIs
depending on how you define them, whether
you
include sudden death in the report.
Well, separately, there were
data
presented on two of the trials that were
included
among the 18, and I just added up the
number of MIs
405
and I come up with the number 8 to 10
when I define
it as non-fatal MI and fatal CHD. So, you already
have a negative balance. What happened in the
remaining 16 trials?
The explanation that was given
was that in
the second CABG trial that got involved
in the
analyses, they subtracted the number of
events when
the patient was on the I.V. formulation
parecoxib.
I looked it up and it turned out to be
one case.
So, that doesn't explain the discrepancy,
so the
explanation that was given was not
satisfactory.
DR. HARRIGAN: Could I have slide D116,
please.
This is Table 20 in the
briefing document,
I can't give you the page number. So, as Dr.
Furberg points out, this is a table that
shows
placebo 2,468 and the 4,087 patients from
the
valdecoxib studies at doses of 20 to 60
mg. Three
myocardial infarctions in the valdecoxib
treatment
406
group.
Now, Table 22 is an
illustration, it is a
table titled from one of the tables,
there are
Tables 22 through 27 in the briefing
document,
which report on the adverse events in the
rest of
the studies described in that portion of
the
briefing document.
As Dr. Furberg points out, we
reported to
him earlier today that the myocardial
infarctions
that he saw in the general surgery study
and in the
two CABG studies, if they occurred to
parecoxib,
they were assigned to parecoxib. These
are trials
in which treatment with parecoxib took
place for a
certain number of days, and then patients
were
switched to valdecoxib.
If you assigned an event to
both
treatments, then, of course, you are
going through
tables until midnight, because they won't
add up.
You have to assign the event to one
treatment or
the other, they were appropriately
assigned to
parecoxib, and so they are not accounted
for in the
valdecoxib column.
A second reason for a difference is
that
the adverse events in the tables that Dr.
Furberg
was drawing them from are adjudicated
adverse
407
events.
So, these are events that were determined
according to prespecified criteria in
both of the
CABG trials and in the general surgery
trial.
So, aside from the parecoxib
confound, you
wouldn't expect those adverse events to
add up to
adverse events reported in a different
way. This
is frequently an issue in safety summary
documents.
There are a number of different ways to
record
adverse events.
You have serious adverse
events, you have
spontaneous adverse events reported to
marketed
drugs, you have adverse events recorded
in case
report forms in clinical trials. By presenting
them several different ways, you are sure
that you
are giving the entire picture, because
you don't
want to select one picture and be accused
of not
showing the other two, but you can be
guaranteed
the columns will not sum up.
DR. WOOD: But parecoxib is the pro-drug
408
for valdecoxib.
DR. HARRIGAN: It is.
DR. WOOD: So, as far as my body knows
when it gets parecoxib, it has got
valdecoxib.
DR. HARRIGAN: Two points.
One is that
the events are described in the briefing
document
as you see, but they are assigned to
parecoxib. I
don't know if you are suggesting that all
treatment
groups that receive parecoxib, all
patients that
receive parecoxib be transformed to
valdecoxib.
DR. WOOD: I guess the body transforms it
to valdecoxib.
DR. HARRIGAN: It would obscure the data
from the effects of parecoxib, which is
given by
different formulation. Some people consider that
significant, so I think to describe it
under
parecoxib is appropriate. To not put it under
valdecoxib is appropriate. The data is in the
briefing document, it is not hidden, it
is not
suppressed, it is clearly available in
the briefing
document.
The columns do not add up. We
think
there are good reasons why they do not
add up.
409
There are alternative ways to present
safety data.
We are happy to, and frequently do,
re-run safety
data and safety tables with different
algorithms
and different rules.
DR. FURBERG: The numbers just don't add
up.
DR. WOOD: Have you another point, as
well?
DR. FURBERG: No.
DR. WOOD: I suggest that we are not going
to resolve this this afternoon, so why
don't we
defer this to Dr. Temple and his staff to
resolve.
Is that fair, Bob?
DR. TEMPLE: Yes.
Curt agreed earlier
that he would write down exactly what the
concerns
are, and we, not me, will follow them up
and pin
down what is going on.
DR. HARRIGAN:
It is important to us that
members of this committee and the FDA,
and other
health agencies worldwide understand that
we do not
suppress safety data. We report safety data, we
report it in a number of different ways,
we do not
410
suppress safety data.
DR. FURBERG: But it would be much better
if you explained why you did it
differently and
present the data in one way in one table,
another
way in another table, the numbers should
add up if
you have information from two trials and
you have
more events than you have in the pooled
analysis of
18, that has to be explained.
I think there are some numbers that will
be hard to explain away.
DR. WOOD: I think we have got it that
there is still a bone of contention
here. Let's
move on to the three questions that we
were charged
with discussing this afternoon.
Dr. Gross, I think wanted to
make some
comments before we get to the first
question.
Committee Discussion
DR. GROSS: On the first question, I would
like to propose a construct to deal with
the issue
is the increase in cardiovascular risk a
class
effect.
My proposal is to say yes, it is, but the
degree of difference and the time of
difference
411
varies and is different enough that one
or more of
the drugs that we have discussed should
be
marketable with a precaution and/or
warning, and
one or more of the drugs we have
discussed should
not.
A reasonable analogy is
statins. As we
know, they all have potential for liver
toxicity
and myopathy. That is a class effect, but the
degree of this difference and the time
when it
occurs varies and is different enough
that one or
more of the drugs have been marketed with
a
precaution or warning, and one or more
have not.
Tomorrow, we will discuss
specifically the
recommendations on celecoxib, valdecoxib,
and
rofecoxib, but I thought I would start
off the
discussion with this question about a
class effect.
DR. WOOD: Okay.
Dr. Nissen.
DR. NISSEN: Did you mean class effect for
the COX-2s, or are you talking about NSAIDs,
as
well, because the question is asked for
both here.
So, I want to know which of those you
mean.
DR. WOOD: Let me make a suggestion. I
412
think we should start with COX-2s. The data we
have seen is by far the most convincing
for that.
Then, let's move on to any other issues.
DR. NISSEN: So, let me agree that is what
we are talking about then.
DR. WOOD:
Let's have a discussion around
the COX-2s first and whether the
available data
support a conclusion that cardiovascular
risk is a
class effect for all--
DR. GIBOFSKY: Could I just interject and
ask
then that we discuss it in the context of
patients with arthritis versus patients
with other
conditions?
DR. WOOD: Okay, that's fine, the
committee can do that, but remember we
are not
discussing the relative risk-benefit at
this point.
We are discussing whether there is an
effect, a
signal, in other words.
DR. GIBOFSKY: I understand, but I think
it is relevant to look at the populations
in which
the signal has been detected.
DR. WOOD: Do you want to comment on that
413
and save us going to back? Do you think that the
arthritis population will be likely to
have a lower
risk than the other populations?
DR. GIBOFSKY: I am merely saying that I
think that one looks at populations. As we have
heard, there is variability in the
population, and
just as we wouldn't automatically
extrapolate
efficacy data from one population to
another, I am
not certain we can automatically
extrapolate safety
data from one population to another, and
I think we
need to discuss it in the context of the
population
studied.
DR. WOOD: Any other comments on this
question?
DR. ILOWITE: I think you made the point
that this was merely a discussion of
safety, but I
think the way the proposal was worded,
there is
implications about cost-benefit with regards
to
whether they should be approved or
not. Could you
repeat the--
DR. WOOD: You have the question in front
of you.
DR. ILOWITE: The proposal.
DR. WOOD: Dr. Gross made a proposal, but
the question we have got in front of us
is to
414
discuss the available data supporting a
conclusion
of increased cardiovascular risk for
COX-2
selective nonsteroidals.
I think we need to discuss that
before we
get to risk-benefit frankly.
Dr. Nissen.
DR. NISSEN: I think your proposal is an
appropriate one and I would point out
that we have
at least one randomized trial for every
drug that
has been marketed in the class that shows
an
effect.
DR. WOOD: You mean a risk.
DR. NISSEN: That makes the grade in terms
of calling it a class effect, but I think
that
there is clearly evidence of a gradient
in risk,
and that gradient is not only by drug,
but also by
dose.
So, saying it is a class effect
means--let
me tell you what it means to me. It means that if
415
you give a high enough dose of one of
these drugs
to a risky enough patient, you can
produce an
increased risk of adverse cardiovascular
outcomes.
But it doesn't mean that a particular
dose in a
particular population is risky.
DR. WOOD: Dr. Fleming.
DR. FLEMING: I think there is a great
deal of data that is giving us a general
sense, but
there is an inadequate amount of
information to
really get at the specifics, and what I
mean by
that is certainly the indication, the
dose, the
duration of therapy, the nature of
ancillary care,
for example, aspirin use, these are all
factors
that obviously could influence the
answer.
The approach that I took was to
try to
summarize the essence of what I think we
have been
presented in the randomized trials, and I
focused
in particular on those that were the
major trials,
many of them looking at somewhat longer
term
exposure and longer term follow-up.
There are about 15, and just to
quickly
run through them, in the Vioxx setting,
there are
416
23,000 patients from four major
trials. Those
studies indicate something on the order
of 1.4 to
1.5 relative risk, and driven heavily by
VIGOR and
APPROVe, and neutralized somewhat by the
Alzheimer's 078-091 trial although that
trial had
surprisingly considerable excess deaths.
In the Bextra setting, the
Nessmeier 071
trial, the 035, and 069 studies give
about a 2 1/2
relative risk even though it is certainly
heavily
driven by this CABG setting.
In the Celebrex trials, the
CLASS, the
Alzheimer's 001, the APC, the PreSAP,
now, we know
there is the ADAPT, but we haven't been
shown that,
so I did the first four, and we are
looking about a
relative risk of 1.3, driven heavily by
the APC
trial and the 001 study, and neutralized
by the
CLASS study and the PreSAP that were more
neutral.
The etoricoxib, the EDGE trial,
and the
other three that we were presented give
us a
relative risk of about 1.625, and in the
lumiracoxib, it is about 1.18 relative
risk from
the TARGET trial.
Now, to put these into context, if we
were
trying to show--I am just going to give
your four
scenarios--a doubling. By the way, I am working
417
off a 1 percent background rate, and that
is just
about what these data show in the
aggregate, in
73,000 patients, about a 1 percent
aggregate rate
for the primary cardiovascular endpoint
of
cardiovascular death, stroke, and MI.
If you were trying to show a doubling, it
takes 88 events or about 5,000
people. If you were
trying to show a 50 percent increase,
it's 256
events, 20,000 people. If you are trying to show a
33 percent increase, it's 508 events,
40,000
people, and if you are trying to show
just a 20
percent relative increase, from 1 to 1.2,
it's
1,265 events or 115,000 people.
Where we are, if you ignore all
those
factors that I was arguing we can't
ignore because
the answer isn't the same, but if you put
all this
into a single pool, these add up to
something in
the neighborhood of a relative risk of
about 1.4 to
1.45.
So, essentially, what it would
have taken
to discern that is an aggregate data of
about
70,000, although the observed results
that you
would have to have, you would have to
have a study
on the order of about 5,000 people,
because an
observed result of 1.55 or 1.45 is
statistically
418
significant when you have about 6,000 or
7,000
people.
So, the point is when you look
at the
aggregate, we have substantial data to
say there is
conclusive evidence here in the aggregate
that
there is the cardiovascular risk.
Now, what can we say
individually? In the
Vioxx setting, where the risk is about
1.43, one
would need to have, with that observed
rate, you
would have needed to have data on about
6- to 8,000
people.
We have data on 23,000. That is why the
evidence there is very clear.
In the Bextra setting, we only
have data
on 3,000 in the Nessmeier 035 and 069
trials, but
the relative risk is 2.58, and for a
2.58, you need
less than 2,000 people, hence, that is
why it is
419
statistically significant in that setting
although
it is only in the CABG setting.
It is also in the etoricoxib
setting with
the relative risk of 1.625, we would have
needed
less than 5,000 people. We have 17,000, so it is
statistically significant in that
category.
In the lumiracoxib setting, we
have a
relative risk of 1.18. That would have taken over
40,000 people for that relative risk to
be
detectable. We only have 18,000. So, it is
suggestive of a modest or moderate excess,
but it
is not proven because of the smaller
sample size or
because of the smaller effect.
In the Celebrex, where it is
about 1.29,
it would have taken 20,000 people, if you
observed
that in 20,000, it would have been
marginally
significant. We observed it in approximately
12,000, so it is suggested, but not
established.
Now, a lot of this, this is
looking at
things in a first pass. It is suggestive that
there is something going on in all of
these cases,
but at very different levels is what at
least the
420
data show, but the data aren't conclusive
for us to
be able to say in a reliable way what is
the
indication, what is the dose, what is the
duration,
is it in aspirin, not in aspirin, but
globally,
there certainly is an effect that is
going on here,
and for three of these five, it seems to
be
conclusively established, and for the
other two,
more modest effects that are suggestive.
DR. WOOD: Thanks.
Dr. Shafer.
DR. SHAFER: I don't know that this will
help our discussions at all, but I think
one of the
things we need to address is what is a
COX-2. It
has been assumed, I think, that we are
going to go
with the company's definitions when they
way we
have a COX-2 drug, COX-2 selective, but,
in fact,
if you go to Warner's review in FSAAB
here, from
2004, we see that Meloxicam, Sulindac,
and as we
have heard, even diclofenac is
potentially
considered a COX-2.
Should we include these drugs
in the
discussions? We certainly won't have the
evidentiary evidence that we have.
DR. WOOD: Let's stick to the drugs for
which we have got evidence, otherwise, we
will be
here until midnight.
421
DR. WOOD: Charlie.
DR. HENNEKENS: I want to support the very
crucial statements of Steve and Tom
here. It does
appear to be a class effect, which varies
by drug
and by dose, but the magnitude of that
risk, which
I also estimate to be 1.4 to 1.5, is
lower than one
would have guessed based on the early
data-dependent stopping of some of these
trials,
based on the reported research, and the
media
coverage of all of this, so I think it is
important
to get Tom's quantitation and Steve's
caveats about
dose and drug and magnitude in there.
DR. WOOD: Dr. Abramson.
DR. ABRAMSON: I want to go back to this
issue of definitions because I don't
think that it
is that simple to say that the coxibs
that we are
talking about are the only drugs that we
need to
discuss.
The concept of diclofenac lite,
that
422
Garret proposed, or has stuck, but I
think that
term could be applied to Meloxicam,
nemesulide. I
think what we are stuck with is that
assuming there
is a class effect, we haven't excluded
the fact
that that class includes those other
COX-2
preferential drugs, that we might agree
that COX-2
inhibition is at fault here, and that is
giving
rise to some of these side effects, but
it isn't
precisely due only to those drugs.
Now, those drugs happen to have
done the
long-term studies, they have done them
frankly at
2X dose compared to their comparators,
and frankly,
when you look at the randomized clinical
trial
development program with a relatively few
placebo
arms, the drugs look relatively
comfortable.
So, if you conclude that there
is an
increased risk because of COX-2 effects
and
hypertension perhaps, then, I don't think
it's
really fair to restrict the discussion
simply to
those drugs that got marketed as coxibs,
the
randomized clinical trials, especially if
you do
agree that perhaps Naprosyn has a modest protective
423
effect, I think don't give a bye to these
other
drugs.
I think we have to look at
certainly the
case of celecoxib, that drug is relatively
comparable pharmacodynamically to the
other several
drugs, so I think it is a much more
complicated
question than simply saying the COX-2
coxibs in
this discussion, and we have to do apples
to apples
if we are going to make recommendations.
DR. WOOD: Dr. D'Agostino.
DR. D'AGOSTINO: My comment is very much
the same.
I am concerned about taking all these
individual studies. I think, you know, sort of the
potential is clear, but are we really
lumping just
because there is a direction on these
here.
Tom, for example, the question
about
splitting the arthritis populations
versus the
other populations, the arthritis
populations come
basically from the old clinical trials
where
adjudication was a problem and things of
that
nature.
So, how much do we believe that data and
how much do we want to draw this
inference?
So, I don't have a problem with
sort of
coming up with some sort of global
statement that
we are concerned, but I am concerned at
this point
424
about quantification in a very heavy way,
and we
just may be overdoing it in terms of how
we are
sort of answering this question.
DR. WOOD: Just to make sure I understand
your point, you are concerned about
putting a
number on it?
DR. D'AGOSTINO: I am concerned about this
global, I mean for us to say that it's
all COX-2.
DR. WOOD: But you would be comfortable
naming names?
DR. D'AGOSTINO: I don't know what I would
be comfortable with. I am uncomfortable with the
sort of global statement that we have
seen a number
of studies--
DR. WOOD: I am just trying to draw out
what you are saying. You would be more
comfortable--
DR. D'AGOSTINO: The only thing we said in
terms of separating was the arthritis
studies. Are
425
we comfortable with the arthritis
studies, do we
have enough information, do we feel
comfortable
enough with the adjudication process, the
recognition of the cardiovascular events
in those
studies?
I mean I think I am much more comfortable
when we come to these new studies.
DR. WOOD: So, what you are saying is that
the dilutional effect of these old
studies may be
substantial.
DR. D'AGOSTINO: Exactly, and I don't know
how we are actually dealing with that.
DR. WOOD: So, that is important for
people to understand. Do you want to develop that
a bit?
DR. D'AGOSTINO: What is that?
DR. WOOD: So, what you are saying is that
the studies that didn't have a
cardiovascular
endpoint--
DR. D'AGOSTINO: And trying to get
adjudication. They were showing a
signal. We
already there is a signal.
DR. WOOD: So, they may be diluting the
426
effects from when Tom adds on the back of
the
envelope--is that reasonable, Tom?
DR. FLEMING: Well, I fully agree that the
best analysis of this is one that we
don't have the
time to summarize here right now, but it
is one
that will drill down in all of these
dimensions as
best we can.
Almost certainly, the answer is
here, if
we had unlimited data. We have about 75,000
people.
That is a lot of insight, although we need
far more than that. We do have another 30,000
coming along shortly.
In essence, though, what we
really need,
if we had the ideal, is that ability to
drill down,
as Ralph says, by indication, and by
dose, by
ancillary care, looking at whether or not
it is an
aspirin or not an aspirin, by duration of
therapy.
These are all things that we have seen
the data
suggesting that there is very likely
these factors
are influential.
So, essentially, my attempt was
to say, in
a very crude way, what do you see from
10,000 feet
427
here, but then acknowledge exactly, as
Ralph said,
that you really do need to drill down.
DR. WOOD: It is not likely to be less
than the numbers you gave. It is likely to be
more, right?
DR. FLEMING: Well, my own sense about
this is this is the weighted average of
the
compilation of all these different
settings, and so
in all likelihood, in fact, with
certainty, there
will be settings where it is less, there
will be
settings where it is more.
Can you, for example, give
Celebrex at a
low dose with a short enough duration
that in wide
settings, it would be safe. That is still entirely
possible within the context of what we
have said.
Those are issues that we really need to
understand.
DR. D'AGOSTINO: And I am concerned
somewhat that if we give this global
statement,
that we sort of can't get back to the
question you
just raised, can we look at Celebrex at a
low dose,
because somehow or other, we are saying
it's in all
the COX-2s.
So, I want to be just careful
in how, the
answer to this question, how it comes out
quantitatively and what it locks us into
in terms
428
of further discussion.
DR. FLEMING: What I would say is
that--and I do agree that we need to look
beyond
these five or six products--but what I
would say in
several of these products, not just
Vioxx, in the
certain settings that we have looked at,
in my
view, there is evidence that establishes
there is
an excess risk.
There are other products where
there is a
suggestion, and we are underpowered,
though, to
discern whether or not that
suggestion--it is not a
suggestion of nothing, though, it is a
suggestion
of something, but it is more modest in
size than
the other agents although it could be a
dose issue,
it could be an indication issue that
explains it.
DR. WOOD: Dr. Domanski.
DR. DOMANSKI: I think if one looks, there
is really an attempt here to look very
carefully in
these quantitatives that we can. I think Dr.
429
Fleming provided a remarkable compilation
just now
for us.
But I think if one backs off to
sort of
high altitude and looks at these drugs,
the signal,
as people are using the term, it is
pretty clear
that there is an excess risk conferred by
some or
all of these drugs.
It seems to me the process is
sort of
turning around. We are trying very hard, you know,
the idea is to demonstrate--and it's the
sponsor
who has to do it--to demonstrate safety
and
efficacy, and not necessarily the purpose
of the
FDA or its advisory committees to somehow
demonstrate that the thing is unsafe.
It does look like they are
unsafe, but the
problem is that the studies presented
really are
not very good studies, and, in fact, one
of the
reasons that we probably didn't learn
sooner that
there is probably a real problem with these
drugs
is because of the relatively poor studies
that were
presented for approval.
So, I think it is important to
remember
430
who has got what role. It is theirs to demonstrate
safety and effectiveness, it is not ours
to
demonstrate it's unsafe.
DR. WOOD: Right, but I think the FDA is
looking to us to give them some guidance
here,
right?
DR. DOMANSKI: Well, that is important for
the future particularly, that is, what
studies
should be done next, and that is a
legitimate
concern.
DR. WOOD: Bob, do you want to say
something?
DR. TEMPLE: Just that it is the company's
job to show that it is safe, but we sort
of have to
say what would constitute adequate
evidence, what
sort of level of risk do you have to rule
out, how
long, and things like that.
Of course, we are in the
process of
learning about those things as these data
come in.
As I said before, I don't think anybody
would have
thought you need a four-year study, but
that is
sort of on the table now, and it wouldn't
have been
431
before.
So, it is helpful to know what
kind of
risk is plausible to rule out and all the
things
that Steve said before, I mean you have
got to
worry about what doses to study.
We encouraged everybody to
study high
doses to rule out GI distress. Whether that was
wise in retrospect, I am not sure, and I
think I
probably had something to do with it a
long time
ago.
I am not sure that was the best thing.
We
want to be really sure you couldn't make
an ulcer.
I, at least, wasn't thinking about maybe
making
something else.
So, all of those questions are
things we
need help with even though, yes, it's the
company's
job to bring the data forth.
DR. WOOD: Dr. Farrar.
DR. FARRAR: In terms of the specific
question that we are addressing, I also
just want
to point out that there is a second part
to that,
which it says, also, discuss the possible
mechanisms of action for an increased
432
cardiovascular risk with these
agents. I think
that has bearing related to the fact that
if we
accept that the in vitro selectivity of
the
COX-1/COX-2 analyses at least have some
bearing on
at least their metabolic process.
What I am struck by is the
variability of
the agents with regards to other factors
indicating
that simply COX-1/COX-2 inhibition is not
their
only action, and that the ones with some
of the
higher risks are not necessarily the most
COX-2
selective.
What that suggests to me is
that we really
don't understand the process yet well
enough to be
able to say that it is a group selective,
because I
am not sure what a COX-2 selective one
is, where do
you draw the line.
I think a more appropriate way
to say this
is to say that clearly, the role of COX-1
and COX-2
inhibition are important in the process
of both
anti-platelet and perhaps platelet
aggregation, and
that those with a more predominant COX-2
component
need to be studied carefully for the
potential
433
excess cardiovascular risk.
I have a great deal of
difficulty, though,
saying it is the group of drugs that have
been
called that by the pharmaceutical
industry. I
think that that is being very
short-sighted about
this.
In fact, the data that we have seen in
these presentations make me want to go
back and
look at ibuprofen with regards to a whole
host of
issues that we hadn't thought of
before. So, I
think it is very important that we keep
in mind
that there is not a distinct relationship
between
those numbers specifically and that we
need to be a
little bit broader in terms of that look.
The other point I would like to
make is
that we clearly need to differentiate in
terms of
what we are considering between the
placebo trials,
which have been done primarily in cancer
prevention, and the comparative trials
with other
agents.
As has been brought up many times, none of
the agents are the same, and so the
comparisons
there need to be carefully considered.
As such, I am in favor of a
statement that
says that we are consciously aware that
COX-2 is an
important component of this issue, but
that all
434
agents that claim to have, really all
agents that
are developed in the future and all the
current
agents need to be carefully looked at for
the
balance between the cardiovascular, GI,
and other
risk factors including the hypertension,
including
the pulmonary edema that we have heard so
much
about.
DR. WOOD: Right.
One pragmatic way
perhaps that we could handle this is
there are two
drugs of whatever the class we are
talking about is
that are left on the market, and for
which we have
a number of randomized trials recently,
and we
could consider them as a sort of present
tense
evaluation, and for future tense, other
drugs that
may have signals that we don't really
understand,
and certainly drugs that were likely to
be marketed
in this area, this space, and whatever
that means,
and would need some sort of evaluation.
So, that would sort of divide
up our work,
435
so that we would be considering what to
do about
the ones that are out there, what to do
about the
ones that are potentially out there, and
I guess a
third group is what to do about other
drugs that
may or may not fit into this class or may
not be at
some extreme of this class.
Is that sort of capturing the
essence of
what you are saying?
DR. FARRAR: That is certainly one way of
dividing up the work.
DR. WOOD: Let's think of it in terms of
that as we move forward.
Dr. Holmboe.
DR. HOLMBOE: Actually, a number of the
things that I was going to say have been
said. I
would just add one caveat, Tom, to what
you said,
that if you look at these trials, over 40
percent
of the patients never made it to the end
of the
study, which means that we probably don't
have over
70,000 patient observations. We probably have
about 20- to 30,000 less who actually
made it to
the end of the trial.
That, I am very concerned
about, and you
look at these trials also, although they
look
similar when they are first randomized,
that is the
436
purpose of randomization, the populations
that get
to the end don't, so I think that we are
also
lacking some very important information,
what
happened to a fairly large number of
individuals
who started but never got to the end of
the trial.
DR. FLEMING: Just to respond to that,
that is a key point. Now, the analysis that I did
yielded approximately 7- to 800 events,
so we are
getting the total number of events that
we would
have needed from a 70,000 person trial,
but your
point is still well taken.
We are not underpowered because
of the
lost to follow-up, but there may be a
bias here
that we all talked about earlier, that if
you
really wanted to get the most insightful,
reliable
assessment, you need to have high quality
follow-up, so that is something that is
still a
relevant point. They are unequal in their quality
of study conduct in the area of
follow-up.
DR. D'AGOSTINO: This was part of my
concern in terms of what I was trying to
raise,
that we have studies, but we don't have
studies,
there is a lot of problems with it.
DR. WOOD: Dr. Dworkin.
DR. DWORKIN: Ralph, I have a question to
437
follow up on what you were saying
earlier. If I
understood you, you were saying that you
are
uncomfortable with a global statement of
the sort
that Dr. Gross was making because you feel
there is
some kind of heterogeneity amongst the
data, of the
type that Tom summarized.
But then it seems to me you are
between a
rock and a hard place, because if you
believe that
there is a great deal of variability in
the results
with respect to risk, how could we
possibly
discriminate amongst the different drugs.
DR. WOOD: We have done that lots of times
before.
DR. D'AGOSTINO: The qualities of studies
are different. I think the arthritis studies is
where we get the CV information, they
weren't
438
designed to get the CV information,
cardiovascular
information, so I think there is a signal
there,
but I don't know how to interpret it.
I think there is the problem of
lost to
follow-up and things of this nature, and
all of
those things make me very uncomfortable,
and sort
of making a global statement and then
living with
that global statement.
I think it is clear or
hopefully it is
clear what I am concerned about. We don't want to
be locked into, by making a global
statement, later
on saying that no matter what drug we
look at, we
have an answer for, and it may be low
dose of
Celebrex, that may be viable, and not
unsafe.
We really need to worry about the
studies
that we are going to suggest, and if we
absolutely
thought there were safety problems, why
are we
suggesting them, why aren't we just
saying stop the
studies and get the drugs off the
market. I think
there is a lot of room for maybe there is
something
going on that is safe, and we want to
really pin
the issues down in good clinical trials.
DR. WOOD: Well, we have dealt with drugs
within classes before. I mean a statin was
removed, but the other statin is on the
market, and
439
troglitazone was removed, but the other
drugs
stayed on the market.
I guess the difference here,
which is only
fair to point out, is that this is
thought to be
producing toxicity through the primary
mechanism of
action.
At least that is one of the postulates,
but we certainly should deal with them as
individual drugs, I think, rather than as
a class
of drugs.
Dr. Abramson.
DR. ABRAMSON: I guess what I am thinking,
it is possible to accept the fact that
many of
these toxicities are via the COX-2
mechanism, but
recognizing that all of the class of NSAIDs, by
definition, when they are effective, are
inhibiting
COX-2, and I am still troubled by the
population
data which shows signals with
indomethacin and
Meloxicam, and by older data which shows
congestive
heart failure particularly with the
non-selective
440
drugs.
So, I think that we have to
look at again
the entire class, and particularly if you
look at
the CLASS and the TARGET trials, why is
ibuprofen
and diclofenac behaving pretty much like
Celebrex
and lumiracoxib, so if there is an
assumption on
our part that this class of drugs, even
the highly
selective COX-2s, increase by 1.4, 1.5
the relative
risk, why is ibuprofen and diclofenac
looking
pretty comparable in those large
population trials.
One answer is that they, in
themselves,
whether diclofenac is rofecoxib lite or
not, but
they themselves are imparting a risk, but
they
themselves have not been subject to these
long-term
placebo-controlled trials that we see in
APPROVe
and ADAPT.
So, therefore, I think the
COX-2 mechanism
may pertain, but it cuts across all
degrees of
relative selectivity.
DR. WOOD: Dr. Furberg.
DR. FURBERG: Well, I spent about five
years looking for a definition of class
effect, and
441
so far I have been unsuccessful. There is in the
literature no definition of class
effect. The
closest I came was an FDA definition of
class
labeling, and that was not a good one.
So, I think the working
definition of a
class effect would be that members of a
particular
group or class share common actions in
the broad
sense, and I think that would apply to
the COX-2s
in my reading. They provide pain relief, GI
protection, raise blood pressure, cause
fluid
retention, have the undesired effects on
cardiovascular risk, so in my mind this
is a class
and sharing a lot of actions, and that would
include the increased cardiovascular
risk.
DR. WOOD: I am going to take two more
questions on this topic and then I would
like to
move us to, I guess, considering which
drugs, to
answer Question 1, which drugs, rather
than a
class, which drugs we see a
cardiovascular signal
with, which is one way to approach the
problem.
You are the next question,
Steve.
DR. NISSEN: What I wanted to make sure we
442
got to is this issue of mechanism, which
is
actually in the question here, and the
reason it's
important is that I am not quite ready to
accept
the hypothesis that one can predict from
the COX-2
selectivity and duration what is going to
happen in
these drugs.
Let me see if I can explain
that because I
think it is very important as we think
about how to
go forward here. I see a broad spectrum of blood
pressure changes that don't seem to be as
tightly
linked to the COX selectivity as one
would guess.
Lumiracoxib, for example, which
is very
COX selective, doesn't appear to have
much effect
on
blood pressure. Rofecoxib has the
largest
effect on blood pressure by far and is
relatively
COX selective, and they are very
different.
So, for the FDA, I think if you
want to
characterize the drugs, not only do we
need
clinical trials around looking for GI
safety and
cardiovascular safety, we need a
standardized
method to look at the effect of these
drugs and
their intended doses on blood pressure,
and they
443
ought to all be subjected to similar
scrutiny, so
we can compare apples to apples and
wherever
possible with active comparators, let us
understand
that.
Now, why do I say that? Because Bob and I
have sat at many a meeting and looked at
blood
pressure drugs, and I can tell you the
data on the
relationship between relatively small
differences
in blood and cardiovascular morbidity and
mortality
is rock solid across a huge number of
drugs and
interventions, and you almost can predict
what will
happen.
So, we need to know--and as I
sit here, I
can't tell you that drug X in this class
has Y
blood pressure effect and drug A has B
blood
pressure effect--and so we don't know,
and we can't
inform physicians about that unless we
have better
data on blood pressure.
So, I am making an appeal that
we get to
that level of specificity, and that is
not a very
big trial to do that. Bob, what do you usually ask
for in the blood pressure study?
DR. TEMPLE: Well, if you use automated
pressure monitoring, I think you can get
a decent
answer with 20 or 30 per group, maybe
40. It's
444
very easy.
DR. WOOD: Let's move on. I am going to
give you the last word in a second. After we get
Dr. Gross's comment, we are going to
divide this
first question into three things, which
drugs do we
see a cardiovascular effect of, and the
secondly,
we can ask whether we see a class effect,
whatever
we understand that at, and I am not sure
we do, and
then the third question that is in
Question 1 is
what do we see as a mechanism.
So, let's divide them into
these three
things and let's move to an answer.
Peter, last word.
DR. GROSS: What we say here today about
these is going to have a significant
impact in
molding public perception, and if we
conclude that
there is a class effect for the selective
COX-2
inhibitors, and don't say the same thing
about
non-selective NSAIDs in general, then
people are
445
not going to want to use COX-2 inhibitors
at all
and they will be using the non-selective
NSAIDs,
which from the data presented, doesn't
look as
though many of them are better from a
risk point of
view.
So, I just issue that note of
caution.
DR. WOOD: Does the FDA want us to go
around the table asking people for an answer
to
each of these questions, the subsets of
these,
John?
DR. JENKINS: I think we really viewed
these questions as things to stimulate
your
discussion, not necessarily things that
are
amenable to yes/no answer. The yes/no answers come
tomorrow.
DR. WOOD: So, can we move on?
DR. JENKINS: If you think you are done
with No. 1.
DR. WOOD: We are done with No. 1.
DR. TEMPLE: I just wondered if people
could come to grips a little bit with
some of what
Steve said and some of what other people
said. I
446
absolutely don't want to put words in
anybody's
mouth, but what I heard people say was
they think
the class has at least the potential for
having
this problem because of the imbalance and
because
of the stuff we have heard about before,
and that
you need to look at each drug to see
whether that
is manifested at a particular dose-dose
interval
and all the rest.
I just wondered whether that is
getting
close to what people are saying or not,
and I
absolutely am not giving my view on it, I
am just
suggesting it.
DR. WOOD: Let me try and answer that and
then we will go around and ask other
people.
I think what we are saying is
almost the
same as the GI effect before the GI
effect or not
was worked out for the so-called COX-2
inhibitors,
that I see an effect, a cardiac effect
from
valdecoxib, certainly from Vioxx, and
from
celecoxib, and there is a dearth of data
on the
nonsteroidals, the other nonsteroidals at
this
stage in terms of cardiac safety, and we
are not
447
going to be able to decide that even on
Friday, it
seems to me.
In the presence of that signal,
the
prudent activity would be to go look at
it sometime
in the future, but we can't do that
between now and
Friday night. So, that is sort of where I come
down.
Dr. Abramson.
DR. ABRAMSON: I might have a slightly
different view, because I mean I think of
the class
more broadly as it is defined now to
include both
the COX selective and non-selective drugs. I think
that there is a signal probably for all
of these
drugs, maybe by different mechanisms
perhaps. I
think we have under-recognized that in
the
population, I think physicians have not
been
concerned enough about blood pressure
changes.
So, my view is that maybe there
will be
different mechanisms, but that each of
these drugs
is suspect as having an increased
relative risk
when used chronically, whether it is
ibuprofen or
the
most selective COX-2.
My own view is, as I said
earlier, is this
is not dissimilar to the late '90s, and
until you
prove otherwise, this is GI warning that
these
448
drugs may cause cardiovascular risk or GI
warning
it may cause serious adverse events, and
I think
each of them should be held to that right
now until
someone proves otherwise, because I think
it would
be wrong based on the evidence to assume
that three
drugs have a cardiovascular risk, and
several of
the others don't, simply because we don't
have the
evidence.
I am also concerned about some
of the
research that was talked about by one of
the public
speakers.
At most of our universities, these
studies have actually stopped because of
concern
that these drugs are not as safe than the
non-selective drugs.
I think, particularly in cancer
and
others, we are doing the public a
disservice by
prematurely picking out these drugs as
being unsafe
and stopping some very important research
where the
risk-benefit might even be more important
than in
449
arthritis.
DR. WOOD: Dr. Shapiro, I missed you, I am
sorry.
MS. SHAPIRO: That's okay.
I think I
agree with you, and I think it is hard to
properly
answer this question unless we ask
ourselves why it
is being asked, and if it's being asked
because the
FDA wants some broad-brush, uniform
regulatory
approach to this, what I hear people
saying around
the table is that that would not be
appropriate for
each and every one of the drugs that are
in this
possible class.
But if we are saying that we
think that
drugs that are related in composition,
structure,
this, that, and the other thing, should
raise a red
flag, which is what I think you are
saying, that is
what I think we want to say, and I think
we are
getting hung up on this class effect
definition
because we haven't gotten behind and asked
why we
are being asked the question.
DR. WOOD: Dr. Furberg.
DR. FURBERG: I think I disagree with
450
Steve Nissen, and I think it is a mistake
to focus
on one mechanism of action. Members of a
drug
class, they don't have to share all
mechanisms of
action.
In fact, I don't know of any drug class
where all the members share all
mechanisms of
actions, so the term is more loose and
relative.
DR. WOOD: Dr. Shafer.
DR. SHAFER: Actually, I think Dr. Nissen
said that it's not all one
mechanism. I think that
is exactly your point.
DR. NISSEN: Exactly my point. My point,
Curt, was that these drugs do differ by
some of
them have much more of pressor effect
than others,
and that seems to be dissociated at least
somewhat
from their COX selectivity, and so I want
to
characterize the drugs individually, not
necessarily collectively.
DR. SHAFER: Continuing that same line of
argument, Bob, in answer to your question
that you
had raised, if there was all the
FitzGerald
hypothesis, then, the class effect makes
a ton of
sense, because you would say okay, you
look at the
451
COX-2 selectivity, we kind of go on the
list, and
we do our cutoffs.
We have the blood pressure
data. I point
out once again we do have the aspirin
data in some
very big trials. The effect should have
gone away
in the presence of aspirin particularly I
point out
again to the APPROVe trial, the
thrombotic risk was
3.25.
We have talked about this on
and off, and
you haven't been feeling well, so we
haven't had a
chance to really get together and
discuss, at risk
of my health, despite having lunch at
Chuck E
Cheese, but I am concerned because we
haven't
explained the aspirin effect, and
aspirin, unlike
the other drugs, doesn't go away, it
doesn't have a
pharmacokinetic component. I mean that should have
clearly made a statement if the aspirin
effect had
reversed these prothrombotic effects.
Steve, I think that argues to
your point
that there are several mechanisms. One is
certainly in part the FitzGerald
hypothesis
although there is partly a class effect,
but the
452
aspirin also shows that there is
something else
going on.
DR. TEMPLE: Then, how do you characterize
the class? I mean it sounds like most people think
you are characterizing the class as one
with a
preference for the COX-2 receptor, for
that one,
and if you can't do that, it is hard to
know how to
go forward.
DR. SHAFER: Can I answer that?
DR. TEMPLE: But I agree with you about
the aspirin, it's a fly in the ointment.
DR. SHAFER: It seems to me that you can
look at where we have data that is
consistent with
the FitzGerald hypothesis, that is
consistent with
it, that you can say these drugs are
behaving as in
class, and certainly for the coxibs, as
Dr. Fleming
presented the data, it appears that they
are all
behaving in a way that is consistent with
a class
effect.
Where we don't have more
specific data
that would say they are behaving in this
fashion,
and I would point out these are the
COX-2s that we
453
don't have data because they are older
drugs, but
they appear to be COX-2 selective, I am
reluctant
to include those in the class and sort of
damn them
because of where they show up on some
table. At
the same point in time, I am reluctant
just to give
them a get-out-of-jail-free card, if you
will.
I think that something needs to
be noted
that they are potentially at risk for
this effect.
DR. WOOD:
Dr. Cush, then Dr. Hennekens.
DR. CUSH: I would support what Steve said
and that I think that we came here with
the
spotlight focused on Class II specific
agents, but
we become more curious as we have seen
all of them
fall, but then seen all the other drugs,
the
non-selective drugs also seem to have
some of the
same failings, we don't want to focus
solely upon
the COX-2 specifics, but I think that we
can start
there and then extend our concerns to the
other
agents, as well.
It doesn't have to be, it can
be linked to
COX-2, and that may be where we start,
but it
obviously needs other investigation to
look for a
454
mechanism of action.
DR. WOOD: If Raymond Pickey were here, he
would say show me the data that tells you
that
these other drugs have this effect in
published
trials.
DR. CUSH: Well, one would be I guess some
of the observational data.
DR. WOOD: Randomized, published
randomized trials.
DR. CUSH: Well, I think the only one we
really have is the Norwegian study.
DR. WOOD: Is that a randomized trial?
DR. CUSH: I believe it was. Well, they
were randomized to--
DR. WOOD: That showed aspirin also had a
negative effect.
DR. SHAFER:
Alastair, that is the reason
the Challenger blew up, the sort of show
me it's
safe, prove to me it's safe or I am not
going to
make a statement.
DR. WOOD: That is not the issue at all.
I mean we have got to be careful, I
think, rushing
455
ahead of credible data on the basis of
rumors of
war that are brought in from outside.
I mean we have got four randomized
and
controlled trials for three drugs, and we
have got
some news of other drugs, it seems to me,
that are
not--and documented very well. That is not giving
anyone a get-out-of-jail-free card, but I
think we
have got to sort of go through this in an
orderly
fashion.
Otherwise, we will be regulating on rumor
forever, and I think that is a very
dangerous step
to take.
DR. TEMPLE: You do have some diclofenac
data i comparison to some of the drugs
that are of
interest, so you have some. It's not the
placebo-controlled trial you are dreaming
of, but
you do have that, and you have naproxen
and several
comparisons, as well.
DR. WOOD: And that looks pretty good.
DR. TEMPLE: Naproxen looks good,
ibuprofen looks the same as--there are, I
didn't
count them up, three or four control
groups of the
older ones scattered around.
DR. FLEMING: Well, we can be specific
because Bob is right, we do have--I mean
basically,
because of all of these other studies
that were
456
done for the COX-2 inhibitors, there is a
lot of
data on naproxen and a lot of data on
diclofenac,
and diclofenac in the etoricoxib trial
and in the
CLASS trial more or less came out looking
like the
COX-2 inhibitors, while the winner is
naproxen.
Basically, in the VIGOR trial, in the
etoricoxib trial, very much in the
lumiracoxib
trial, it came out positive. Now, we are going to
hear something tomorrow about the ADAPT,
but
looking at these others, it sure looks
like
naproxen is a winner, and it does look
like the
theory that was put forward that
diclofenac is
COX-2-like is at least supported by the
trials
where it was studied.
DR. WOOD: Right, but all we can say is
they look the same as another drug where we
are not
absolutely certain of the effect of that
other
drug.
DR. FLEMING: That is true although we
457
have a lot of other studies on the other
drug, and
it is always you have got to be careful
when you
say A is better than B, and then B is the
same as
C, is C worse than A, but there is that
kind of
evidence.
DR. WOOD: Which is what I am concerned
about.
Dr. Hennekens.
DR. HENNEKENS: I would say I am
struggling with trying to gain this
clarity, but as
I view the drugs that either have been or
are
marketed with regard to cardiovascular
risk, the
picture that emerges, begins to emerge to
me is
that rofecoxib, ibuprofen, and possibly
valdecoxib
are in one bin, diclofenac and celecoxib
in another
bin, naproxen in a third bin, and then
aspirin in
the fourth bin, going from concerns about
hazard to
neutrality to benefit.
DR. WOOD: Other comments? Dr. Farrar.
DR. FARRAR: Two quick--well, I guess
every time we mention aspirin, it never
ends up
being quick--but two quick comments, one
of which
458
is that I am not as concerned about
aspirin
knocking out the issue of the COX-1/COX-2
problem
primarily because, in fact, aspirin is a
surrogate
marker for people with cardiovascular
disease.
If you look at the actual rates
in all of
the aspirin groups, they are at least, at
least 2
to 3 times the rates in the non-aspirin
groups to
start with. So, I think that there is an issue
there.
I think the second issue has to
do with
what was just discussed in terms of the
comparison
of drugs, and just to emphasize the fact that
what
we are talking about is we have data for
there
being a risk factor in the
placebo-controlled
trials, primarily the best data, which we
will have
a whole lot more of in two months or
three months,
of the cancer prevention trials to tell
us what the
level of risk is.
Then, we have the comparison
data that Bob
Temple was just talking about in terms of
the
non-selective versus the selective that
say that
they have very similar levels of risk.
The third point just to make is
that all
of this discussion about risk, I don't
want to
imply that I think that this risk is big
enough to
459
actually warrant the continued hold on
all the
trials that we have going, and I think,
in fact,
what it suggests to me is that we need to
continue
with trials to understand better what the
data is
telling us.
DR. SHAFER: May I respond to the aspirin
point?
This confusion that you raise came up when
I first raised it, I guess it was just
yesterday,
but the risk that we are talking about is
not
aspirin versus non-aspirin, because
clearly,
aspirin will be a marker for increased
risk.
What we are talking about is
the risk of
rofecoxib in the case of APPROVe, the
risk of
rofecoxib versus the comparator in those
patients
taking aspirin, so that the increased
risk of
cardiovascular events has been evenly
distributed
between the two groups, because that is
the blinded
comparator variable.
So, we are talking about the
risk of COX-2
460
versus non-COX-2 in those patients on
aspirin. It
is different from the risk of aspirin
versus
non-aspirin, which as you say is, of
course, that
risk is confounded. But in this case, that risk is
evenly distributed between the two
groups.
DR. WOOD: Go ahead, Dr. Farrar.
DR. FARRAR: I think the problem is that
what you are saying is that aspirin is
somehow only
a COX-1 inhibitor and therefore it has a
role there
that somehow should balance the COX-2 or
there
should be some other process going on.
There is no question that
aspirin and its
indication of increased cardiovascular
disease has
an effect on the relationship of the COX
problem.
We have seen multiple examples in the
cardiovascular risk, in the group who
have the high
cardiovascular risk, there is a different
response
to the COX-2 problem than in the lower
group, so
there is no question about that.
But I would argue that aspirin
is as very
different drug in terms of how it works,
in terms
of its binding to the sites, so all I am
saying is
461
that I am not sure that that obviates the
need to
say that there is an issue there with
COX-1/COX-2
that we need to look at more thoroughly.
DR. WOOD: Dr. Manzi.
DR. MANZI: I actually have a problem
making inferences about diclofenac and
naproxen in
studies where I think we have a difficult
time
feeling comfortable with the results in
relationship to the COX-2s. I mean the trials that
are really driving the signal here are
the
placebo-controlled trials of long
duration.
So, to feel that we don't have
enough
information to really feel comfortable
with COX-2s,
and then to try and extrapolate to the
comparators
in those, I think is dangerous.
DR. WOOD: That is what I was saying, too.
You know, it's ten past 5:00, just to
draw
everybody's attention to that.
John, you are saying that you
don't
necessarily want a vote on this, is that
right?
So, I guess the question is, is there
further
discussion on this specific question that
the
462
committee feels they can't hold until
tomorrow?
Tom?
DR. FLEMING: I share the caution in that
last comment, but I will just note that
methodologically, it is the exact problem
we run
into or situation we run into in non-inferiority
trial designs, because you have
placebo-controlled
trial of agent A, and now you want to
look at
whether B is adequately safe, and you are
looking
at B against C, the new agent, and if C
is the same
as B, that was shown to be non-inferior,
or you
knew what its relationship was to no
treatment, it
is that non-inferiority issue.
Nevertheless, many of us have
concerns
with non-inferiority settings, but that
is the
methodologic challenge.
DR. WOOD: That is my concern, as well.
Let's move on Question No.
2. We may have
discussed this a lot already, but this
really
addresses the contributions and
limitations of the
currently available observational studies to
the
assessment of cardiovascular risk for the
463
non-selective and COX-2 selective--and
let's not
bog down in what we mean by that. In particular,
discuss the role of such observational
studies in
informing regulatory decisions about
postmarketing
safety issues.
Now, let me ask a clarification
question.
Does this mean we just sort of ignore the
randomized trials here or take them as a
given, or
how do you want us to handle that?
DR. JENKINS: I think the idea here was to
get your thoughts on how we should
consider and
weigh these studies in a mixture where we
have some
control trials, we have the observational
trials.
Sometimes they don't agree with one
another.
Sometimes the observational data come at
a time
when we don't have the control data.
We are trying to get your take on what
weight should we place on these data as
we are
trying to make regulatory decisions.
DR. WOOD: So that we could modify the
question to sort of include the
randomized trials
and
say how do we relatively assess these and weigh
464
them up?
DR. JENKINS: Sure.
DR. WOOD: All right.
So, that is a
helpful clarification.
Comments on that question? Yes.
Dr.
Stemhagen.
DR. STEMHAGEN: A couple things. I think
I want to make sure that it is
understood, in my
view, that they are definitely
supplementary to the
randomized clinical trials.
I think we all recognize that
the value of
randomized clinical trials is the
randomization,
that we don't have the selection bias
that
certainly takes place in observational
studies, but
nevertheless, when we think about the
magnitude of
the studies that we have, we have over
many
hundreds of thousands of patient years of
exposure,
we have in the cohort studies.
In the case-controlled studies we
have
more than 25,000 cases. We do have a very rich
data set.
I think we have talked a lot
about the
465
fact that we have got a number of studies
and we
see a lot of consistency in the results
between
those studies. There was an issue of maybe they
are all biased in the same direction.
I think they were conducted in
very
different ways, many of them, and many
very
different databases. We also see some data on dose
response, which is another suggestion
that there is
something going on and that the data
should be
believed.
I think if we talk about lost
to follow-up
in some of the randomized trials, in some
of the
very stable populations that we have in
some of the
databases, we actually do have long
follow-up,
although ideally, we would like these
studies to go
on longer. None of them are really as long as we
would like, and part of that I think is
the data
being on the market or available within
those
databases at the times that the studies
were done.
Another thing that really is
different
with these studies is we are not just
talking about
volunteers. When we do our clinical trials, we are
466
talking about volunteers. In our
databases, we
really have the totality of patients, of
cases, of
exposures.
So, I think we have got a
somewhat
different groups of patients. The clinical trial
patients are essentially a subset to
that. We also
are looking at actual use doses, which
are somewhat
different doses perhaps than in a lot of
these
clinical trials where we have talked
about high
doses are pushing the dose.
So, I think they are different
pieces of
information. The endpoints that we are looking for
are very hard endpoints, and I think we
have talked
about, and there was some evidence, that
in some of
these studies, there are adjudications,
the same
way there are in clinical trials when the
medical
records are collected.
There have been some validation
studies
looking at the ascertainment of MI and
feeling that
it is very complete. So, I think we can feel
reassured that in these closed populations,
we
probably have identified the cases that
we are
467
interested in, and we also have a lot of
data, not
necessarily exclusive, on the
confounders, and
there have been adjustment for
confounders.
So, I really want to urge that
when we
look at the data, we don't just dismiss
the
randomized clinical trials, but they are
telling us
something. They do have some patterns, and they do
show some differences between the
products.
DR. WOOD: Dr. Cush.
DR. CUSH: I think there is obviously a
value for observational studies, but one
thing I
keep hearing is that the FDA is not properly
empowered to mandate that postmarketing
trials be
done until maybe a significant issue like
this
comes up.
This kind of public health
issue sort of
underscores some of the weaknesses of the
current
MedWatch system where common events like this
are
not going to get reported on new drugs,
because
people get heart attacks and heart
failure and
uncontrolled hypertension, and I think
that one
thing I would like to see come out of
this is that
468
Congress and others empower the FDA, so
they can do
postmarketing trials that need to be
done, either
mandate it or as they need to occur, and
if they
can, mandate registries as they need to
be done, as
well.
That is certainly right now
what I think
is a big hole in our current safety
system. We
heard today from the patients, they want
to know
that we are going to help them. That
mainly means
they want to know that we are going to
give them
medicines that are safe.
DR. WOOD: Dr. Bathon.
DR. BATHON: I would like to take the
example of naproxen for a minute where it
seems
like from observational studies, it has a
neutral
effect on cardiovascular risk, at least
that was
the overwhelming notion, whereas, in
randomized
trials it seems to be more protective.
I would like to explore for a minute why
that discrepancy, if it is true, why it
might be
true.
I would like to posit that in the randomized
trials, we have people taking drug every
day or at
469
least we think they are taking it, and
they are
taking it in the appropriate dose to have
consistent COX-2 or whatever, COX-1 and
COX-2
inhibition.
In observational data, those
are driven,
NSAID drug use is driven primarily by
acute pain
syndromes and osteoarthritis, where
people, if we
go to the acute, somebody has back pain
for a few
months, a lot of the people using those
drugs might
be on them for a few weeks or a few
months.
The proportion of patients like
the
rheumatoids or the bad OA patients who
might be
taking them every day is probably
relatively very
small in that group. Even within the OA group, I
think a lot of us probably have OA in here, some of
us who have gray hair or getting gray,
even the OA
patients do not take the drug every day
on average.
The rheumatoids tend to, the OA
patients
don't, and then the acute pain syndrome
people or
the back pain are more intermittent.
So, I wonder if the difference
between
observational data and the clinical trial
is driven
470
by the fact that we are looking at very
different
treatment regimens, treatment durations,
and so
forth.
So, I think the randomized trials are more
valuable here than the observational
data.
DR. WOOD: Dr. Holmboe.
DR. HOLMBOE: I would just make a couple
of points. If we agree on No. 1 that there
actually is harm, then, I think yes, you
are going
to have to do observational studies. I mean it is
going to be hard to randomize somebody to
study
harm.
I think that we can take some
comfort even
though that the effects are different,
that the
observational trials were reasonably
consistent
with a lot of the randomized controlled
trials that
were presented today.
Second, I think a poor
randomized
controlled trial actually may be worse
than a good
observational study. As I mentioned
earlier, a
number of these studies had over 40 percent
dropout
with these patients not being followed,
and I think
that that is an opportunity for the FDA
to follow
471
these people out to see if there is
something
inherently different about those
populations who
aren't continuing on the study drug.
The third point I would make,
that with
regard to meta-analysis, it is very
important that
the trials be fairly homogeneous in the
way they
were done. In all the stuff reported, I did not
see anybody talk about a test of
heterogeneity to
see if they really truly could be
combined.
While I understand that because
the events
are so low, you are trying to pool risk,
there is
some danger in pooling studies that are
quite
disparate. So, I think that is something that
needs to be taken into consideration.
The last thing I would say is
that I think
there is a real lesson here potentially for
the
FDA.
The comparator drugs were approved before we
truly understood the biologic mechanism
of these
drugs.
Our understanding of COX-1 and COX-2
occurred long after the original comparator
drugs
were approved.
So, it is a real challenge I
think for the
472
FDA to go back and say wait a minute,
could these
comparator drugs potentially be a lot
like the
drugs that we are now studying, that we
think are
being proposed as different, but, in
fact, may not.
So, I think that that is real lesson, it
has
created a lot of the confusion we are now
having to
deal with, because a lot of the comparator
drugs it
turns out actually are very similar to
the COX-2s
that we are evaluating.
DR. WOOD: Dr. Day.
DR. DAY: Concerning the 40 percent
dropout rate in the randomized trials, we
have all
the sponsors here, and they have lots of
data and
computers, and so on. Would it be useful to get
the percentage dropout for each of the
target drugs
and the comparators and/or placebos in a
giant
chart before tomorrow to see, and then
try to get a
breakdown of what the reasons were for
dropout?
Do they retain that information
when a
patient drops out, what the reason is, or
is that
on file somewhere?
DR. TEMPLE: They always provide it. The
473
question is how reliable it is. A lot of them say
administrative reasons, and it really
requires
people to pursue that question, interview
the
patient, and while that is properly done
sometimes,
it isn't by any means always properly
done.
DR. DAY: So, the breakdown isn't
possible.
What about the percentages for each of
the groups that we have seen just in
these studies?
DR. TEMPLE: Pretty much all studies know
how many people stopped and completed and
when.
DR. DAY: Do we know?
Have we been given
those data?
DR. WOOD: Well, I guess the Kaplan-Meier
curves, and under each Kaplan-Meier
curve, I think
there is a number of patients at each
point.
DR. D'AGOSTINO: Part of that was dropout,
but part of that was the way they
planned, you
know, follow-up on the individual. The individual
could, for some reason or another, say
they are not
going to take the drug anymore, and they
only
follow them 14 days, so that was
legitimate in the
study.
A dropout that just disappears was sort of
474
illegitimate, that was not split up.
DR. WOOD: It is still a dropout, I mean
the person didn't complete the study.
DR. D'AGOSTINO: Well, it followed the
protocol. I mean you can't now go back
and say they
should have done something.
DR. WOOD: Dr. Paganini.
DR. PAGANINI: One of the things that I
was surprised at here was the lack of
information
on the older NSAIDs, and that is one of
the things
that we are trying to deal with is what
is the
difference.
That then speaks to continued
observational studies in the postmarket
venue where
if we had had that, we would have at
least had some
sort of observational anchor to put some
of the
newer drugs on.
Let me also add that while we
always look
at prospective randomized controlled
studies as
being the be-all and end-all, there is
now an
emerging--and I will ask some of the
biostat folks
to comment on this--a developing thought
process of
475
having a wild arm, the wild arm being
what is
usually and customary done when doing
something.
For example, if you do a dose
of a drug,
or you do an amount of O2 delivery or
some sort of
a respiratory issue in the ICUs,
frequently, when
you enter into a study which is randomly
controlled, you have one arm versus the
other arm,
and they are fixed arms, but there is now
a third
arm that people are starting to ask for.
It's a wild arm, what do people
usually do
outside of the study, and I think that is
a very
important issue for when you are using
drugs in a
common, out-of-the-box way where
everybody is using
the drug.
So, postmarketing observational studies
might be considered the wild arm for some
prospective randomized controlled trials
in that
same era.
DR. WOOD: Dr. Nissen.
DR. NISSEN: It is interesting. We like
our observational studies when they show
us what we
want to see, and we just hate them when
they show
us what we don't want to see.
I have lived through this with
the
estrogen business. I had people tell me that it
was absolutely unethical to do a trial of
476
postmenopausal estrogens because
everybody knew
they were beneficial, every observational
study had
shown it.
So, it is important that we use
observational studies as
hypothesis-generating
studies.
If you see a signal in an
observational
study, it is an indicator that you need
to do a
randomized controlled trial, and that is
how we
ought to use them. If we get too far beyond that,
we are going to get into the women's
health
initiative kind of problem again.
It comes up every generation as
another
example of this, where every
observational study
tells us one thing until we do a
randomized trial,
we find exactly the opposite.
DR. D'AGOSTINO: I want you to recall that
the Framingham studies said just the
opposite, it
was the observational study that didn't
agree.
DR. NISSEN: Thank you, Ralph, you are
477
usually right.
DR. WOOD: You were down next to speak,
Ralph, is that your question?
DR. D'AGOSTINO: Oh, is it my turn for my
question?
DR. WOOD: Yes.
DR. D'AGOSTINO: When we have
placebo-controlled trials, randomized
controlled
trials, I mean in some sense it is I
think the gold
standard, and when you have positive
comparators,
randomized controlled, it's the next
level, I think
that we have a lot of data that is well
developed
in terms of the studies.
We have questions about the
dropout, and
so forth, and I raised them also, but I
think the
randomized controlled trials have put us
in the
situation where we can minimized in some
sense the
observational studies.
Yesterday, I made my comment about
torturing the data. We can torture the
observational studies forever and ever,
but I think
our weight should shift on the
placebo-controlled
478
trials.
DR. WOOD: I agree with that.
Dr. Fleming.
DR. FLEMING: Maybe just to be specific
here about different kinds of
observational
studies, there is passive surveillance
and active
surveillance. Passive surveillance has been widely
used, for example, in vaccines, childhood
vaccines,
and with the Veer system.
Essentially, it worked really
well when
you are trying to detect rare events and
events
that are proximal to the time of the
intervention.
So, introsusception with rotovirus and
encephalitis, and anaphylaxis, et cetera,
have been
assessed fairly well.
The problem with those, and we
heard
naproxen experiences in what I would call
passive
surveillance, the problem is if you have
events
that occur with more regular frequency in
the
background, it is going to be almost
impossible.
There is under-reporting, you don't have
denominators.
So, a step up is the
large-linked
databases or the active surveillance
systems, and I
think this is what a lot of what we have
been
479
talking about with these observational
studies.
They give us numerators and denominators,
they give
us more complete ascertainment, but they
still have
unavailability often of confounder
information on
aspirin use, smoking, outcome specificity
and
sensitivities are less reliable.
We have talked earlier today
about how it
is extremely difficult in that context to
do a
valid ITT type analysis and have a time
zero cohort
and minimize lost to follow-up, and
ultimately, you
are not randomizing, and randomizing
doesn't solve
all problems, but it does, in essence,
eliminate
the systematic occurrence of imbalance.
It doesn't eliminate randomly
occurring
imbalances until you have large numbers,
but you
cannot, with covariates, go back and
adjust for
what is different in an observational
study,
because I always say the known and
recorded
covariates are just the tip of the
iceberg, so you
480
are left with a great deal of uncertainty
about
bias.
Where they are very effective
is
understanding natural history,
understanding event
rates, understanding covariates,
understanding how
people are treated, but we really want to
use them
to understand causality, does
intervention have an
effect.
Essentially, if it is a very
large effect,
you can get some reasonable senses, but
in most
cases, I think they serve a very useful
purpose,
but it's hypothesis generation, it's
development of
clues.
So, if we look at the overview
that David
Graham gave, my sense is he was able to
give us
insights about a wide array of issues
that we have
not yet got adequate randomized trials,
so
specifically, the nonspecific NSAs, what
does it
look like there, and issues about dose,
but I would
call those hypothesis generation or
clues.
I would be very reluctant for
the majority
of what we saw from those analyses to
take those
481
results as established. It rather gives us a guide
because we can't do randomized trials in
every
setting.
It gives us a guide for how to design
those trials and where the most pressing
questions
are.
So, the observational studies go
hand in
hand, but the ultimate answers in most
cases really
come from the randomized trials.
DR. WOOD: Right, and the estrogen studies
shouldn't be forgotten, right?
Dr. Morris.
DR. MORRIS: I think Tom said a lot of
what I wanted to say, but a lot
better. In terms
of causality assessment, living through
what the
Agency of Healthcare Policy and Research
went
through for outcomes, I think the
conclusion is
unless you randomize, you are never
really sure.
In terms of observational
studies, I think
it is interesting that like event rates
or
something like that, where we think it is
so much
better, yet, I was struck in the
discussion today
of some of these drugs is how much the
event rates
482
varied by center or study or country.
What isn't done in observational studies,
what could be done, is more of a
population-based
sampling, so we have a better
understanding of how
much or how well that particular database
is
representative of the broader population
of the
U.S., so we can do some kind of sampling
or
extrapolation and get much better event
rates,
where I think observational studies can
really do a
much better job than clinical trials
because they
can measure naturally occurring events
much better.
DR. WOOD: Dr. Domanski.
DR. DOMANSKI: You know, one always hates
to admit ignorance, but I want to pursue
this
business of a wild arm. I mean I have
seen some
pretty wild arms in clinical trials, but
never as a
third one.
I don't understand where that
is, I have
not heard of that one, and I would like
to learn
more about it. Can you explain that?
DR. PAGANINI: I will give you an example
of an NIH-VA study that is now ongoing
looking at
483
dose of dialysis delivered in which there
is a high
dose delivered and then there is a low
dose
delivered. Then, there is the thought process of
putting a third arm on there is what is
everybody
delivering anyway, so it is whatever the
wild type
is, to see if, in fact, people are
artificially
placed into one dose versus a second
dose, and
that, in and of itself, is an artificial
placement
of patients as opposed to what people
usually do.
So, therefore, what is the
comparison
between one dose versus a second dose
versus what
is usually and customarily done.
DR. DOMANSKI: But don't you usually use a
registry for that kind of question, that
is, how
well does it represent practice I guess?
DR. PAGANINI: It could be retrospective,
but in effect now what they are doing is
a
prospective collection of data of what is
normally
done in that particular institution when
people are
off study.
DR. DOMANSKI: Again, registries can be
prospective, of course. I am
having trouble seeing
484
the difference. I mean are those people
randomized, as well?
DR. PAGANINI: No.
DR. DOMANSKI: Okay, so it's a registry.
DR. PAGANINI: It's just a registry.
DR. WOOD: Dr. Hennekens.
DR. HENNEKENS: I would view the strengths
and limitations of observational studies
to be a
function of the effect size. For
the moderate to
large effects, we can make safe clinical
and policy
decisions based on consistency of the
data from the
observational studies.
As the effect sizes get
smaller, however,
it's a two-fold problem because now the
effect
sizes we are seeking are as big as the
amount of
uncontrolled and uncontrollable
confounding that is
inherent in the designs.
There is a certain seduction
from these
large-scale databases because you have a
large
number of data you control confounding
on, you
could get very robust p values, so you
begin to
believe that you have really discovered
something,
485
but I agree strongly with Tom that for
small to
moderate effects, they are useful to
formulate, not
test, hypotheses, so what Dr. Graham told
us this
morning are useful to formulate hypotheses.
If people took them as serious
evidence
that this indicated harm, he might be
right, but it
would have nothing to do with the data
that we have
seen.
I conclude with the statement, I have the
privilege to know Sir Austin Bradford
Hill who, on
this question, and I think Rich would
agree with
this, he said, "Don't let the
glitter of the tea
table detract from the quality of the
fare."
DR. WOOD: Dr. Elashoff.
DR. ELASHOFF: Two comments.
One, in this
situation, especially when there is very
specific
evidence that the relative risk may vary
over time,
looking at the standard way that
observational
studies lump it all into patient years is
bound to
be misleading.
A second point has to do with
the fact
that in a randomized trial, when you are
comparing
events, the analysis per se tends to be
pretty
486
transparent, but in an observational
study, in
order to understand it in detail, there
are many
covariates, pretty fancy footwork in the
statistical realm, and it may not be very
easy to
tell exactly what was done or to think of
reproducing it.
So, the observational study
tends to be a
lot less transparent in terms of the way
it has
been analyzed.
DR. WOOD: Dr. Friedman.
DR. FRIEDMAN: Two points.
One, if I can
follow up a little bit on this wild arm,
if you
will.
As Dr. Wood knows well, this whole issue
came up, to my dismay, if you will, about
a year
ago when we were dealing with the ARDSNet
issue,
and I think the general conclusion there
was that
it, in general, is not a very good way of
answering
a specific question. It might contribute in some
fashion, but in general, it is not all
that
helpful.
Second, I am looking at the specific
question here, and it says discuss the
role of
487
observational studies in informing
regulatory
decisions about postmarketing
safety. It seems to
me that one of the things we might do is
suggest
ways that the FDA can improve some of the
postmarketing surveillance issues.
For example, we have talked
about all of
the difficulties in using observational
studies,
and I don't disagree at all with any of
them, but
if some of them are planned ahead of
time, with
good ways of collecting data in
consistent ways, we
won't completely eliminate all of the
problems, but
we can reduce them, and I think we ought
to at
least consider that approach.
DR. WOOD: Dr. Platt, last comment on
this.
DR. PLATT: To emphasize that point,
taking everyone's thoughtful comments
into account,
it seems to me we have to be careful not
to let the
best be the enemy of the very good. I think that
Tom Fleming's reference to the CDC's
large databank
for vaccines is quite on point. It seems to me
that there is every reason for FDA to
require, as
488
part of the approval process, that there
be a
substantial and organized observational
set of
studies that give at least a sense that
generates
hypotheses that would allow us to
recognize the
possibility that there is a signal of
events that
never be seen in clinical trials, events
on the
order of 1 or 2 or 3 per 1,000.
It is possible to do that with
what in the
scheme of these discussions we are having
would be
a relative small investment, and we
wouldn't have
to rely on the occasional observational
trial or
the clinical trial that shows up to start
a
discussion like this.
It seems to me that that is a
very easy,
relatively small step for CDC to take, to
have
every manufacturer of a new drug commit
to doing a
reasonable observational study.
DR. WOOD:
But, Richard, isn't that the
problem that Tom highlighted ages ago,
that that
sort of registry approach will pick up
events that
are relatively rare in the background,
like
devastating encephalitis or something
like that
489
relatively easily.
But where you have got a
background noise
that is as high as MI, it is going to be
extraordinarily difficult to pick that up
from that
kind of study.
DR. PLATT: Well, in the vaccine field,
the large-linked database has been
extraordinarily
useful for things like febrile seizures
after a DPT
immunization, and that is a relatively common
event.
So, I don't take the point that you can't
make reasonable observations about even
relatively
common events.
DR. WOOD: Bob, do you agree?
DR. TEMPLE: Well, just to make the same
distinction you were making. You can look for
introsusception or something that
basically is very
unusual, but how to find an increase in
the rate in
the rate of MIs requires a structured
study and a
plan to do it, and you sort of have to
have a
hypothesis or you don't know what to look
for.
It is totally different from
liver, you
know, from gross hepatotoxicity, which
comes in
490
through the AERS pretty well actually,
maybe you
could stimulate those, but it is totally
different
when you are looking at a change in
something that
has a high background rate.
DR. PLATT: The fact that it's challenging
doesn't mean that you can't learn
something useful,
and it is pretty clear from the
observational
studies we have that we can learn
something useful
about that.
DR. TEMPLE: I was reacting to what you
said, should we have the capacity or have the
ability to get people to do studies once
something
emerges or once a question arises, or
once you know
something about the drug class, I am not
challenging that at all, that's fine, but
to have
it in place as a mechanism for sort of
automatically putting stuff up, I guess I
don't
know what that mechanism is.
There has been talk about
encouraging
places to report, and we have an arrangement
with
some liver centers, and those things are
fine.
Those might be ways to find hepatotoxins
maybe
491
faster than we do now, but that still
doesn't
answer the question of a change in the
rate of a
common event, which is a fundamentally
different
problem, requires a study, not a report.
DR. PLATT: Well, the model of the
large-linked databases I think gets
around the idea
of having to have active reporting. I think that
there is a lot of ability to capture the
outcomes
that are of interest.
Obviously, you don't look for
every
outcome for every drug, but you can make
up the
list of things that you care about for
certain
classes of drugs, and it is possible to
use
automated systems to take you a long part
of the
way in understanding whether there is a
problem
that needs serious analysis.
DR. TEMPLE: Can I propose an alternative?
I think what you are really saying is the
thing you
are worried about with drugs, where there
is a high
background rate of something, is always
cardiovascular outcomes.
So, I think what you are saying is you
492
might want to look for any chronically
used drug at
cardiovascular outcomes, and that you
could
probably put in place.
DR. WOOD: Wait a minute. Are you
suggesting that we insist on a
cardiovascular study
for every drug that we get approved? I mean that
would make it prohibitive to approve any
drug.
DR. TEMPLE: No, no.
We used to fund more
of them than we do now, that's a problem
that other
people will discuss, and certainly I
won't, but we
have access to databases, whether it's
California
Medicaid or whatever, and one can do
that.
It doesn't seem inconceivable to
me--and I
am talking about something that other
people know
more about than I do, so I should
probably shut up,
but I won't--I can imagine that a couple
of years
into the approval of a drug that is
widely used,
you could ask the question at certain
sites, can we
see an increase in cardiovascular risk.
I am not sure how many other
high
background events it is that are common
in the
population that we are really that
worried about.
493
Maybe that is something that we could
think about.
DR. WOOD: So, if we could just sum up
where we are, what the committee is
saying, I
think, is that we are impressed as the
primary data
source, and that the primary data source
should be
randomized and controlled trials, and
observational
studies may be good for hypothesis
generation, and
I guess the third point is that the AERS
database
is of almost no value in detecting
adverse events
that are common in the background in a
situation.
Is that sort of fair for what
we have sort
of got out of this? Do people disagree with that?
Yes, Dr. Farrar.
DR. FARRAR: There is one specific point
to this question, which is that all of
the
non-experimental studies that have been
presented
here, I would certainly suggest, and I
would hope
people would agree, are hypothesis
generating at
best.
Every single one of them is confounded by
indication.
The best example is the
indomethacin one
where it is only used in people who are
sicker than
494
people who aren't. So, I think there are clearly
examples.
What I was hearing before was a
discussion about what we might do, and I
just
wanted to be clear that what we might do
is very
different than what we have right now.
DR. WOOD: You put it much better than I
did.
That is what I was trying to say.
Yes, Dr. Jenkins.
DR. JENKINS: I found this discussion to
be very interesting because I think you
all know
there has been a lot of Monday morning
quarterbacking about what FDA has or has
not done
in this class, and a lot of that has been
based on
observational study results, many of which
fall
into the range of what we have been
calling small
to moderate, I think, at best.
I don't think we need to
revisit that
here, but I think the questions we have
going
forward, first of all, we have to look at
the data
set we have today, and you have to look
at the data
set you have tomorrow on answering the
questions
about what do we do now, and
observational studies
495
are part of that data set. We have controlled
trials that are part of that data set
also.
I think we are also interested
in hearing
your thoughts on going forward. I suspect that
this going to be a mining exercise for
everyone who
does observational studies in the world
probably.
They are going to be looking to do
another COX-2 or
another NSAID observational study.
We are going to see more and
more studies
published, and as I think someone said,
it often
becomes attractive to say, "Oh, look
at that, you
have got a very small p value, yeah, the
relative
risk is only 20 or 30 percent or 40
percent, the p
value is very small, the study was very
large, FDA,
you should take regulatory action, you
should take
this drug off the market, you should
restrict its
use, whatever."
You are telling us you view
them primarily
as
hypothesis-generating, and that they should lead
to controlled clinical trials. The reality is even
if we have the authority that we might
like to
mandate those trials, it is going to take
years to
496
get those controlled clinical trial data,
and there
is the pressure between people wanting
you to act
based on the observational data versus
the
scientific desire to wait until you get better
controlled clinical trials.
I would be interested in having
the
committee say a little bit more of your
thoughts
about, you know, what do we do in the
future in
this class when we get the next
observational study
that is touted as wow, this really shows
something,
FDA, you should take action.
DR. NISSEN: Can I suggest some courses of
action? One of them is that as people
have pointed
out, the strength of the association, I
mean the
hazard ratio is really important, and if
somebody
comes up with something which suggests 2
or 3, that
is very different from a 1.3.
The other obviously is to have
a rigorous
process for looking at the quality of
it. One of
the things I have learned from several of
you is
that there is observational studies and
then there
is observational studies, and some of
them are done
497
very well, and some of them are not done
so well.
The FDA has the expertise to
evaluate
that.
Now, the problem is, of course, if it gets
into the political arena, you get a lot
of
political pressure, but what we would
want you to
do in the public interest is look at the
strength
of association, look at the quality of
the study,
and make a decision on whether there is
enough
there to put a warning out.
We have seen some strange things go on,
like the warning around naproxen, that
was clearly
based upon pretty weak evidence. So, I think
having a good standard is where you have
to kind of
hold your ground.
DR. WOOD:
The other thing, in response to
your question, is if we walk through the
scenario
here, the first signal was from a
randomized
clinical trial, and the question I guess
then is
what would we need to strengthen that observation
because it wasn't against placebo and all
the
problems there were with it.
It would seem to me that we
don't need is
498
a bunch of observational trials. That hardly is
going to convince anyone, it seems to
me. What we
do need is an appropriately powered
randomized
trial that looks at the issue directly,
and I am
not so sure how long that would
necessarily take.
It only took 2 1/2 thousand
people and
approved to get the data. The question to which we
don't know the answer, in fairness, is
would it
have taken less time if we had done a
larger study,
and I don't know the answer to that, no
one knows
the answer to that, but it is certainly
potentially
possible that we could have gotten the
data quicker
if we had done a larger study and the
effect
appeared faster.
We don't know the answer to
that, but that
is one approach. I guess, responding to your
question, it certainly seems to be in the
public
interest that you should have the power
to ensure
that that kind of a study gets done, and
that is
something certainly people should hear
and hear
loudly, I think.
DR. O'NEIL: Could I say something
499
relative to a point that Janet Elashoff
had brought
up?
The general process for the review of
randomized controlled trials, such as the
ones we
have been reviewing, is we have the data
in, there
has been a strong movement for
prospective
specification of events, even blind adjudication,
we look at the protocol very
seriously. We
actually have the data in hand. We actually can
re-analyze, regroup, adjust, stratify, do
many
things.
We are normally not in a
position to do
that on observational studies. We don't have the
same level of process review for ran
observational
study.
In fact, it is not even clear what the
prespecified hypotheses were, even if you
wanted to
say the best that the observational study
could do
is to generate a hypothesis. However, there are
many of them that have confirmed
important things
for us, the last of which was a protocol
that we
played a heavy role in, and that was
propanolamine
and its association with CVAs.
That was a five- or six-year
prospective
500
case-controlled study that was done, that
we
reviewed the protocol. We had a heavy hand. In
fact, David had a heavy hand in how that
was
designed, and that turned out to
essentially
support a regulatory conclusion.
The point I am making here is
that if we
do open the door for observational studies,
we have
to have a different way of actually
having access
to the data, the quality of the data, and
give it
the same level of attention that we do in
the
review of randomized trials but for the
fact that
it's not randomized.
Right now that is not in place,
so we are
talking about trials being balanced
against
observational studies where the standards
for the
trials are dramatically higher than the
standards
for the observational studies, not that
they
couldn't be better balanced, but I think
that is an
important issue.
There has been a society, ISPE,
the
International Society for
Pharmaco-Epidemiology has
tried to put good principles in place to
sort of
501
say these are how you would do these
studies, but
we really don't have a process that would
require
that along the same ways that we would in
these IND
type studies or the larger randomized
trials that
we are seeing for the safety.
DR. WOOD: Tom.
DR. FLEMING: Just to reinforce some
comments that Bob was just making, and
Larry
Friedman was making earlier, and Steve
Nissen, as
well.
Not all studies are the same, we know that
is true of randomized trials in terms of
their
quality, it is certainly true in
observational
studies.
Stuart Pocock more than 20
years ago put
forward criteria for what you would want
to do if
you were doing an observational study
that would be
as reliable as possible.
Essentially, it is just like a
randomized
trial, it is very complicated and takes
considerable effort to ensure that you
are putting
in the structures. You can then have the
sensitivity and specificity issues
assessed or
502
addressed by independent committees. You can do
your best to try to define time zero
cohorts.
You still don't have
randomization,
though, and ultimately, the level of
reliability is
increased, but it still doesn't match the
reliability of a randomized trial, as
Charlie
Hennekens was saying, until you are
persuaded that
the signal exceeds the potential
magnitude of the
bias, you can't be confident that the
result is
reliable.
So often what we are looking at are effect
sizes that aren't, in fact, larger than
the
magnitude of the bias, so that leads us
down the
pathway of needing randomized trials.
John, getting back to your point,
if you
have a profoundly low p value, this may
be obvious,
but it doesn't mean we know the
truth. There are
two fundamental aspects around the
truth. One is
variability and one is bias, and I can
have 100
trials put together and give me a highly
precise
estimate.
I mentioned yesterday, you just end up
with a precisely biased estimate, and
that is my
503
concern in the absence of randomized
trials.
I believe these are very useful
clues, we
need these results, but just because you
have
profoundly low p value doesn't mean we
got at the
truth.
DR. WOOD: That is what happened with the
estrogen studies, of course.
DR. HOLMBOE: I just want to make one
point because we keep hearing about the
estrogen
study.
There is one very important fundamental
difference here. Estrogen had been posited to have
a positive effect on cardiovascular
mortality in
observational trials, so it made a lot of
sense to
use randomized controlled trials to prove
that
hypothesis.
The hypothesis here is that
COX-2
inhibitors are harmful, therefore, you
are doing a
randomized controlled trial that in
investigating
harm, not benefit, and I think we have to
keep that
in mind.
DR. WOOD: Good point.
Now, we are going to move on,
Steve, to
504
the next question. The next question is discuss
the available data regarding the
potential benefits
of COX-2 selective nonsteroidals versus
non-selected nonsteroidals, whatever they
are, and
how any such benefits should be weighed
in
assessing the potential benefits versus
the
potential risks of COX-2 selective agents
from a
regulatory perspective.
DR. JENKINS: Dr. Wood, could I make a
comment about that as you get started
about this
particular discussion point? We put this in here
for a reason, because clearly, we didn't
want a
three-day meeting to just focus entirely
on risk,
because the decisions you need to give us
advice on
have to be balancing risk and benefit.
I think here we are
particularly
interested in hearing your views about
benefit in a
wide range of categories.
You know, this class of drugs
was
developed for the GI effect, so we are
interested
in hearing your conclusions about the
benefit of
these drugs on the GI toxicity, but there
is also
505
other areas. Any input you have on their efficacy
for pain relief for the treatment of
inflammatory
conditions will be useful.
I am also interested in hearing
your
comments about the value of choice. We heard that
from some of the people in the public
hearing
today, that, you know, don't limit my
choices, and
we hear that a lot from physicians, we
hear that a
lot from patients, but we often are also
hearing a
competing view that if you have got one
that looks
like it is safer than the others, then,
you don't
need the others, but that is at odds with
the idea
that people like to have choice, because
people
don't respond the same to every drug,
they may be
allergic to one drug or whatever.
So, in this context of benefit,
I would
like you to cover a lot of different
areas, and not
just to gastrointestinal benefit, but
that is
clearly one of the major focuses of
benefit here.
DR. WOOD: Okay.
Dr. Nissen.
DR. NISSEN: A couple of things. One is I
haven't seen any compelling evidence that
in terms
506
of pain relief, that the drugs are
actually more
effective, and if such data is available,
I would
love to see it, but I don't find it
there, so I
think that is a little easier for me.
I don't think we can minimize the
importance of the GI aspect. There is actually two
things, one of which was talked about
interestingly
by the public, but not necessarily by us
or the
companies, and that is, you know, patient
quality
of life and patient preference.
Any of us, I have certainly
taken NSAIDs
and gotten gastritis from them, and it is
not fun,
you know, having your stomachache, and
people who
have that every day, you know, there is a
suffering
related to that, that we heard from the
public, and
that has to be taken into account as we
think about
these drugs.
In addition, I would be the
first to say
that a GI bleed is not a benign
event. If these
drugs were drugs that were better for
treating
acne, and they caused cardiovascular
harm, that
would be one thing, but the events, the
GI events
507
here are serious events.
They are not as
life-threatening as a
stroke or a heart attack, but they can
be, and they
don't produce the permanent disability
that a
stroke or an MI does. You know, I take care of
people with heart failure, and if you
have had a
big MI, and your pump doesn't work, your
life is
changed, the rest of your life is going
to be
different.
Most people with a GI bleed
recover, and
so as I weigh these events, I don't
discount GI
benefits, but I have to give them less
credence
than the kind of hard, permanently
disabling
effects of MI and stroke, and I also
think we
absolutely have to factor in here the
sort of
suffering of patients who just don't
tolerate the
conventional NSAIDs, and I think that
compassion
has to come into our decisions.
DR. WOOD: Dr. Fleming.
DR. FLEMING: A great deal of the focus on
the data we have had presented to us
relates to the
cardiovascular risks and relates to the
confirmed
508
complicated upper GI events, so if I
start by
focusing on that, it looks as though in a
crude
estimate that we might be having the rate
of these
events using the COX-2 inhibitors rather
than the
non-selective NSAIDs.
It looks as though that might
be, in 1,000
people, preventing 5, 6, 7, 8 events,
something on
that order. If we took a relative risk of 1.4 as
the relative risk for the increase in
cardiovascular events, that would be
about 4
events.
So, coming back to what Steve
is saying,
when you look at it in that context, yes,
these
ulcerations are important events, but 7
per 1,000,
how is that up against 4 events that are
strokes,
MIs, or cardiovascular deaths, I don't
think it
adds up.
If that is the whole picture, I
would have
a concern, but in a number of settings,
it isn't
the whole picture. We have heard about the
oncology setting. We have talked about, truly, we
haven't talked about efficacy. We have only had a
509
number of comments stated that the pain
relief
seems to be about the same.
Well, if it is the same, then
that balance
that I was saying concerns me as not
being a
favorable balance, but we heard a lot of
people
testifying, and I will be the first to
say open
sessions at these meetings are not random
samples
of the entire public, but we still heard
a lot of
comments that reflected the fact that
there seems
to be some differential protection or
pain relief
in certain patients.
Can we quantitate that? Can we, in fact,
more scientifically, rigorously establish
certain
subpopulations where there really is a
differential
relief?
Then, the benefit to risk shifts, or in
the oncology setting, the benefit to risk
shifts.
The bottom line here, though,
is to me the
issue isn't so simple as choices. The issue is
informed choices, and it takes the kind
of
scientific studies to reliably identify
what are
the true benefits and risks, so that
patients are
in a position to make an informed choice,
and part
510
of the challenge to this, as one of the
speakers at
the public session pointed out, is it is
not always
the case that what might be learned by
those people
doing the studies is being effectively
transmitted
to the bedside or to the patients and
their
caregivers, and that is the other aspect,
as well.
So, it is critical to follow a
strategy
here that allows us to reliably address
benefit to
risk and allow patients to make an
informed choice.
DR. WOOD: Dr. Hoffman.
DR. HOFFMAN: I think for the last two
days we have been hearing appropriate
angst about
damning a class of agents for which there
is a
measure of efficacy, both in regards to
pain and GI
events because of newly-discovered
adverse events,
but I feel like we are walking on eggs in
trying to
get away from a consistent observation
that is the
dose-response effects, relative risks
that we are
looking at in terms of cardiovascular
endpoints.
We have heard this from experts
at the
FDA, independent investigators. We have even heard
it from the thought leaders of industry,
there
511
seems to be a consensus to the effect
that there is
a class effect.
I do take Steve Abramson's
point that all
of
these drugs are not pure in their effects in
terms of COX-1 or COX-2, but this is the
data that
we have, and it seems like there is a
consensus
about a class effect, and there also is a
consensus
in acknowledging that the patients that
we enter
into randomized controlled studies are
probably the
people least at risk that we may not see
in our
practices, who come in with 3 or 4
comorbidities
that may have excluded them from being in
this
trial and actually having seen even a
clear signal.
The data, of course, that we
would like to
have is something that we don't have, and
that is,
the old standards of treatment for pain,
whether
it's the arthritis pain of OA/RA or
postoperative
pain, with NSAIDs plus PPIs over a long,
extended
period of time.
We would all like to know the
data for
that over 2 or 3 years compared to the
COX-2s,
which I don't think any of us are saying
should, as
512
a class, be taken off the market, but
certainly
should be used at the lowest safest dose.
Now, at the lowest safest dose
we don't
even know their efficacy qualities. We don't know
whether at the lowest safest dose we have
the same
benefits in terms of preventing peptic
ulcer
disease, treating pain effectively,
decreasing
inflammation effectively, and that it
seems is the
data that we need to have.
I am a little concerned, as a
footnote to
that, about the issue of choice. I think it is our
obligation to provide patients choice
within the
realm of relatively safe medications, but
most of
us would not give as a choice a narcotic
analgesic
to a patient with, say, fibromyalgia.
I don't think we should keep
drugs on the
market because of public pressure if we
have a
signal that we feel is a very strong
one. We
shouldn't give people a choice if we
think that
choice is uninformed and potentially does
harm.
Now, I am not saying that for
the class of
COX-2 inhibitors, I am just saying that we
need
513
more data to be able to provide for
ourselves
adequate information to make that choice
and give
our patients informed choice.
DR. WOOD: Dr. Cryer.
DR. CRYER: We were asked to kind more
widely consider the potential
benefits. As I see
it clearly, one of the benefits is GI,
and I will
comment on it, but I do want to reiterate
some of
the comments that I personally don't see
the
benefit with respect to efficacy.
I think the clinical trial
experience to
date has pretty consistently indicated
that the
efficacy is similar to the traditional
NSAIDs. We
did see some provocative data with
etoricoxib today
suggesting greater efficacy in one trial
than
naproxen, but that wasn't replicated.
So, overall, I have to think
that the
efficacy is the same as we have with the
traditional NSAIDs. I appreciated the testimonials
of the patients about their individual
efficacy
responses, but my conclusion about that
is those
are anecdotes and it is consistent with
the
514
clinical experience that we have with
efficacy of
NSAIDs, which is that there is variable
and
idiosyncratic, unpredictable responses
between
patients, and it is very common that you
will have
one patient who responds to one NSAID and
does not
respond to another.
I do think that we would still
be giving
these patients a wide range of choices
given that
there are 20 other NSAIDs available in
the U.S.
among which they can choose.
The benefits clearly I think
are in the GI
tract, but I will say that my conclusion
is that
the GI benefits are less than previously
speculated.
If you look at the three
outcome trials
which we have, that looked at GI
benefits, we have
VIGOR, CLASS, and TARGET. The results in the VIGOR
are clear, but I think that was clearly
also of a
manifestation of the comparator, and one
of the
things that I would like to be remembered
is that
the comparator NSAID matters.
One sees a greater degree of GI
benefit
515
when one compares against naproxen than
when one
compares against diclofenac, so I do
think there is
value from the CLASS trial. I know that there was
a GI benefit shown against ibuprofen.
In the TARGET trial, those GI
estimates
are overestimated primarily because they
enroll a
low risk group of individuals and in a
lower risk.
We have consistently seen in trials that
when you
have low risk GI group, the relative risk
is higher
although the absolute risk in a low risk
population
is very low.
So, the benefit is going to
depend on the
comparator. It is probably less than the 50
percent that you suggested it to be,
because that
50 percent is based upon the VIGOR
trial. It is
probably closer to maybe a 30 percent
benefit that
I would estimate.
It also narrows when you
consider low dose
aspirin.
In the face of low-dose aspirin, there is
no apparent GI benefit. So, I think we also need
to modify our estimates based upon the
population
that would be using or not using low-dose
aspirin.
So, my conclusion about the GI
events is
that, yes, there is a benefit, it is not
as large
as we thought, the appropriate target
population is
516
smaller with respect to the target
group. It could
be low risk people not taking low dose
aspirin, but
this event doesn't happen very commonly
in low
risk, and when you look at the high risk
people in
whom these drugs were originally
targeted, several
data sets suggest that the high risk
people do not,
in fact, have any appreciable benefit of
GI risk
reduction from a COX-2 specific
inhibitor.
Final comments about other
areas of
benefit.
Dyspepsia isn't one that is very
convincing. When you look at the dyspepsia data
from the clinical trial experience, it is
only a
few percentage points reduced. Dyspepsia, I
consider mostly a nuisance symptom for
which we
have other very safe therapies to
effectively deal
with this.
Finally, from the GI
perspective, the
polyp story could be another potential
benefit, but
with regard to the polyps, we have to
remember in
517
every trial we have seen, we are only
modestly
reducing the polyps and ultimately, we
don't reduce
cancer risk unless we eliminate
adenomatous polyps,
so it doesn't really change our algorithm
in terms
of how we would manage these patients,
which would
be colonoscopy and polypectomy.
DR. WOOD: Before you finish that, there
are only two drugs on the U.S. market
now,
celecoxib and valdecoxib, so let's review
the upper
GI safety for them first.
Is there a study that you are
aware of
with valdecoxib looking at complicated
ulcers that
showed in randomized fashion that there
was a
safety signal?
DR. CRYER: No.
Wait, what do you mean by
safety signal?
DR. WOOD: GI benefit.
Is there a VIGOR
trial for valdecoxib?
DR. CRYER: No.
DR. WOOD: So, confining our discussion to
the two drugs that are on the U.S.
market, there is
no VIGOR equivalent, if you will, in
valdecoxib,
518
right?
DR. CRYER: Correct.
DR. WOOD: Now, for the other drug that is
on the U.S. market, celecoxib, the
published study
didn't show the full data set. For the full data
set for that, there wasn't benefit
either.
DR. CRYER: Correct, but we did have the
benefit of--
DR. WOOD: I understand, but there is
always a benefit--I mean there is
mortality
problems halfway through, too, that
disappear, that
we ignore when we get to the end of the
trial.
So, for the two drugs that are
on the U.S.
market now, we have no clear randomized
data that
show GI benefit given the endpoints that
were
predefined and the end of the trial, not
the trial
that was published without the complete
data set.
The TARGET trial looks at a
drug that is
not on the U.S. market. So, our job is to evaluate
the
two drugs that are on the U.S. market, it seems
to me.
DR. CRYER: So, I agree with your comments
519
about the fully published results in JAMA
for the
class, however, we did have the benefit
of
reviewing the full class results in the
FDA hearing
four years ago, and it is based upon that
evaluation that I am deriving my
conclusions of the
full data set in which there did appear
to be a
demonstrable GI benefit when compare to
ibuprofen
in people who were not taking aspirin,
certainly
not when compared to diclofenac.
DR. WOOD: But the trial was not--that was
a subsequent analysis taking out the
aspirin. That
wasn't the predefined endpoint.
DR. CRYER: Point well taken.
DR. WOOD: So, I mean just summarizing the
point again, we have a benefit in a trial
for a
drug that is not on the U.S market, but
we are not
prepared to extend a class effect to
cardiovascular
risk necessarily, so I don't think we can
just sort
of step back and say that we are going to
give a
class benefit to GI benefit either extrapolating
from studies of drugs that are not on the
U.S.
market.
DR. CRYER: Just because they are not on
the U.S. market does not reduce the
validity of the
observation, for example, with
lumiracoxib, and
520
just because this was not absolutely
predefined,
and the benefit was recognized in, let's
say, a
post-hoc perspective, I still think there
is
recognized benefit in the data that we
see in terms
of assessing the GI benefits of celecoxib
versus
ibuprofen, and lumiracoxib versus its
comparators.
DR. WOOD: But the non-aspirin group also
had a cardiovascular risk, right?
DR. CRYER:
Absolutely.
DR. WOOD: I mean as we are doing Tom's
sort of analysis, when we take out that
aspirin
group and say, wow, there is a GI benefit
there,
when we take out that aspirin group we
find there
is
a cardiovascular risk. So, you know, we
can't
have it both ways.
DR. CRYER: Well, I would say that the
cardiovascular risks extend to both
groups, aspirin
and non-aspirin.
DR. WOOD: Right, but it was clear
521
in--okay, Dr. Fleming.
DR. FLEMING: Just to pursue a bit
further, Alastair, what Byron is saying,
there are
two aspects that I hear you saying that
are really
critical to the comments that I had made
earlier.
One is that I might be
overestimating the
actual GI benefit when I say you are
having maybe
it's a 30 percent.
DR. CRYER: It depends on the comparator.
DR. FLEMING: But the other, even more
important thing to me that you are saying
is that
in spite of what might appear in the open
session,
which we know is anecdotal, the
scientific data you
are saying repeatedly are showing in the
RA, OA,
CABG settings where we have done studies,
that
there is not a difference in the pain
relief and
the efficacy.
I would like to get more sense
about that.
If
that is even close to true, then, there should
be an incredibly low threshold for what
you would
accept in additional cardiovascular
events, because
the only thing you are getting relative
to
522
nonspecific NSAIDs then would be a very
small GI.
So, it seems like the efficacy
here about
the pain relief is a key issue.
DR. WOOD: I think the company wants to
say something.
DR. KIM: Mr. Chairman, if I could, I will
just make a comment, please. As I said yesterday,
at the time that Merck withdrew Vioxx
from the
market, we based that decision on the
available
data that was available to us at that
time, and we
also stated that we thought that it would
be
possible to continue to market Vioxx with
a
labeling change that incorporated the
results of
the APPROVe trial.
But we decided and we concluded
that the
most responsible course of action to
take, given
the information that we had at that time,
and the
availability of alternative therapies,
was to
voluntarily withdraw the drug from the
market.
We have heard over the past two days new
data and we have seen in the New England
Journal
new data on some of these alternative
therapies.
523
Merck's interpretation, as you have
heard, of these
data are that we are dealing with a class
effect,
and the major question on the table right
now is
how large is that class.
We are a data-driven
company. If this
committee and the FDA agree that what we
are
dealing with here is a class effect,
then, I think
it would be important for us to take the
implications of that conclusion into
consideration
with regard to Vioxx, particularly given
the unique
benefits that Vioxx provides, one of which you are
alluding to.
So, I just wanted to make that
point.
DR. WOOD: So, just to understand, what
you are saying is that if we think the
cardiovascular effect is a class effect,
you would
consider putting Vioxx back on the
market.
DR. KIM: What I am saying is that at the
time we withdrew the drug from the
market, we did
so because of the availability of
alternative
therapies and the science that was
available at the
time.
That science has progressed. We
are now
524
engaged in a discussion around that
science.
There are unique benefits to
Vioxx, one of
which is it is the only COX-2 inhibitor
with proven
reductions in gastrointestinal events,
another one
of which it is the only coxib which is
not
contraindicated for patients with
allergies to
sulfonamides, and the third is that we
have heard
numerous reports, and you have heard a
few today,
from patients, including patients with
chronic
debilitating pain that Vioxx was the only
drug that
relieved that pain.
DR. WOOD: Okay, good.
Dr. Farrar.
DR. FARRAR: I wonder if I could just be
very clear that so far I don't think we
have talked
about benefits. The point I want to make is that
what we are talking about with the GI,
quote
"benefit" is, in fact, a
reduction of risk. No one
that I know of takes coxibs of any kind
for an
upset stomach.
I think what we need to do is
focus on the
benefit to the patients, and we heard
some of that
525
in the public forum today, and I want to
be as
clear as possible about the issue of that
benefit.
There are two ways of measuring
benefit,
and, in fact, in outcome trials, there
really are
only two summary statistics that are
possible. One
is a mean or a median or some central
tendency with
a spread, standard deviation.
The second is a proportion, and
it is a
proportion of responders, it a proportion
of people
who die, which is the easiest, and in
pain
management, we get into all kinds of
arguments
about how much improvement you have to
have to be a
responder.
If you look at the data, we are
used in
most of our clinical trials to looking at
means and
standard deviations, and if you look at
means and
standard deviations, it is very hard to
find a
difference between any of the NSAIDs and
acetaminophen, any of them.
If you ask patients about what
works for
them, in clinical practice, every patient
will tell
you that one works and that one
doesn't. "I get
526
sick with that one, I don't get sick with
the other
one."
That is not something that we
measure
typically in our clinical trials. If you look at
what level of drug is effective, with
almost any
NSAID, it is never, it is never above 50
percent in
terms of patients who actually go on
using the drug
in a chronic process.
What we are talking about is
trying to
identify less than 50 percent of a
population who
respond to a drug, and I can tell you
from clinical
practice, as any of you who have taken
patients
with rheumatoid arthritis know, people
like
specific drugs because they don't cause
side
effects and because they do have an
effect.
I think choice actually is a
very
important issue. Granted, we don't want
to provide
choice if there is an absolutely huge
risk
associated with that choice, but I think
it is
really important to understand that pain
kills in
the same way that the drug potentially
can kill.
I think it is very important to
understand
527
those two principles, the principles of
the
difference between a proportion and a
mean value.
Now, I am obviously talking to the
converted here,
but I think the issue really is looking
at those
issues.
We don't have any good trials, any that
look at switching behavior within our
patient
populations, so there is no data that I
know of
that will help inform us about the need
to and
exactly how to go about this process, but
I do know
that in spite of all of our understanding
of what
goes on with the COX-1/COX-2 pathways and
the
inflammatory pathway, that when it gets
down to
using it in the patient, the issue is, is
it
absorbed, does it cause local effects,
does it get
to the active site, once it's at the
active site,
are there enough receptors for it to then
cause the
effect that we are looking for, a whole
host of
factors that we really can't measure and
haven't
measured yet in terms of metabolic process.
My honest sense from the data
that we have
heard here is that the drugs that we are
528
considering today, the two, perhaps
three, has to
do with the relative benefit of those
drugs.
What is very clear is that
there are
people, and a large portion of people,
who have
trouble with the current list of what we
call
non-selective COX inhibitors, and that
there is a
very important role for the more
selective COX-2
group, however we want to define that.
I think it is also, however,
very
important to understand that not
everybody should
be
on a COX-2 predominant agent, and one of the
problems that we are struggling with
right now is
the fact that because they were marketed
as being
safer, there was a very large push to
switch people
over who may not have needed to be
switched.
So, I think that the issues
that we need
to consider are there is very good data
that these
drugs are effective at least in some
segment of the
patients in whom they are tried.
There is I think reasonable
data to
suggest that the potential risks is not
clearly
very different between them, at least not
the data
529
that we have to date, and that from that
perspective it is going to be important
that we
carefully think about how we then go
about
controlling those drugs.
I would end with just saying
that I agree
absolutely it is about informed choice,
and that I
think that there needs to be a fairly
large amount
of information in the label and
information
conveyed to patients and physicians to
help them
make those choices.
DR. WOOD: Dr. Gibofsky.
DR. GIBOFSKY: I am particularly pleased
about the nature of the conversation
because as a
student of medical history, it reminds me
that the
first treatment we had for arthritis was,
of
course, willow bark, and we told our patients
to
ingest willow bark in order to get
salicylates,
which, of course, have an
anti-inflammatory effect.
So, if only our patients could take
aspirin,
perhaps we wouldn't need the whole class
of
non-selective and selective COX-2s, but,
of course,
they can't. There are problems just with aspirin
530
in the treatment of arthritis at the
doses they
need it.
I am intrigued by the comments
that, well,
you know, an MI is an MI and you are
dead, but a GI
bleed, you get up, you get over it with
no long
lasting effect, and that may be true for
the people
who survive, but as Dr. Cryer showed us
yesterday,
and the best data set we have from Dr.
Singh, 16
percent of patients who have a GI bleed
die, so for
them, it's a fatal event and one that
they are not
going to get up and continue on.
I don't want to get into a
discussion of
the GI benefit and whether, in fact, it
was
achieved with one agent versus another,
but what is
clear is something that hasn't been
remarked yet,
and that is for patients going to
surgery, who are
going to require anticoagulation
following their
surgery, and that is particularly in
large part
patients who have arthritis and are
undergoing
joint replacement surgery, the risk of a
traditional nonsteroidal with an
anticoagulant
appears to be far worse in terms of
bleeding later
531
on than the risk of being on a COX-2
because of the
lack of platelet inhibition.
So, certainly there is a
benefit for
patients in that group who are going to
go to
surgery and require concurrent
anticoagulant.
With regard to the issue of
patient
choice, there is several sets of
data--and we heard
one--showing that when you give a patient
two
different medications, in one study, the
ACDA
study, looked at acetaminophen versus
diclofenac,
another one, the PACES study, looked at
celecoxib
versus acetaminophen, and you asked
patients
without knowing which drug they were
getting, in
which arm, patients expressed a
preference for
either diclofenac or celecoxib over
acetaminophen
in the treatment of their arthritis.
The other issue with regard to
choice is
that we have also recognized, even in the
pre-COX-2
days, that not infrequently, patients
develop what
is called a tolerance to the agent that
they were
on, that the latest data set we had
suggested that
inside of 18 months, patient who were
taking
532
medication for their arthritis
chronically had to
be rotated among agents three to four
times in that
period of time.
So, the necessity for multiple agents
in
our armamentarium, the necessity for
agents that
allows for this individual idiosyncrasy
that we
have heard of is quite important.
As was alluded to, there can be
two
patients in the waiting room on the same
drug, one
will swear by it, one will swear at it,
and so it
is for that reason that we need to have,
not just
one agent in a class, whatever we define
that class
to be, but sometimes several. Sometimes they are
agents of allergy or idiosyncrasy which
necessitate
having more than one agent available.
I think it is for all those
reasons that
we have to consider that in the benefit
part as
long as we are discussing benefit in the
last part
of the day.
DR. WOOD: I think we have to be really
careful accepting this data, this
15-year-old data
from Dr. Freis. I mean he has published, he
533
published multiple updates on that, and
people keep
showing that same data, and that data
isn't what is
in his latest revision.
DR. GIBOFSKY: Accepted.
Dr. Cryer?
DR. CRYER: I would like to
comment on
that, and I think your point is well
taken. While
I showed the 16,500 data yesterday, at
the same
time I said that that estimate, based
upon more
recent evaluations, is probably an
overstatement of
the actual mortality risk, GI risk
attributable to
NSAIDs.
Dr. Singh has showed me more
recent data
which he has conducted in the U.S., which
has shown
that the risk has dramatically decreased
in the
U.S.
That is probably related to several factors
included in which is the eradication of
HP, the
introduction of PPIs into the U.S.
marketplace, as
well as the introduction of COX-2
specific
inhibitors.
The most recent estimates that I
have seen
would suggest that the mortality is about
half of
what Dr. Singh previously suggested it to
be.
DR. GIBOFSKY: Accepted, but even the
mortality rate of 8 percent in a
population is
unacceptable.
534
DR. CRYER: It is not 8 percent, it would
be 8,000.
DR. WOOD: It is much lower than that, and
if you look at the curve, the fall
occurred long
before COX-2s were on the market.
DR. CRYER: You are correct.
DR. WOOD: The data are out to 2000 on his
paper, and that fall had occurred by
1998, so that
is
before any of these drugs were on the market.
My point is that we keep
throwing this
100,000 number around, including from the
industry
people, when the data is 15 years old,
and the
author has updated it multiple times, and
that is
not reasonable, guys.
Dr. Singh.
DR. SINGH: As the author of the papers
that you are discussing--
DR. WOOD: I am talking about Dr. Freis's
paper, which was actually published. Yours is an
535
abstract, I think.
DR. SINGH: Also, the 16,500 was from my
paper that we estimated with the Aramis
data set,
and that, you are right, it is not 15
years old,
but that is about '94, '95 data, and now
that we
have newer data sets, that was an
estimate from the
Aramis data.
The latest work now is actually
on real
hospitalizations based on the nationwide
inpatient
sample, which is a much better estimate
of what is
really happening than an estimate from a
small
patient population.
When we go back and look in
'93, '94, of
what the total number of deaths that the
Federal
Government said occurred in the United
States, we
were off by 32, that's it. It was like
16,486.
That is how far we were off by, just to
let you
know in terms of an estimate.
This is also true that now,
today, the
latest data set that we have available
from 2002,
that has dropped significantly, and the
death rates
are more like 8,000.
But the other place where we
underestimated was the hospitalizations. We
underestimated the hospitalizations, they
are not
536
108,000, there are a lot more than that.
The mortality rates today have
gone down
tremendously, and the mortality rates today
are
probably more in the 5 to 6 percent
range, and that
is where Byron is correct, as well.
Then, as far as the trend is
concerned,
the data that I showed you today is based
on 483
million hospitalizations. We are not counting
about 50 hospitalizations and then
extrapolating it
to the country. There are 483 million
hospitalizations and 3.68 billion patient
years.
Yes, the trend line started
going down way
before the COX-2s were introduced, but
then there
are two sharp years of decline. The trend line
actually, if you look at my slide, is
very
interestingly correlated with PPI use,
and I showed
data to Byron from the same data set,
that it also
explains it very nicely because the
duodenal ulcer
rates have gone steadily downward, which
would be
537
attributed primarily to PPI use and H.
pylori
eradication therapy.
The gastric ulcer rates and the
gastric
ulcer hemorrhage rate have not gone down
in the
same fashion. They went down when the '94-'95 H.
pylori eradication campaign started. Then, they
plateaued off pretty much, and PPIs
haven't done
very much to gastric ulcers until 1999,
when the
gastric ulcer rate dropped dramatically.
In 1999, there is a 22 percent
drop per
100,000 prescriptions sold in this
country. I
don't know what it is because of. Coincidentally,
in 1999, January 1, celecoxib was
introduced. I
don't know what it is because of.
DR. WOOD: Let's move on.
Dr. Dworkin.
DR. DWORKIN: Much of what I wanted to say
has already been said, but I just want to
emphasize
that while there are no differences on
average in
pain relief amongst these drugs,
certainly none
that are replicated, as Byron pointed
out, that
there is a great deal of variability in
response,
538
and I think there is every reason to
believe that
some patients respond better to one drug
than
another, so you have variability in the
pain
benefit, and you have to consider at the
same time
there is variability in the tolerability
of the
drug.
So, there are two sources of
variability
in patient response, which at least to my
way of
thinking provides a really solid basis
for there
needing to be a choice amongst several
drugs,
because you have the variability in the
pain
benefit amongst patients and the
variability in
their tolerability.
DR. WOOD: Dr. Cush.
DR. CUSH: I prefer to say that these
drugs are equally potent between the
COX-2 specific
and the non-selective drugs. I think there is a
variability, but that speaks to the need
for
choice.
Every rheumatologist at this
table will
tell you they cannot manage in any
effective or
compassionate way osteoarthritis or
rheumatoid
539
arthritis using just Tylenol and aspirin
and
ibuprofen. That would be a gigantic step
backwards.
So, they are equally
potent. I think when
it comes, however, to the risk,
thankfully, this
risk is incredibly low, but we would like
to make
it lower, and what we need to put forward
is that
we need a strategy for risk modification
that is
going to extend to all these drugs that
we are
examining here, much in the same that
occurred with
GGI, I think that we can start with some
recommendations and then make it the
responsibility
of the manufacturers to come up with
studies that
will further define how we can best
reduce the risk
in people who may need to receive these
medicines.
DR. WOOD: Dr. Morris.
DR. MORRIS: Let me focus on the question
that asks about the weighting, because
what we have
is--I guess everybody interprets this
question
differently, but what I interpret it as
is how do
you look at these non-comparable outcomes
and how
you trade off a TIA from a gastric ulcer
or
540
something.
I think what we can do is we
can describe
the effect and we can describe the
probability of
the effect, but what we don't know is
what is the
right way to weigh those things, and I
would make a
plea that probably the right way is to
try to
involve in some way the views of patients
in that
decisionmaking.
I don't mean that
qualitatively, I mean
that quantitatively, is in quality of
life type
data where people have looked at various
outcomes,
looked at it on a single scale, and apply
some of
those ways, so we understand how patients
view it,
and go beyond just medically what we
think patients
should evaluate it, but how they actually
do
evaluate it, and try to use some of the
input of
those data.
That literature suggests that
we get it
wrong, that there is things worse than
death, and
we always think of death as the worst
thing to
happen in a medical outcome, but yet from
a
patient's perspective, being paralyzed by
a stroke
541
is perceived as worse, and we need to
understand
patients' evaluation of these outcomes,
so we can
make those weightings better for them.
DR. WOOD: Ms. Malone.
MS. MALONE: Obviously, this is
complicated. I agree with most of what the
previous speakers have said especially
Dr. Cush,
Dr. Gibofsky.
A big problem is like Dr.
Gibofsky had
said about having choice and trying
different
drugs, and having a period of time when
they would
work, and then they wouldn't be as
effective and
you would have to try something else.
That is why the need for choice
is there.
I have spent the last 35 years probably
on each of
the NSAIDs that are still available, and
went
through that, and the frustration and the
pain, and
just--it's very difficult, so when people
give this
anecdotal information and say that they
have found
something that works for them, they are
going to
fight for that.
We have to be able to prove to
them that
542
the risk far exceeds the benefit, and we
have to be
able to show that, and we can decry
anecdotal
evidence as not being sufficient enough,
but, in
reality, it all comes down to anecdotal
evidence.
It all comes down to the personalization
of it,
what happens to me when I take this drug,
what
happens to me when this drug is not
available.
But I think behind everything is the whole
element of trust, and they place their
trust in us,
in FDA, and we can't give in to pressure,
okay, but
we can't give in to pressure either
way. We have
to keep an open mind about it and realize
what they
are going through and try to put
yourselves in
their shoes.
DR. WOOD: Dr. Platt.
DR. PLATT: In the spirit of supporting
informed choice, it seems to me we could
do a very
much better job than we do by using the
existing
data that FDA already has to provide good
information to patients about the risk
stratum that
they inhabit.
Saying that there is an overall
1 1/2 or 2
543
percent difference in the risk of a GI
complication
or myocardial infarction is not doing the
best
service to most people who take those
drugs.
I would imagine that those data
can be
used to support predictive modeling that
would
allow a fair amount of discrimination so
that
individuals could be told that people
like them can
expect a risk of 1 in 1,000 or 1 in 100
or 10 in
100, and that would make it a lot easier,
I think,
for individuals and their doctors to make
thoughtful decisions about the tradeoffs
of the
benefits and the risks.
It seems to me those data are
there and it
would be a straightforward thing to make
them
available. We do that with breast cancer all the
time.
The NIH did a tremendous service I think to
the public by providing good predictive
models that
let women know what their risk of breast
cancer is
to help them decide whether to take
preventive
action.
I think we could do it with these drugs.
DR. WOOD: Dr. Bathon.
DR. BATHON: It is interesting that you
544
would say that because that is, in fact,
what most
of us rheumatologists have been doing for
the past
four months with every single clinic
visit, is
weighing the benefits and the risks based
on the
data that exist right now, and it is a
difficult
endeavor.
I think that we are really
hearing from
our patients, and we heard this today, we
are in a
different era of patient-doctor
relationships, and
patients want to be a collaborator in these
decisions, and they want to know the
information.
I think that the way I am
thinking about
this problem right now is that these
drugs, whether
they are selective or non-selective, are
another
risk factor in the GI complications and
the
cardiovascular complications that we have
to weigh
along with their blood pressures, their
diabetes
status, their BMIs, their family history,
and
everything else to come to a final
decision about
what we recommend with their input.
Until we see an unequivocal
cardiovascular
risk that outweighs all those other
factors, I
545
think that is the appropriate approach
with the
patient is to put the drug in with all
the other
risk factors and try to come up with the
best
benefit-risk ratio that exists for that
individual.
DR. WOOD: Dr. Hennekens.
DR. HENNEKENS: I find Question 3
extremely complicated in a number of
dimensions. I
am attracted to Tom's formulation of
benefit to
risk, but I think we also have to
consider these
arthritis patients with regard to the use
of
selective coxibs.
As a group, they are at maybe a
double the
risk of heart disease of their
non-arthritis
counterparts. They are also suffering terribly
with pain.
From that perspective, the data we saw
over the last two days on naproxen was
somewhat
reassuring to me, but for the patient who
has
gastroesophageal reflux disease or an
allergy to
aspirin or non-selective NSAIDs, I think
there the
benefit-to-risk obviously shift although
even here,
I think they have to have their
cardiovascular risk
546
factors managed aggressively, and I would
add three
more dimensions.
One is I am not reassured at
all by the
data that are available on the
short-acting
non-selective NSAIDs with regard to risks
and
benefits, and I think we need a lot more
data
there.
I am also not reassured by data
we haven't
reviewed that acetaminophen is either
sufficiently
efficacious or much safer, and then
finally, the
problems with high doses of aspirin are
real.
I do point out, though, the UK
TIA trial
of 2,400 people that gave aspirin 1,200
mg in a
placebo-controlled design for 5 years,
the rate of
GI side effects attributable to the
aspirin was 14
percent, significant bleeding was 3.3
percent, but
this flies in the face that 25 percent of
the
people on placebo had GI side effects and
1.6
percent of them had a significant GI
bleed, so I
think nothing is straightforward here.
DR. WOOD: Dr. Nissen.
DR. NISSEN: Just one brief comment, and
547
that is, one of the things I am
struggling with for
all of you, and maybe some of those that
either
deal with these diseases can help me with
this, is
that the people at greatest risk for GI
bleeding
are the older and more frail individuals
who are
also at the greatest risk for
cardiovascular
disease, and so finding the sweet spot
for the
drugs becomes a little bit harder.
There obviously are certain
populations
where it is obvious, but the big
populations where
there is risk, is it not true--I think I
heard from
Byron that older people are at greater
risk for GI
bleeding, and I can assure you they are
at greater
risk for coronary disease, so the
question is how
does it tilt in any given patient. It is not so
easy to figure it out.
DR. PLATT: But you can quantitate it. I
mean it seems to me you could tell the
patients
what individually, approximately what
they could
expect on both dimensions, and for a lot
of
patients, they would be high on both, but
at least
they could make an informed decision
about that.
DR. CUSH: But it's the same situation as
the GI problem. We know what the risk factors are,
and age is a risk factor, and we counsel
patients,
548
and we probably should tell the ones who
might be
willing to accept some small risk,
because they
don't seem like they are at risk just
because of
their age, but they don't have any other
factors,
and the same thing can happen here with
regard to
the cardiovascular risk if we have some
appropriate
guidelines.
DR. CRYER: Steve Nissen, I think you have
got it exactly right and that there seems
to be a
great degree of overlap in those who are at
GI risk
tend to be, not uncommonly, the same
patients who
are cardiac risk. They are older, they may have a
previous history of cardiovascular
disease, and
other risk factors which are common to
both risk
considerations, GI, and cardiovascular.
DR. WOOD: Ms. Malone, do you want to say
something?
MS. MALONE: Yes, I do.
Just what Byron
has said, all of that brings in the
importance of
549
the doctor-patient relationship, and
today, with
the health care climate that we have, I
have heard
patients say how difficult it is to go in
and get
an amount of time when you can talk to
your doctor,
have a relationship with him, and
especially, as
people become older, and where I live in
South
Florida, there are many elderly people
who do not
have family around, so they are going to
their
doctor by themselves, and they are
dependent on
that doctor's viewpoint.
They will say, "Well, what
do you think?"
I used to say if I were your child, and
then it was
if I were your wife, now it is getting to
be if I
were your grandmother, you know, with the
age of
everyone, and I hope I live to say if I
were your
granddaughter.
But that is very true, and
again it is not
a simple situation, and whether we need
some sort
of health educator to assist the doctor
to be able
to explain this to the patients, so that
they are
not taking valuable doctor-patient time,
but
something needs to be done.
DR. WOOD: Thanks.
Dr. Ilowite.
DR. ILOWITE: I wanted to talk to a few
550
pediatric issues about these agents, the
granddaughter. First of all, about choice, there
are far fewer choices in pediatrics.
There is only
three NSAIDs approved, only two liquids
and none
any longer that are available as
once-a-day dosing
regimens.
The second issue is about
tolerability.
Certainly, children have fewer serious
gastropathic
events, but they do have a lot of
symptoms, and it
is often difficult to get children to
take
medications that give them even
bellyaches.
Third, is the risk of cardiovascular
disease, which is very low in
pediatrics. A new
clinical research network called CARRA,
Childhood
Arthritis and Rheumatology Research
Alliance,
organization polled its 130 members of
whom 92 or
71 percent responded, and there were no
events of
myocardial infarction or stroke that
couldn't
otherwise be accounted for easily that
were
551
attributable to these agents.
Lastly, is the issue of exposure. It is
likely that children with chronic
rheumatic
diseases are going to be on these agents
longer
even than adults, and the cumulative risk
is of
great concern.
I think it would be very
important to try
to get some insight into the pathogenesis
of this,
not just the frequency, so that early
markers could
be explored in children who are exposed
before they
exhibit the clinical endpoint.
MR. LEVIN: I haven't spoken for two days,
so now I may go on. A couple of thoughts. One is
I am all about informed choice, but the
question is
how informed is the choice, I think, as
others have
raised, and I want to point out that I
think we
have this sort of mythology of a changing
environment which is patient-centered in
which
there is this sort of partnership.
With all due respect to the
clinicians
around the table in the room, I don't
think that
characterizes most people's experience in
the
552
health care system today. I think it is totally
unrealistic. We have 45- to 50,000 people who are
uninsured, who have very haphazard access
to care,
certainly don't have an ongoing
relationship
probably with a practitioner who is going
to sit
down and run through the benefits and
risks in the
alternative therapies and help them make
an
informed decision.
We know from studies of how
much time
physicians have with patients and what
they convey
when they prescribe a drug, that is far
from the
role of the learned intermediary that is
sort of I
think mythic, and we need to get over.
I agree with Lou that we need
to ask
patients what they want and what their
experience
is, but on the other hand, we have a
regulatory
context here. We have 1906, we have 1938, we have
1962.
For better or worse, the Congress has
decided that there is a role for
government to play
in protecting the public from harm.
So, I don't think we can just
sort of
slide this all off on patients and physicians
553
supposed in this Nirvana good, up-to-date
information, making intelligent choices
through
this very difficult, complex issue.
The Government does have a
responsibility,
and that is why we are here. We are being asked
for I think advice on how government can
best meets
its responsibilities under statute to
protect the
public health and to do what it has to
do.
We all recognize that there are
lots of
things that need to be improved, I
believe, in the
way new drugs come to market, because I
have sat
through this before when we are chasing
the train.
The train is out of the station, folks,
it is going
down the track very fast, and we are
trying to
catch up to it and figure out what do we
do.
You know, it is heading for the
crossing,
there is a car on the track, how do we
stop the
train.
It is too late. We are always
going to
hear from patients no matter what the
drug, "This
drug worked for me, it's wonderful, it
changed my
life."
I believe them, I certainly
empathize with
554
them.
There will always be that appeal.
So, I
guess we have a complex task, the train
has left
the station, but we can't abrogate our
responsibility, and we can't pretend the
Government, through the FDA, doesn't have
a
statutory responsibility here to protect
the public
health.
We can't just say put
information out
there, make it transparent, let this
mythical
doctor-patient relationship sort of
bubble up and
make things all right, because it's not
going to
happen that way.
DR. WOOD: Helpful comments from our
consumer representative.
Dr. Manzi.
DR. MANZI: First, I would like to
congratulate the members of the panel who
I thought
have brought some very relevant points to
the
table, and I agree with most of them, but
it is
interesting to me how many times I have
heard the
term "safe alternatives" used.
I look at our first question
about
555
weighing the benefits of the COX-2s
versus the
non-selectives, and I think the
assumption, as we
are trying to deal with the coxibs, is
that there
is, quote "safe alternatives"
in the non-selective
agents that we would feel comfortable
having our
patients turn to in the event that these
other
COX-2s were not available.
My question would be, or I
guess my
challenge to my other panel members would
be to
provide data that has been obtained with
the same
rigor and had to undergo the same scrutiny
as the
drugs that we have just looked at to
prove that the
other non-selectives are safe
alternatives.
I don't think we have it. I think we have
signals actually to the opposite
potentially. So,
I just think we have to keep that in mind
as we are
making decisions that patients are going
to have to
turn to something, and do you feel
comfortable
saying that the alternatives are safe.
DR. WOOD: Another way to think of the
same thing, though, is that if we were
sitting here
thinking about approving these drugs
right now,
556
would we approve drugs with a clear
cardiac risk in
randomized clinical trials.
I think that is an important
question for
the committee to address because if we
don't
address that, we will either not be able
to address
it for drugs coming up in the future
and/or we are
going to apply a different standard to drugs
that
are on the market, and I understand all
these
points, but I think it's--maybe I am
wrong--I think
it's highly improbable that the committee
would
have approved any of these drugs given
the safety
signal we have got right now.
I think it is highly improbable
that the
FDA--I am talking about from randomized
clinical
trials--I think it is highly improbable
the FDA
would have approved drugs if they had had
all the
randomized studies they have right now.
That doesn't mean they wouldn't
have
approved them eventually perhaps, but
they
certainly wouldn't have approved them on
that
basis.
Is that fair, Bob?
DR. TEMPLE: I think it varies depending
557
on how you view various collections of
data, but
some of them I think probably would not
have made
it.
DR. WOOD: All right, some of them we
would not, but that is a fair comment.
DR. MANZI: Could I just comment?
DR. WOOD: Sure.
DR. MANZI: I would argue that that would
depend on the need for the drug, and it would
also
depend on the alternatives available, and
so I
think it is hard to look at it in
isolation.
DR. WOOD: Fair point.
Dr. D'Agostino.
DR. D'AGOSTINO: The Framingham study has
generated many risk assessments. They are in the
cholesterol guidelines. Cardiac risk assessment
tools do exist. Would the physicians use them? I
am not sure that cardiologists use them,
nor other
classes of physicians to automatically
use them and
sit with the patient and go through that,
but they
do exist, and if you could build a
scenario for
that, it would be possibly very useful.
But one of the things I wanted
to really
mention isn't just the existence of these
tools,
but there seems to be something
synergistic about
558
taking the drug and your cardiac risk, so
it is not
just a matter of telling you you are
diabetic and
how likely you are to have a heart
attack. This
drug seems to double that or triple that,
and so
forth, so you will be presenting very
high risk to
the subjects, and I am not so sure how
easy that is
to do, but it should be kept in mind that
there is
an elevated risk beyond the normal
cardiac risk.
DR. WOOD: Unless someone else has a
burning question, I am going to give Ms.
Malone the
last word.
MS. MALONE: I feel the need to speak up
for rheumatologists. I had been on this panel I
believe starting in 1995, and as a
consumer rep. I
filled someone's term, and then I had my
own term.
So, I was on it for five years, and then
I came on
as a patient rep intermittently.
From my 35 years dealing with
rheumatologists and being on the panel, I
have to
559
say that rheumatologists, on a whole, are
a unique
set of doctors. They are in there for the long
haul and I have always felt that when I
was on this
committee, if I were not here, that the
voice of
the
patient would still be heard.
I find that I don't think there
is one
rheumatologist on here who would not
spend time
with their patient, who would not spend
time
educating them and listening to them
albeit it it's
not a half-hour, but I think they do have
the
ability to form a relationship with them,
and I
applaud them for that, and I disagree
with Arthur
on that point.
DR. WOOD: Stephanie, I will give you the
last word and then we are stopping.
DR. CRAWFORD: Thank you so much, Mr.
Chairman.
I simply can't quite leave without at
least attempting to address this stunning
near
cliffhanger that we were given about 40
minutes
ago.
I am going to ask, if I
may--and please
forgive me if I get your name wrong, it's
not
560
listed on my papers--I think it was Dr.
Kim from
Merck.
Thank you.
Yesterday, I asked the question
to Dr.
Braunstein about what was or were the
deciding
factors in the extraordinary step that
Merck made
in deciding to voluntarily withdraw
rofecoxib. I
am not sure I heard a clear-cut answer,
so I am
going to ask you something very related
to this
last question we have been addressing
from the
opposite side.
Tonight, what considerations
would you
weigh or would you ask this committee to
consider
when we deliberate tonight or tomorrow in
determining the benefit of potential
re-introduction of rofecoxib, or if you
wish to say
this class, where the benefits would far
outweigh
any issues of safety concerns?
DR. KIM: Thank you for that question, and
I will say that it has certainly been a
very
educational and informative day, two days
actually,
listening to these discussions. I think the issues
are
complex, and I think that all of the complex
561
issues are being brought up.
As I said, Merck believes,
based on the
new data that has just become available,
that what
we are dealing with here in terms of
cardiovascular
risk is a class effect.
The thing that we are
struggling with,
which you are all struggling with, is
what does
that mean in terms of the size of the
class, and,
in particular, is it limited to just
inhibitors of
COX-2 or does it include inhibitors of
COX-2 that
also now have an effect on COX-1.
The only point that I was
trying to make
was that at the time that we decided to
withdraw
Vioxx from the market, we did so based on
the
information that was available to us at
that time,
knowing that there were alternative
therapies and
that there were questions that were
raised by the
APPROVe trial.
Now, where the science has
progressed to,
where we see, we think, and we look
forward to your
decisions, but we think we are dealing
with a class
effect, then, I think we are no longer
dealing with
562
a situation where Vioxx is unique in its
cardiovascular risk, but instead is a
member of a
class.
Then, I think it is important
for
us--again, we are looking to you, this
committee
and the FDA, for your evaluation of
whether or not
you agree with our interpretation that
this is a
class-specific effect, but if that is the
case,
then, I think we need to take a look at
the unique
benefits that Vioxx provides, which I
mentioned,
and actually a fourth benefit which was
already
mentioned, that is, that Vioxx is the
only COX-2
inhibitor which has been proven to reduce
the
events, serious GI events, as compared to
naproxen.
Vioxx is the only COX-2
inhibitor that was
approved that is not contraindicated in
patients
with allergies to sulfonamides, and Vioxx
was the
only COX-2 inhibitor with approval for juvenile
rheumatoid arthritis in addition to the
fact that
we have heard numerous reports from
patients, some
with very chronic debilitating pain, that
Vioxx was
the only drug that worked for them.
With that, I will leave it to
the
committee. We really await your decision on this
issue.
563
DR. WOOD: Okay.
It's never the last
word, is it.
DR. STRAND: May I finish the answer to a
question that I was asked yesterday?
DR. WOOD: Who are you?
DR. STRAND: I am Dr. Strand and I
responded to you yesterday about the use
of COX-2s
in patients, the benefit-risk
profile. I simply to
say that with Dr. Hochberg we authored an
editorial
in 2002 after the introduction of the
data from
CLASS and VIGOR to point out that there
is benefit
with these COX-2s, which is at least numerically
preserved from a GI point of view, both
from TARGET
and CLASS data, with a baby aspirin, and,
in fact,
most of the cardiovascular risk may be
abrogated by
co-administration, and we certainly don't
have to
then worry about the potential
interaction as has
been demonstrated with ibuprofen.
So, I think it is important in
your
564
deliberations to consider that point.
Thank you very much.
DR. WOOD: Kimberly tells me the committee
has to meet in the lobby in 15
minutes. I think
that is pretty optimistic, but good luck.
(Whereupon, at 7:00 p.m., the
proceedings
were recessed, to reconvene on Friday,
February 18,
2005, at 8:00 a.m.)
- - -