Page
201
1 DR. KATZ: I
want to respond. The
2 consent issue
is something that we have
3 discussed and
elected not to put in what I
4 presented
today because, No. 1, we knew it
5 would come up
here and, No. 2, the issue that
6 we think is
paramount is what are we trying to
7 learn and what
are we trying to get with a
8 seven-day
platelet and we're talking about
9 adequate
clinical safety.
10 The consent issues
need to be
11 hammered out and I
would like to hear about
12 how you would
approach community consent for
13 this if it's
appropriate at some other point.
14 As far as sons and
daughters and
15 grandsons of
PASSPORT, I think this is our
16 last shot. I don't
think we are going --
17 absent pathogen
reduction methods I don't
18 think we are going
to probably revisit this a
19 whole lot in public
after a decision is made
20 here.
21 I can't answer
Harvey's implied
22 question about how
many people are not getting
Page 202
1 platelets they
need because I don't know how
2 many people
who get platelets need the
3 platelets they
are getting now. It is an
4 important
point but it's not something I
5 certainly can
deal with from the blood center.
6 We are very
concerned about our
7 ability going
forward to keep platelets on the
8 shelves at
hospitals and that is really the
9 genesis of
where we are coming from with what
10 we think is doable
and we'll hammer out the
11 details where the
devil lives going forward.
12 DR. SIEGAL: Dr.
Fleming.
13 DR. FLEMING: Dr.
Katz' comments,
14 I think, lead into
the two fundamental issues
15 that I want to
discuss. I think the FDA
16 recognizes this
because these are the two
17 fundamental issues
they are asking us to vote
18 on.
19 One of them has to
do with how we
20 define the endpoint
and is the endpoint going
21 to involve the
bacterial contamination
22 component. The
other relates to whether or
Page 203
1 not there is
going to be resampling for
2 contamination
at day five. Let me turn to
3 that issue
first.
4 It seems to
me the fundamental
5 structure for
this is, the standard or control
6 here is
platelets at day one to five. The
7 experimental
is platelets at day six/seven,
8 allowing
platelets to day six/seven, which
9 could occur in
one of two ways.
10 As Dr. Katz has
laid out, it could
11 occur with the
bacterial contamination
12 assessment at day
one at hours 24 to 36
13 without a
subsequent assessment versus the FDA
14 proposal that if
it's on the shelf at day five
15 you would do the
reassessment before you
16 release
it.
17 Those are two
different -- I'm
18 going to call it
two different experimental
19 strategies. If we
had unlimited resources we
20 could do all three.
We could have the day one
21 to five compared to
what Dr. Katz is proposing
22 versus compared to
what the FDA is proposing.
Page 204
1 If we have
one shot, the issue is
2 what Dr. Katz
is proposing has as its benefit
3 convenience
feasibility costs. We've heard
4 about how it
would be difficult for many of
5 these sites to
be bringing these back and
6 having the
reassessment done at day five.
7 The advantage of what you're
8 proposing is
convenience and practicality and
9 feasibility
and lower cost. The disadvantage
10 to that is it seems
logical to presume that
11 that reassessment
at day five has an enhanced
12 sensitivity to
whether or not there would be
13 contamination.
14 And so it seems so
much more
15 plausible that if
you do reassess at day five
16 that the success
rate at day six/seven that
17 you're going to
meet the noninferiority
18 assessment will be
much greater. And so it's
19 a judgment here but
if you take the approach
20 of saying let's see
if it's safe without a
21 reassessment and
you win, that's a great win
22 because
accompanying that win is convenience
Page 205
1 and
feasibility.
2 But if you
lose and, in fact,
3 there are some
suggestions that you may lose
4 based on what
we know from PASSPORT and what
5 we know from
the Irish data, then you'll wish
6 that you had
done what the FDA said and said,
7 "Look, maybe
it would have been okay at day
8 six and seven
if we had just done this scene
9 at day five
with enhanced sensitivity. It
10 seems like that is
the choice we have to make
11 relative to that
second issue.
12 The first issue is
an endpoint
13 issue. The first
issue is essentially the FDA
14 is saying, "Should
we also do a day seven
15 assessment because
they would like to have a
16 bacterial
contamination day seven assessment
17 as an additional or
as a primary endpoint to
18 go along with the
sepsis transfusion reaction
19 endpoint which, by
the way, I love that
20 endpoint because
it's a clinical endpoint.
21 It is exactly what
we are trying
22 to address. But
there are legitimate issues
Page 206
1 that have been
raised about that endpoint
2 including
passive versus active surveillance
3 issues and
including the sample size.
4 Somebody
alluded to the sample size. Let me
5 actually get
into a little bit about the
6 sample
size.
7 These
calculations are very
8 preliminary.
The sense that I have from what
9 has been
presented to us is if what we are
10 comparing is a day
one to five septic rate
11 against a day six
or seven septic rate, if we
12 are just using what
the American Red Cross had
13 indicated which is
three per million and we
14 wanted to rule out
a doubling, and I'm going
15 to come back to
that because essentially what
16 is an unacceptable
increase in risk? Just for
17 the sake of
discussion I'll say rule out a
18 doubling. Then it
would essentially take 20
19 million
people.
20 However, if we
acknowledge that
21 there is under
reporting, and I think there is
22 a lot of
concurrence on the need for some
Page 207
1 level of
active surveillance, some data that
2 I said could
increase this rate by 10 fold in
3 which case the
actual rate in the control arm
4 isn't three
per million, it's three per
5 100,000 and
then ruling out a doubling would
6 be a two
million person trial rather than a 20
7 million person
trial.
8 The
contamination endpoint has a
9 background
rate of three per 10,000 and hence
10 ruling out a
doubling would take 200,000. The
11 contamination
endpoint would take one-tenth
12 the active
surveillance sepsis endpoint,
13 although the active
surveillance sepsis
14 endpoint would be
assessed in all people, the
15 contamination
endpoint would be assessed in
16 maybe only 10
percent so you may be awash.
17
It may be a couple million people
18 for the trial size.
That, by the way, is a
19 very crude back of
the envelope calculation
20 but it is a sense
of where we are. Now, the
21 question that
hasn't been addressed at all
22 today is I
completely agree with the way Dr.
Page 208
1 Katz and the
FDA have indicated this would be
2 a
non-inferiority assessment.
3 The concept
here is we have day
4 one to five
platelets. Can we use day
5 six/seven
without an unacceptable increase in
6 safety risks
because of the importance of the
7 added
availability? That is the way it's laid
8 out. Well, the
fundamental question is what's
9 unacceptable?
I've been indicating I'm using
10 a
doubling.
11 If the baseline
rate here is an
12 active surveillance
sepsis is three cases per
13 100,000 and you are
ruling out a doubling,
14 then a positive
result would be a one-third
15 increase. You are
ruling out 100 percent
16 increase.
17 A positive result
would be an
18 observed one-third
increase so you basically
19 would be saying
this is positive if the number
20 of active
surveillance sepsis cases per
21 100,000 is
increased from three to four or one
22 case. Does that
meet the laugh test? Does
Page 209
1 that make
sense to everybody?
2 If so, then
these numbers are
3 about the
right size. If what we're saying
4 is, yes,
because of the convenience and
5 importance and
feasibility and efficiency of
6 having day
six/seven we would allow in those
7 day six/seven
units to have four per 100,000
8 sepsis
transfusion cases rather than three but
9 we haven't
discussed this issue at all and it
10 has a lot of impact
on what the size of the
11 studies would
be.
12 Dr. Katz, FDA, has
anyone thought
13 about
this?
14 DR. KATZ: Yes. We
think that the
15 sample size is not
as big as yours because I
16 think we're kind of
looking at Ros Yomtovian's
17 true active
surveillance and find the rates
18 higher than what
Ann Eder and Dr. Benjamin, et
19 al. in the Red
Cross did so we didn't think it
20 was quite as
daunting as you do.
21 DR. FLEMING: Just
as an aside, so
22 I'm using a figure
of three per 100,000.
Page 210
1 DR. KATZ:
Wasn't there 10 or 20,
2 Mark, in true
active surveillance? I was
3 looking
through your slides here.
4 DR. FLEMING:
The Red Cross is .3
5 per 100,000 so
I multiplied by ten.
6 DR. KATZ:
Unfortunately the
7 numbers here
-- I should have memorized it but
8 I haven't.
Yes, it's too small for me to see.
9 Mark, can you
enlighten me? It was your
10 slide, I
think.
11 DR. BENJAMIN: Can
I just correct
12 the American Red
Cross passive surveillance
13 data?
14 DR. KATZ:
Yes.
15 DR. BENJAMIN: Some
of the data
16 shown to you today
was preliminary data shown
17 at the ABC meeting
last year. The current
18 rates for septic
reaction report to the Red
19 Cross are running
somewhere between one and
20 100,000 to one in
120,000 transfused products.
21 We will report
later in the year the results
22 on a million
transfusions, but we are still
Page 211
1 updating that
data.
2 DR. FLEMING:
I was using a figure
3 of three per
100,000 with an active
4 surveillance
so higher than your figure.
5 DR. BENJAMIN:
Right, I think you
6 mentioned
three per million for passive
7 surveillance
and that's not right.
8 DR. FLEMING:
Correct. The
9 figures I gave
were .3 per 100,000. I used
10 three per million
for passive surveillance Red
11 Cross and 10 fold
that, three per 100,000 for
12 active
surveillance. I think that is the
13 numbers I'm still
hearing.
14 DR. KATZ: I
thought Ros' was
15 higher.
16 DR. FLEMING: To
maybe 10. As
17 high as
10.
18 DR. KUEHNERT: And
I think it
19 needs to be taken
into consideration that part
20 of Dr. Yomtovian's
active surveillance was to
21 do microbiologic
culture of the units and then
22 call back the ward
and say, "Did you see
Page 212
1 anything
untoward in the patients." "As a
2 matter of
fact, they spiked a fever." And
3 then they
worked it up. That was a
4 combination of
both clinical surveillance and
5 culturing.
6 MR. JEHN:
Well, ideally in our
7 circumstance
we would have the vital signs in
8 front of us
and not have to call back. That's
9 the point.
10 DR. FLEMING: So
just with the
11 numbers that we are
hearing, I have not heard
12 anything to take us
above 10 per 100,000.
13 Even if it got up
to there, my two million
14 number would drop
to just under one million.
15 It doesn't have a
huge affect.
16 DR. SIEGAL: Just
one quick comment
17 is that another
alternative, and I would be
18 interested in Dr.
Katz' assessment of
19 feasibility, would
be to get as good clinical
20 data as you can and
it sounds like that needs
21 more discussion in
an active surveillance
22 approach, but are
there a subset of centers or
Page 213
1 the ability to
then collect microbiologic data
2 at five days
as well. I'm coming at--and are
3 there ways we
can look at those two different
4 kinds of data
and what some of the endpoints
5 would be. I
certainly agree with Dr. Fleming
6 that we should
think through in advance of the
7 study what is
acceptable.
8 The other
comment I was going to
9 make as
certainly an ID clinician and has
10 somebody who has
been involved in clinical
11 trials, but not so
much in blood transfusion
12 so I would be
interested in Matt and others
13 experience, but is
the lack of precision of
14 these definitions
and how important it's going
15 to be that the
instrument be applied in a
16 standardized way
across centers. I think you
17 do have a control
group unless I'm not
18 understanding the
study proposals which would
19 be, let's say,
people transfused on day X
20 versus people
transfused at seven days. If
21 your definition of
"sepsis" is imprecise
22 because it's not a
microbiologic diagnosis on
Page 214
1 the patient,
the power of the study then with
2 an imprecise
measure becomes even more
3 challenging. I
think we can't design the
4 actual numbers
and the measures right here but
5 I think we
have to bear in mind the
6 imprecision of
some of the clinical endpoints
7 in
understanding the power of the study. But
8 I would like to ask could there be -- actually
9 you're not going to have as many people at
10 seven days as five days and before. If you
11 got data on half or less of the people
12 microbiologically, could you also have a
13 secondary endpoint or some kind of endpoint
14 that is microbiologic in comparison?
15 Otherwise, I think you could end up with a
16 hell of a lot of clinical data and without an
17 answer.
18 DR. KATZ: Well, we
thought about
19 it and we discussed
it. At the end of the day
20 to get to tier two,
the surveillance cultures
21 in the original
PASSPORT, we got under 5,000
22 in about two years.
It was like pulling
Page 215
1 teeth. It's
very, very difficult. If we have
2 to do that,
it's another hurdle and it
3 requires a lot
of discussion. That's why we
4 didn't put
surveillance cultures in what I
5 showed you
today because it was so hard to get
6 them in the
original. It's because it is out
7 of control of
those of us with the BacT/ALERT
8 machines.
9 DR. SIEGAL:
Dr. Bianco.
10 DR. BIANCO: I want
to change a
11 little bit the tone
of the discussion. I
12 think that
everything that is being said and
13 the precision that
we are looking for in these
14 studies is very
appropriate and would be very
15 appropriate if the
standard apheresis platelet
16 was the only
product available.
17 There are a lot of
products on the
18 market that is the
blood derived platelet that
19 is not tested, that
is much cheaper than the
20 apheresis platelet
that is not cultured and is
21 available and
created by every blood center.
22 Whatever
difficulties are created
Page 216
1 in the use of
the apheresis platelet, that is
2 much more
expensive and the recruitment and
3 everything is
more difficult, you are going to
4 simply
facilitate the use of the other type of
5 platelet and
we are not going to gain any
6 safety. We are
going to lose safety so I
7 think that has
to be part of the equation.
8 If you could
get rid of the whole
9 blood derived
platelet, the uncultured one,
10 then you could have
the discussion, you could
11 have the study much
more precise than what we
12 are talking about
now.
13 DR. SIEGAL: Dr.
Fleming, you were
14 trying to say
something.
15 DR. FLEMING: I'm
sorry. There
16 were a number of
issues. I wanted to just
17 endorse what Dr.
Katz was talking about that
18 the bacterial
contamination assessment while
19 I'm saying it would
take one tenth the sample
20 size, you wouldn't
have that assessment in all
21 units and so it
probably would take a similar
22 sample size as the
sepsis endpoint.
Page 217
1 The other
comment that I wanted to
2 follow up, the
FDA was talking about the
3 importance of
having a precise definition of
4 the endpoint
of septic transfusion reaction.
5 I want to
endorse that as well and we've
6 talked already
about why having active
7 surveillance
is important, capturing all the
8 events,
increasing by as much as ten fold and,
9 hence,
reducing the sample size by ten fold.
10 But the other
reason it's
11 important is
essentially as at least a
12 suggestion we have
talked about ruling out an
13 increase from three
per 100,000 to six per
14 100,000 which is
achieved if you have an
15 increase of no more
than one per 100,000.
16 But if there is
fuzziness in the
17 definition of
sepsis, and many cases of sepsis
18 are occurring for
independent reasons, then
19 you would have
similarity in -- you would
20 dilute out what
would be an unacceptable
21 increase if, in
fact, day six/seven is
22 unacceptable if it
truly doubles.
Page 218
1 That could be
diluted. If it's
2 truly three
versus six that you have three or
3 four other
events in both arms from unrelated
4 causes, then
you have diluted out the excess.
5 And so to
ensure that you don't miss an
6 unacceptable
excess, it is very important to
7 have very
careful documentation of what it is,
8 of the
mechanism you are trying to address
9 here which is
specifically septic transfusion
10 reactions.
11 DR. KUEHNERT: Just
a general
12 comment and then a
specific one. It just
13 seems like we are
trying to determine what may
14 be a very, very
small difference between day
15 seven and day five
and then on top of it we
16 are trying to use
endpoint measures which may
17 also be very, very
imprecise. It's trying to
18 hit a target
blindfolded.
19 But one way I
think that -- of
20 course, there is a
lot of data on sepsis and
21 what signs and
symptoms are associated with
22 sepsis so one could
look at that, the
Page 219
1 evidence-based, and come up with a definition.
2 One other way to be more precise
3 would be to have a low threshold for doing
4 recipient blood cultures. Of course, that
5 involves another procedure for the patient but
6 is something to be considered because that
7 might not be something that is automatically
8 done.
9 Certainly if
there is fever and
10 other obvious signs
of sepsis it's going to be
11 done but it might
not be, for instance, in the
12 absence of fever.
So if that is worked into
13 the protocol, it
might be beneficial as far as
14 determining the
true nature of some of these
15 cases.
16 DR. WAXMAN: The
only problem with
17 that is that won't
pick up potentially
18 endotoxin that's in
the contaminated platelet
19 pack that is given
to the patient.
20 DR. KUEHNERT:
That's true, it
21 won't detect
endotoxin. But if someone, I
22 mean, if there is a
unit with a lot of
Page 220
1 endotoxin, I
mean, it's going to be a gram
2 negative and
it's likely going to be present
3 in a blood
culture. You're right, that is a
4 limitation.
5 DR. ZIMRIN:
The other problem is
6 a lot of the
patients are going to be on
7 antibiotics so
that would limit the utility of
8 the
cultures.
9 DR. KUEHNERT:
I wonder also if
10 that kind of
information is going to be
11 collected because,
of course, antibiotics is
12 going to blunt any
sign or symptom and blood
13 culture results so
that would be important to
14 note,
too.
15 DR. BRACEY: A key
driver here is
16 the issue of
availability and safety and
17 shifting from one
type of a component, the
18 whole blood
derived, the apheresis platelet
19 toward whole blood
derived platelets.
20 One large
collector that collects
21 about half of the
nation's blood supply, I
22 believe, is not
looking at extending beyond
Page 221
1 five days. I
would like to know from the Red
2 Cross and
maybe Dr. Benjamin to let us know.
3 Have you seen
outdates that are
4 problematic in
terms of -- sorry, losses of
5 apheresis
platelets that are problematic and
6 have you seen
a shift in your system away from
7 what is the
cultured platelet toward whole
8 blood derived
platelets?
9 DR. BENJAMIN:
It's an interesting
10 question. I'm not
sure I can actually answer
11 it. The Red Cross
works hard to get the
12 platelets out of
it's facilities and into
13 hospitals. Unlike
some other blood centers
14 that will take
platelets back and take the
15 responsibility for
outdates, we generally
16 don't do that so it
is the hospitals that are
17 taking the hit for
outdates. We don't have a
18 good measure of
what is happening in the
19 hospitals on
outdates.
20 Having said that,
we don't outdate
21 very much at the
Red Cross because every
22 single platelet we
make there is demand for.
Page 222
1 Over the last
two years we increased our
2 production by
some 10 percent and our updates
3 didn't change.
Our updates run one to two
4 percent, very,
very low. So our impression is
5 that there is
unmet demand out there. We
6 don't know how
much of that is from outdates.
7 DR. KATZ:
That's very important
8 what Richard
just said. For example, he sees
9 a one to two
percent outdate but he doesn't
10 know what's
happening in the hospital.
11 Whereas the data
that I shared with you from
12 BSI and my center
represents actually what
13 outdates.
14 For my center we
do our platelets
15 on consignment.
They go into the hospital and
16 if they are not
used in some agreed-upon
17 interval, they go
to another hospital that has
18 need. So we
actually know what our outdate
19 rates are. I think
the numbers around the
20 country, those who
do it the way we do it, are
21 kind of converging
on post-PASSPORT in that 10
22 percent range.
Every outdate isn't an outdate
Page 223
1 and you need
to understand that.
2 DR. FINNEGAN:
Question for Dr.
3 Katz. I really
like the concept of the two
4 arms. In other
words, having most of the
5 people do the
passive but having a small group
6 of people
because it seems to me like this is
7 important to
your group and if it is, then
8 there should
some buy-in from some of the
9 larger
centers.
10 My actual
hypothesis would be that
11 the passive is
probably just as good. But I
12 think if you do two
arms two things, you can
13 make a lot of
people happy and you can
14 actually see
because it would be less
15 expensive to do it.
Is that even remotely
16 feasible?
17 DR. KATZ: At this
point -- it's
18 an interesting
question. Do I have to have
19 100 percent buy-in
to distribute a seven-day
20 platelet in which
case I have to have two
21 inventories and now
it becomes a very
22 complicated CGMP
issue.
Page 224
1 If I have to
have a five-day and a
2 seven-day
inventory it's a very complicated
3 CGMP issue.
Our original idea or the most
4 feasible thing
was to get a good sample of the
5 big users to
agree to send us back all the
6 data so it is
functionally a two-arm study.
7 DR. SIEGAL:
Jay.
8 DR. EPSTEIN:
I would like to
9 probe the
feasibility issue of the five-day
10 culture a little
bit by posing a question to
11 you, Louis, or to
Celso.
12 The way the issue
has been framed
13 is that there is
agreement from the scientific
14 standpoint that a
day-five culture would be
15 informative,
especially if cultured -- I'm
16 sorry, if coupled
with a cultured outdate
17 essentially day
eight.
18 Dr. Fleming has
made clear an
19 argument which has
always been in FDA's mind
20 which is that it
takes smaller numbers to get
21 a useful answer
which then also means you get
22 it
quicker.
Page 225
1 So then the
question is are there
2 ways to make
this feasible in the hospital.
3 The naive
point of view is every hospital can
4 do a blood
culture on a patient so is there
5 not a way for
the transfusion service to get
6 a culture done
on a sample from a platelet
7 sitting in the
inventory.
8 We need to
ask the operators that
9 question
because the argument is coming to us
10 from the operators,
"No, no, we can't do
11 this," but it's not
entirely clear to us
12 whether there are
ways to do this.
13 DR. GLYNN: Can I
also add
14 something -- I'm
sorry -- to Jay's comment?
15 DR. SIEGAL: Go
ahead.
16 DR. GLYNN: Also
this is only for
17 the duration of a
study. Right? That's what
18 we're talking
about. We're not talking about
19 doing this ad
infinitum. Is that right?
20 DR. EPSTEIN: Well,
a little bit
21 it depends on the
outcome because, after all,
22 if what you
discover that you have a high
Page 226
1 contamination
rate on day five that you can
2 effectively
lower by day seven by doing a
3 five-day
culture, there will be an argument
4 that you
should then maintain that as a
5 condition of a
seven-day platelet.
6 It's been
argued ideally you
7 should grab a
day three culture to assure
8 safety of a
day-five platelet. A day-five
9 culture
assures safety of a day-seven
10 platelet. I think
that it's data driven. If
11 what we find out is
that there is a de minimis
12 difference, then
you would argue against it.
13 If you find out,
as in the Decourt
14 study that the
culture -- I don't mean the
15 Decourt study, the
Irish study, Murphy's
16 study, that the
rate is actually higher on day
17 seven than it was
on day one, then you would
18 have an argument to
do it.
19 Of course, their
finding was that
20 culturing on day
four didn't get you to safety
21 on day seven. This
brings in a whole other
22 issue that we
haven't dissected in detail
Page 227
1 which is what
are the slow-growing organisms.
2 The problem
is that you've got
3 slow-growers
that you seem to miss not only on
4 day one but
also on day four. Because if you
5 look at what
you grew out on day seven, it's
6 a different
spectrum of organisms than you
7 grew out on
day one.
8 There is a
whole other discussion
9 that we
haven't had about what is the
10 pathogenicity of
these slow-growers that you
11 pick up late and
might miss as late as day
12 four. The short
answer to your question is it
13 would depend on the
outcome because that is
14 what we'll tell you
whether there is a benefit
15 of a five-day
culture.
16 DR. FLEMING: Can I
just very
17 briefly to your
response? It's a great
18 question.
Assessments that are being made
19 purely from the
perspective of having the
20 ability of a
scientific study to assess an
21 outcome is specific
to the scientific study.
22 For example, if we
say yes to
Page 228
1 question 2(b),
that's an issue that relates to
2 having added
information for the scientific
3 study. But
2(a), which is the point I tried
4 to make
earlier on, 2(a) isn't just collecting
5 data to assess
for the scientific study. It's
6 changing the
strategy.
7 There are two
choices that we
8 have. There
are two strategies to consider.
9 One strategy
where you are only looking at
10 bacterial
contamination assessments at hour 24
11 and not subsequent
to that which I think is
12 Dr. Katz' proposal.
The other strategy, which
13 is FDA's proposal,
is before you use day
14 six/seven you've
got to reassess at day five.
15 If you go with the
FDA strategy
16 and you get a
positive result, then that day
17 five assessment is
inherent to the strategy.
18 It stays forever
because all you've proven is
19 that it's safe as
long as you validate it safe
20 at day
five.
21 I would come back
to Dr. Katz and
22 ask the question
that strategy has far more
Page 229
1 likelihood of
being successful. Your strategy
2 is more
aggressive. It's saying, "I think you
3 are safe at
day six and seven without even
4 assessing at
day five." It has the benefit,
5 if you're
right, of feasibility, practicality,
6 cost and
convenience.
7 If it turns
out that, no, when you
8 don't reassess
at day five you miss the slow
9 growers and
they are going to cause sepsis in
10 a few cases per
100,000 than condemning this
11 approach, then you
have missed the opportunity
12 to see that there
was a way to use day
13 six/seven and that
was with reassessment.
14 But the
fundamental point for Dr.
15 Glynn's question is
the day five assessment is
16 inherent to the
strategy. It doesn't go away
17 if you put that in,
whereas the day seven
18 assessment is just
an outcome assessment.
19 DR. RENTAS: To
answer your
20 question coming
from someone that is running
21 a hospital blood
bank right now, I think the
22 biggest issue with
the day-five culture will
Page 230
1 be having to
hold a unit for 12 hours after
2 doing the
testing. That will be a big issue
3 for everyone
out there.
4 Our other
issue out there will be
5 what type of
testing do we do at day five.
6 Are we talking
about rapid testing and use
7 that to
validate for QC, or are we talking
8 about
BacT/ALERT. Those are two issues in
9 there that we
haven't talked about.
10 DR. GLYNN: The
other issue, I
11 think, is also if
you do not know -- if you
12 don't have the
data, how can you really make
13 a decision? Unless
you culture at day five
14 how can you
decide?
15 DR. KATZ: Well,
just to
16 reiterate, we think
that it is clinical
17 outcomes assessed
by a method agreed upon by
18 sponsors,
investigators, and FDA that is what
19 we're after, not
culture results.
20 If we look at
100,000 -- well, the
21 sample size has to
be calculated based on
22 reasonable
estimates of clinical events from
Page 231
1 that
surveillance that's available. And if
2 the clinical
events are equivalent between the
3 two, that is
what we want to know within the
4 bounds of the
feasibility of getting an
5 answer.
6 DR. FLEMING:
You've just answered
7 an important
issue. You've justified the
8 endpoint but
the motivation for the day-five
9 culture is
much beyond just the endpoint.
10 It's the strategy.
Is it your sense -- the
11 FDA has advocated
pursuing the safety and
12 efficacy of a
strategy of reassessing at day
13 five and then using
the platelet on day
14 six/seven.
15 Are you not
supportive of that
16 because you think
it's so impractical it's a
17 non-starter and,
therefore, you are willing to
18 go out more on the
end of the limb and hope
19 that you can still
be safe at day six/seven
20 without the
assessment? Is that your
21 philosophy?
22 DR. KATZ: Yes. I
think that is
Page 232
1 really what
we're talking about here is
2 operational
feasibility. Once again I point
3 to the
surveillance cultures, the outdate
4 cultures in
the original PASSPORT.
5 It took us
from the fall of '05
6 until the end
of January this year to get less
7 than 5,000.
Operationally it's very difficult.
8 It's not that
hospitals can't do cultures.
9 It's that with
seven-day platelets the number
10 of outdates is
microscopic, No. 1.
11 Then, as we've
heard, the
12 operational -- it's
hard enough for us to
13 control things and
complex algorithms with
14 good blood
establishment and computer systems.
15 At the hospital
level to control a hold and
16 then relabel or do
whatever they have to do to
17 make it a seven-day
platelet is very
18 difficult.
19 DR. FLEMING: But
are you not
20 worried about your
results from the PASSPORT
21 that you had
initially and from the Irish
22 study even though
it's a different setting
Page 233
1 suggesting
that -- and all the data we see
2 with the major
increase in bacterial
3 contamination,
aren't you worried that day six
4 and seven is
very plausibly going to double
5 this rate of
sepsis?
6 DR. KATZ: Our
look at the bugs
7 involved
suggest that we are measuring
8 something that
is not clinically significant.
9 Our belief is
that these bugs at the levels
10 that they are
present in those cultures at
11 outdate probably
aren't doing anything to
12 patients.
13 When we look at
admittedly our
14 passive
surveillance and find rates very
15 similar to what
everybody else is seeing, I
16 don't think it is
materially worse. That's
17 really the crux of
the matter, do we believe
18 the coag-negative
staph and the propriano
19 bacterium that
we're finding out late are
20 important. From a
clinical standpoint I don't
21 think so but that's
why we have to answer the
22 question.
Page 234
1 DR. SIEGAL:
Dr. Kulkarni.
2 DR. KULKARNI:
You know, I really
3 have problems
with this labeling because it
4 puts the onus
on the patient and the
5 physician.
Like if I have a kid who is
6 bleeding away
and here I'm supposed to explain
7 to them that
there may be a risk of bacterial
8 contamination,
etc., etc., etc., so I just
9 feel very
troubled about it as a clinician
10 that in the midst
of something that I have to
11 explain to
people.
12 It reminds me of
the HIV days
13 where I have to
treat patients with hemophilia
14 saying, "I don't
know what I'm giving you but
15 I've got to give it
to you." I think the
16 testing on day five
and day seven, at least
17 once the results
are available, at least gives
18 me that confidence
that I can tell to the
19 patients.
20 I'm looking at it
strictly from a
21 patient's point of
view. I agree that Matt
22 that the gold
standard would be to do a pre-
Page 235
1 and
post-transfusion blood culture on the
2 patients who
are receiving it. Yesterday I
3 saw a kid full
of mouth sores and all and very
4 low platelet
count that I had to give
5 platelets.
6 If he
developed eclampsia I don't
7 know whether
it came from him or whether it
8 came from
eating or whether it was both,
9 whether it
would enhance. So I think when you
10 do an active
surveillance, I think having some
11 information from
the patients would be
12 helpful.
13 DR. WAXMAN: Fred,
I have a
14 comment.
15 DR. SIEGAL:
Jay.
16 DR. EPSTEIN: Thank
you. Lou, I'm
17 struck by your
point about the 12-hour hold in
18 the FDA proposal.
If there were no hold but
19 cultures were done,
for example, in the
20 morning of day
five, then you would not have
21 changed the safety
of a day-five platelet by
22 not waiting for the
culture and by day six you
Page 236
1 would already
have a 24-hour culture. Would
2 the landscape
be different if FDA did not ask
3 for a 12-hour
hold? A 12-hour hold is an
4 overnight
hold, in fact. It allows you to
5 culture at the
end of the day. But I'm just
6 pointing out
you would have the same effect by
7 day six
whether you called it a hold or you
8 didn't call it
a hold on day five. The issue
9 is whether you
issued a day five platelet
10 prior to the
result. I don't think it's FDA's
11 intent that you
would not issue a day-five
12 platelet. It's the
day-six platelet that was
13 in question. Would
that change the landscape?
14 DR. KATZ: Well, it
certainly is
15 easier. There is no
doubt that it's easier.
16 It is something
that when we've gotten done
17 with this wrestling
match we have to go ask
18 our hospitals, "Are
you willing to do this in
19 order to have a
seven-day platelet?" The
20 advantage is that
critical mass is going to be
21 hard to get but it
doesn't mean we can't ask
22 if that's what we
want.
Page 237
1 DR. EPSTEIN:
But at that point
2 there's no
quarantine, there's no stratified
3 inventory.
It's just like if you take culture
4 at day one and
it's day four and you're about
5 to issue a
platelet you check the culture
6 first. Same
scenario.
7 DR. WAXMAN:
Fred.
8 DR. SIEGAL:
Celso.
9 DR. BIANCO:
The major obstacle
10 there in the
logistics is something that we
11 are not considering
here, it's sampling the
12 unit.
13 DR. EPSTEIN:
Sorry?
14 DR. BIANCO:
Sampling the platelet
15 unit. We don't want
them to stick a needle
16 inside the unit so
you will have to do a
17 sterile connecting.
You will have to have a
18 segment that allows
you to do the sterile
19 connecting. You
will have to mix the bag very
20 well if you want a
representative culture and
21 all the things that
we do in the blood center
22 and to avoid
contamination. There are very
Page 238
1 few hospital
blood banks that can do that in
2 an effective
way and that is my concern. That
3 is why again
we had out of 400,000 platelets
4 that were
studied in PASSPORT I we had only
5 5,000 cultures
at the end of the process.
6 DR. KATZ: I
don't want to be
7 categorical
and say that X, Y, or Z is a
8 nonstarter. I
think the sense of the PASSPORT
9 group is that
if we are asking the hospitals
10 to do more than
they do now beyond some
11 commitments for an
active surveillance
12 program, they are
going to tell us no. Day
13 five with a hold is
certainly a nonstarter.
14 Day five culture
without a hold we can ask but
15 I'm not optimistic
I'll get a response that
16 allows me to go
forward.
17 DR. WAXMAN: Let me
make a
18 comment. I've been
trying to get in.
19 DR. SIEGAL:
Please.
20 DR. WAXMAN: After
hearing all
21 this discussion I
am less enthusiastic about
22 doing this study. I
mean, just look at the
Page 239
1 discussion
that we've been having. At the end
2 of the day
because we're not doing cultures at
3 day five and
if it turns out that we do have
4 an increase or
no change from the previous
5 study of
infection rate, we're not going to
6 know where
we're at. We're going to be in the
7 same situation
as PASSPORT I. So to put all
8 this effort
into the next two years, or two
9 and a half
years or three years, and not have
10 a true scientific
study that we can actually
11 answer some very
critical questions to me
12 seems like a total
waste of time until perhaps
13 technology catches
up with reality of today's
14 world. As we heard
from the German speaker,
15 maybe flow
cytometry where we can get an
16 answer in an hour.
Most hospitals -- I don't
17 know about some of
the rural hospitals in
18 Mississippi but
many of the hospitals do have
19 some capability of
doing flow cytometry but
20 that's probably off
in the future as that
21 technique becomes
validated. I am really
22 totally less
enthusiastic about PASSPORT II
Page 240
1 because I
think we'll be back here in three
2 years talking
about the same issues because we
3 haven't really
rigorously set up this new
4 PASSPORT II to
answer some of the critical
5 questions.
6 DR. McCOMAS:
If I may make a
7 comment as
well. I've heard a lot about the
8 demand for
seven-day platelets but I've also
9 heard people
saying nobody really wants an old
10 platelet. I've also
heard the argument that
11 hospitals won't opt
into the study if it's
12 very difficult for
them. I assume that means
13 they will just go
back to taking up to five-
14 day platelets. So I
guess I'm trying to
15 understand who is
benefitting from this. It
16 kind of comes back
to this issue of the
17 informed consent
which I think was also raised
18 or underscored in
my mind over the confusion
19 of the labeling and
the labeling that says
20 that testing does
not guarantee sterility.
21 The risk increases
with age of platelets. Who
22 necessarily reads
that labeling? It seems to
Page 241
1 me the person
who is doing the transfusion as
2 opposed to the
patient themselves. So we are
3 asking
somebody to take on a risk but not
4 really asking
the end user which I think you
5 pointed out.
But yet, we are not really
6 explaining the
benefits of this. Again, I'm
7 not hearing
the benefits other than an
8 increased
availability but, I guess, perhaps
9 I haven't
heard the point made about the
10 shortage of
platelets in the current system
11 that would perhaps
make it more attractive to
12 a hospital to go
out of their way to do five-
13 day cultures so
that is the argument I haven't
14 seen made
yet.
15 DR. SIEGAL: Okay.
Dr. Cryer.
16 DR. CRYER: Louis,
I buy the
17 problem with the
hospital not wanting to do
18 the extra work. I
guess the question I have
19 is I'm a hospital
and I have five-day
20 platelets now. If I
want seven-day platelets
21 I've got to do some
extra work to fill out
22 this form. I've got
to make a clinical
Page 242
1 assessment on
all those patients. Am I going
2 to want to do
that even?
3 DR. KATZ:
Well, we're not sure.
4 Sample size
becomes critical on that point.
5 It depends on
how you run your system. My
6 hospitals
don't bear the cost of an outdate so
7 they say to
me, "Just keep my platelet
8 incubator at
our agreed-upon levels and seven
9 versus five
days is not my problem."
10 Perhaps we should
change our
11 strategy so that if
it goes into their
12 hospital and it
outdates, they pay for it.
13 That's not been the
business model certainly
14 that we've used. So
on our side we have taken
15 away incentives
from the hospitals to
16 participate in
this. I think it's a fair
17 question that I
don't know the answer to.
18 DR. SIEGAL:
Speaker in the back.
19 DR. FITZPATRICK:
Mike
20 Fitzpatrick, former
member of the Committee
21 and hopefully
you'll take my comments as not
22 conflicted even
though I'm the president of a
Page 243
1 company trying
to make freeze-dried platelets.
2 I applaud the
industry for trying
3 to put
together a study that looks at clinical
4 outcomes as an
endpoint. I applaud FDA for
5 putting
together studies that looks at
6 scientific
data as an endpoint. There is a
7 basic
conflict. The same people that have
8 proposed
clinical outcomes as an endpoint have
9 used the
problems that clinical outcomes are
10 fraught with to
shoot down other studies. You
11 can't get
compliance. It's hard to do
12 assessments. It's
difficult to know whether
13 everyone is doing
assessments the same way or
14 not. CDC showed you
a study that says you
15 need to go a year
out to look at whether the
16 patient really was
transfused something that
17 was harmful. You
obviously can't do that.
18 FDA wants to know
if you have a known pathogen
19 that you are
transfusing to a patient after
20 day five. Seems
reasonable. Clinical
21 assessments are a
roll of the dice. If you
22 transfuse a known
pathogen to an
Page 244
1 immunocompromised patient you will have a
2 totally different clinical outcome than if you
3 transfused a known pathogen to an ER patient
4 who dies in the ER. You don't have anyway of
5 knowing who is going to get that platelet that
6 was contaminated with a known pathogen. You
7 can't predict that next year random chance
8 will be the same and only those non-
9 suspectable patients will get the contaminated
10 units. It seems like the five-day culture is
11 the only one that is going to give you real
12 data to go on to assess the safety of the
13 product. You have lots of data that shows you
14 that the product decreases in safety over time
15 because of bacterial growth. That seems
16 reasonable, too. The design of a study that
17 provides you good scientific data and gives
18 you clinical outcome seems to be the best
19 solution. It's also difficult and will cost
20 money. That is the problem that always faces
21 this Committee. You want to get answers to a
22 question, you want good data, and it cost
Page 245
1 money to do
that. Rather than expending
2 resources
toward a study that is going to
3 raise more
questions than it will provide
4 answers it
seems like you need to make that
5 plea for the
money to do the study in the
6 correct
manner. I think that would be my
7 suggestion to
the Committee.
8 On the flip
side, on the
9 conflicted
side there are products that have
10 been developed, one
of which is in use by the
11 Netherlands which
is a cryo-preserved platelet
12 for bleeding
patients. It's not for oncology
13 patients. It's not
used to be a prophylactic
14 platelet. It's in
use by the Netherlands for
15 their military.
There has been little
16 interest in
adopting that other than the
17 military in the
U.S. because there is no
18 patents around it.
It's not seen as a
19 profitable measure.
But it would allow you to
20 freeze platelets,
do donor retesting, and
21 provide a safer
product and you can hold that
22 culture for as long
as you want. A liofilized
Page 246
1 platelet could
do the same thing. Our goal is
2 to help
provide a safer product, not just make
3 money. For the
Committee I think you have a
4 very difficult
task ahead which is to try to
5 design a study
that provides both good
6 scientific
data based on clinical outcomes and
7 shows that you
can either be as safe as or
8 increase the
safety of the product.
9 DR. SIEGAL:
Question.
10 DR. KLEINMAN: Yes,
Steve
11 Kleinman, AABB. I
notice that the preliminary
12 questions were
different than the final
13 questions. I know
that the PASSPORT
14 subcommittee had
looked at another option
15 which was instead
of re-culture at five days
16 a point of release
test at five days, the only
17 licensed one being
Verax.
18 I wonder why that
option has, at
19 least by FDA, been
taken off the table
20 apparently?
Wouldn't that still give you
21 some enhancement? I
don't know if that's
22 practical. I mean,
I know the test cost a lot
Page 247
1 of money and
hospitals will still have to run
2 it.
3 Training of
staff would be
4 considerable
and it may not be -- it may have
5 all the
logistical problems as does a
6 BacT/ALERT
culture but I'm curious from a
7 scientific
point of view why that -- we know
8 the
sensitivity of the test is less than a
9 culture.
10 Yet, we would be
applying it at
11 day five when
presumably it would have
12 sufficient
sensitivity to tell you which
13 platelets were
likely to cause clinical
14 transfusion
reaction. I'm wondering why
15 that's not in the
mix, at least, in the
16 discussion.
17 DR. VOSTAL: I
think that is a
18 very good point. We
discussed using or
19 advocating a point
of release test in the
20 design of the
study. A couple of issues that
21 came up when we
were discussing this. One is
22 the big difference
in sensitivity between
Page 248
1 culture and
point of release is almost
2 100,000, a
full difference.
3 If you look
at the Irish study
4 they used a
culture at day four and the
5 culture was
actually missing contaminated
6 product.
Sensitivity even at day four and day
7 five is
probably still an issue. To have the
8 most
conservative approach would be to do a
9 culture at day
five.
10 The other problem
we have is that
11 we have no clinical
data for the use of the
12 Verax device. We
are hoping that device will
13 be able to
participate in the PASSPORT study
14 and in
participation it might need to get
15 validated in terms
of its sensitivity to
16 detect contaminated
products. So we actually
17 have no issues with
the rapid bacterial
18 detection device
being part of the study but
19 we think it would
have to be in addition to a
20 culture.
21 DR. SIEGAL: You
know, we're
22 getting past the
point of needing to address
Page 249
1 the question
so I would like further
2 discussion to
be brief if possible.
3 DR. ZIMRIN:
Very briefly, I just
4 wanted to
address Dr. McComas' point. You're
5 right, the
constituency we don't see here is
6 the people who
are not getting platelets. I
7 certainly have
seen people have major
8 hemorrhages
even with a fatal outcome with
9 thrombocytopenia. Usually because they were
10 refractory but having an increased number of
11 platelets available makes it easier to find
12 platelets for patients who do have
13 refractoriness.
14 In Louis Katz'
slide I think 12 of
15 18 hospitals said
they were having trouble
16 filling orders. I
don't know what that means,
17 whether people are
just postponing surgery,
18 whether they are
transfusing leukemics at
19 5,000 instead of
9,000. I don't know. That
20 data has not been
presented.
21 DR. SIEGAL: Yes,
in the back.
22 DR. DUMONT: Larry
Dumont from
Page 250
1 Dartmouth.
Couple of facts on availability.
2 I got the data
from Dr. Thomas Sulo at BSI.
3 During their
short tenure with seven-day
4 platelets they
had an outdate of just over
5 three percent.
Since they have gone back to
6 five-day
platelets that's gone to nearly 12
7 percent.
They've had to increase their
8 collections,
apheresis collections, by 20
9 percent to
maintain their service level with
10 their customers.
That will give you some
11 flavor for the
impact. We know that.
12 We also know -- I
mean, we've
13 gotten into a lot
of design minutia here but
14 on the big picture
we also know with recent
15 data that people,
in fact, today are changing
16 their bacterial
screening strategy by what
17 some people would
say would be in a less safe
18 mode with changes
to the number of bottles and
19 the volumes and the
timing, etc.
20 We also know for a
fact that
21 people are shifting
their supply. Some people
22 are shifting their
supply from apheresis to
Page 251
1 untested whole
blood. And, in addition to
2 that, we know
that people have to delay and/or
3 modify their
TRALI mitigations that they had
4 planned to do
this year because of this
5 significant
impact on availability.
6 Put those all
together and it
7 looks like, in
fact, today there is already a
8 shift
backwards in safety. I think what the
9 passport group
and what the field is telling
10 people is that we
are going backwards. We've
11 got a real
problem.
12 We need seven-day
platelets to
13 maintain a safety
level and a proper
14 availability of our
platelets. We are looking
15 for truly a Phase
IV study with some
16 intelligent design
on how we can properly
17 assess and evaluate
the safety over time.
18 But in fact, today
I think many
19 people in this room
would attest to the fact
20 that we are
marching backwards in safety from
21 where we were when
we had seven-day platelet
22 availability.
Thanks.
Page 252
1 DR. SIEGAL:
Ross and then we
2 should get to
the questions.
3 DR. KUEHNERT:
I was actually
4 going to ask
if the questions were going to be
5 read and then
if we would have an opportunity
6 to ask for
points of clarification on the
7 questions
because there are some things in the
8 questions that
I don't fully understand the
9 meaning
of.
10 DR. SIEGAL: I
think that is
11 reasonable. Would
the FDA like to start with
12 the
questions?
13 DR. FLEMING: Are
we going to the
14 questions? Is that
where we are? Could I
15 just ask one final
quick question?
16 DR. SIEGAL:
Sure.
17 DR. FLEMING: It
follows up on Dr.
18 McComas', I
thought, key point. Comment and
19 then question for
you. My comment is it seems
20 to me that if I'm a
patient there are several
21 different scenarios
that I could have. One is
22 that I could have a
day one to five platelet
Page 253
1 that has been
assessed at day one and found to
2 be
safe.
3 The second
scenario is I could
4 have a day
six/seven that has had a bacterial
5 contamination
assessment at day five. The
6 third scenario
is a day six/seven that hasn't
7 had a
reassessment. The question is what is
8 in the
patient's best interest?
9 My sense is I
don't know about the
10 day six/seven risk
when I've only had a day
11 one but there is
enough evidence here to
12 suggest that it
could readily be doubling the
13 rate of sepsis
related events in transfusion
14 from three per
100,000 to six per 100,000.
15 Here is my
fundamental thought and question.
16 If that's the
case, and it may
17 readily be true,
aren't we much more sensitive
18 to detect those
increases if we do a day five
19 culture rather than
just the day one? If, in
20 fact, you have such
a platelet unit, isn't it
21 much more likely
that if it is, in fact,
22 unsafe for use at
day six/seven that you
Page 254
1 detect it at
the day five reassessment than at
2 day
one?
3 DR. KATZ: I
don't think anybody
4 is arguing
that a sensitive assessment at day
5 five is a bad
thing in and of itself.
6 DR. FLEMING:
I'm not saying it's
7 a bad thing.
Isn't it very plausible
8 understanding
the growth patterns that we've
9 seen that if
you are going to expose a patient
10 to day six/seven,
that patient is much more
11 likely to be
protected if you had a
12 reassessment at day
five than not. Isn't that
13 true?
14 DR. KATZ: Much
more likely, I'm
15 not going to argue
about the adjective but,
16 yes. I think you
can find units that may be
17 clinically
significant with that day five that
18 we don't find now.
I believe that is probably
19 true. Then I ask
myself if that's the case,
20 why am I still
making platelets derived from
21 whole blood and
distributing them?
22 DR. FLEMING: So
the bottom line
Page 255
1 seems to me
that if you were going to provide
2 just day one
to five, that's the safest
3 approach but
then there is a supply issue.
4 Then a step to
expand the supply issue is to
5 allow day
six/seven with an assessment having
6 been done at
day five. The only rationale for
7 expanding to
day six/seven without doing an
8 assessment at
day five if it can be argued
9 that we are
going to run into situations where
10 if not that a
patient had no option and if we
11 made that case. Are
we to the point that a
12 day six -- I could
understand following kind
13 of the reasoning
that I heard from Dr. McComas
14 that if it were a
choice between nothing and
15 a day six/seven
without a reassessment at day
16 five, I would take
the day six/seven but that
17 is the only
circumstance under which I would
18 take it. Do we not
owe patients with the
19 evidence that exist
here for plausibility of
20 increased safety
risks with day six/seven that
21 there is something
you can do. Do a day five
22 assessment which is
much more likely to be
Page 256
1 sensitive than
day one. Am I missing
2 something with
that logic?
3 DR. KATZ: I
don't think you are.
4 The problem is
how can we get the day five
5 assessment
done? Currently available methods,
6 including the
rapid assay, we are hearing from
7 the people
that would actually do the work
8 that is going
to be pretty tough to do.
9 That doesn't
mean that another
10 company with rapid
assay isn't in the wings
11 with something that
has a higher through-put,
12 a little quicker to
get done, that sort of
13 thing, that could
be revisited when the FDA
14 approved them if
they would allow the use of
15 a rapid assay for
that assessment as opposed
16 to requiring
culture, but that's not where we
17 are. I don't think
we're arguing with the
18 point you're
making.
19 DR. SIEGAL: Okay,
if it's brief.
20 PARTICIPANT: Could
I just comment
21 that the whole
blood derived platelets that
22 are prestorage
pooled are required by labeling
Page 257
1 and FDA to be
cultured by BacT/ALERT or AADS
2 so this move
towards whole blood derived is
3 not really
cutting into the safety.
4 The people
that are doing that,
5 the thought is
that day-three pool that is
6 cultured is
probably a safer and to me better
7 quality
platelet product, day three or four
8 anyway, than a
day six or seven as we're
9 talking now,
apheresis.
10 DR. KLEIN: We
don't really know
11 what percentage of
the whole blood derived are
12 pooled prestorage,
or do we? Do you? Are
13 they all being
pooled prestorage and cultured?
14 PARTICIPANT: No.
This is a
15 relatively recent
product approval. Obviously
16 people are better
developing it and using it
17 or using it mainly
because it's a culture-
18 based tested
product.
19 DR. SIEGAL: Dr.
Epstein, you had
20 a point? Jay, did
you have a point to make?
21 DR. EPSTEIN: Well,
just the point
22 that's been made is
correct which is that
Page 258
1 there is an
available system for prestorage
2 pooling which
does require a culture which
3 would make the
prestorage pooled platelet
4 comparable, if
you will, to the
5 bacteriologically screened apheresis platelet.
6 But our impression, and we don't
7 have hard data, is that system has not been
8 widely adopted and we're not exactly sure why
9 not. It is an approved system.
10 DR. KATZ: I can
address it. It's
11 operationally very,
very difficult in our
12 component
laboratories to use it so we are
13 looking at ways to
try to do it and we are
14 doing that for a
number of reasons. One is
15 bacterial testing
and the other is whether it
16 gives us a route to
TRALI mitigation
17 strategies that we
didn't have before.
18 If we can do that
instead of
19 deferring large
numbers of female donors
20 because they have
HLA antibodies of unknown
21 clinical
significance, can we prepool and
22 store platelets
from male donors. The
Page 259
1 operational
aspects of setting up that program
2 are really
daunting, but a number of us are
3 looking at
it.
4 DR. BRACEY:
Another point I think
5 we should make
is that this is a moving target
6 because, as
we've heard, one of the
7 accrediting
agencies will look at adding a
8 requirement
not to use dipstick testing so I
9 think whole
blood derived platelets that
10 aren't bacterially
tested that number will
11 grow as these
accreditation requirements
12 change. It's going
to be about a year or so,
13 I think,
though.
14 DR. SIEGAL: Okay.
So then should
15 we go to the
questions again?
16 DR. VOSTAL: So
just to restate
17 the question,
basically the first question is
18 asking whether the
reporting of an outcome of
19 the study should be
active or passive
20 following of septic
reaction.
21 Actually, we agree
with the
22 manufacturers that
any study that's done
Page 260
1 should follow
septic reaction rates but we are
2 trying to come
to an agreement in terms of how
3 that
monitoring should be done, whether it
4 should be
active or passive.
5 DR. SIEGAL:
Is there any
6 discussion on
that before we vote? We've had
7 a good deal of
discussion I think already.
8 Okay.
9 MR. JEHN:
Again, just to review
10 how we do the
voting now, it's simultaneous
11 voting. I would
like to get the yeas for the
12 first question with
a raise of their hands.
13 Keep your hands
raised until I go around and
14 call off your
name.
15 For the first
question we'll go
16 around the room,
you know, how many yeas do we
17 have.
18 DR. SIEGAL: Yea
means that we
19 favor active
testing.
20 MR. JEHN: Okay.
Keep your hands
21 up until I review
your name.
22 Ms. Baker, Dr.
Bracey, Dr. Cryer,
Page 261
1 Dr. Di
Bisceglie, Dr. Fleming, Dr. Glynn, Dr.
2 Klein, Dr.
Kuehnert, Dr. Kulkarni, Dr.
3 McComas, Dr.
Rentas, Dr. Zimrin, and Dr.
4 Siegal.
5 And nays. Dr.
Finnegan. Dr.
6 Ballow is an
abstention. That was it.
7 Dr. Katz, do
you have any
8 comments?
9 DR. KATZ: I
think what I want to
10 do is kind of
obvious. There is a little
11 unclarity here that
I didn't understand. Is
12 FDA proposing that
as some kind of condition
13 of transfusing
platelets that you get active
14 surveillance or
this in the context of a
15 study?
16 DR. VOSTAL: This
is in the
17 context of a study,
any future study to
18 validate seven-day
platelets.
19 MR. JEHN: Okay.
The votes were
20 13 yeas, one nay,
and one abstention.
21 DR. SIEGAL:
Maureen, did you want
22 to
comment?
Page 262
1 DR. FINNEGAN:
I actually think
2 the proposal
for two arms, one passive and one
3 active, is
probably the best choice.
4 DR. SIEGAL:
Well, let's proceed
5 to the second
question and its components.
6 DR. VOSTAL:
Okay. The second
7 question is,
"In addition to reporting of
8 sepsis does
the Committee agree with the FDA
9 that (a)
additional aerobic and anaerobic
10 cultures should be
performed on day five both
11 to increase the
safety of platelets on day six
12 and seven and as a
baseline measure?"
13 DR. KUEHNERT:
Okay. Here's where
14 I have a question.
So this qualifier "to
15 increase the safety
of platelets." Is this
16 saying that it's
more than just to enhance the
17 study information
or are you saying this needs
18 to be done to
ensure that the patients receive
19 safe blood
regardless of the study? That's
20 what I'm a little
confused about.
21 DR. VOSTAL: Well,
I think, right,
22 there is two
issues. If you are going to
Page 263
1 reintroduce
seven-day platelets to the market,
2 we feel that
there should be some additional
3 safety
intervention that would prevent
4 concerns about
septic reactions. We propose
5 that it should
be a culture at day five.
6 DR. KUEHNERT:
This may not be the
7 case but just
to stress the point. So even if
8 the data would
not help the study at all, get
9 useful
information, you're saying that would
10 this need to be
done in addition to ensure the
11 safety of the
patients?
12 DR. VOSTAL: Yes,
because we feel
13 that we have to --
to get them back on the
14 market we want to
put in this safety
15 intervention and
then we want to have the
16 study to make sure
that the intervention we
17 put in place is
actually doing what we expect
18 it to
do.
19 This is something
that we learned
20 from the initial
PASSPORT study. We put in
21 bacterial detection
early on in storage
22 assuming that it
was going to solve the
Page 264
1 contamination
problem but now in retrospect we
2 see that was
not fully the case.
3 DR. CRYER: So
you're saying -- I
4 mean,
basically what this says is that there
5 is an
intervention that says that you are
6 going to
culture the platelets, day-five
7 platelets. If
that culture is positive, the
8 patient
doesn't get them so patients are only
9 going to get
culture negative day six and
10 seven platelets or
day five culture negative
11 six and seven
platelets.
12 DR. VOSTAL:
Yes.
13 DR. CRYER: I got
it.
14 DR. DI BISCEGLIE:
I mean, this
15 does sort of assume
that a positive day-five
16 culture increases
safety. As Dr. Fleming
17 pointed out, it is
sort of a logical
18 assumption but you
don't have the data.
19 I mean, there is
kind of a jump
20 here from doing
this as part of the study to
21 leaping over to say
that if we reintroduced
22 day-seven
platelets, that's going to be part
Page 265
1 of the
algorithm to use them. I think Matt's
2 question is
quite correct.
3 It's a little
fuzzy, and I'm not
4 quite sure we
are given the choice in
5 responding of,
yes, it's a good thing to do as
6 part of the
study to gather the data. And as
7 Dr. Fleming
points out, it will probably be a
8 good thing to
jump into just saying this is
9 what needs to
be done.
10 DR. SIEGAL:
Jay.
11 DR. EPSTEIN: Would
it be helpful
12 to the Committee if
question 2(a) were split
13 into two parts,
additional aerobic and
14 anaerobic culture
performed on day five to
15 increase safety,
call it (a)(1). Additional
16 aerobic/anaerobic
culture to provide a
17 baseline, call it
(a)(2).
18 That lets you
advise us on the
19 respective benefits
of these two concepts, and
20 it would open the
door potentially if you
21 answered no to the
first part, you don't need
22 it for safety, but
you answered yes to the
Page 266
1 second part,
need it for a baseline, would
2 open the door
to culturing a subset to get a
3 measurement as
opposed to introducing the
4 five-day
culture to presumably enhance safety
5 on day six and
seven. I don't know how the
6 committee
would vote, but it would separate
7 the two
issues.
8 DR. SIEGAL:
I'm not sure how you
9 could get an
outcome so you need to do the
10 control
really.
11 DR. RENTAS: Is
rapid testing out
12 of the question as
far as the FDA is concerned
13 on day
five?
14 DR. EPSTEIN: Rapid
testing can be
15 done after 72 hours
in the context of a
16 previous culture.
I'm just talking about
17 Verax because that
is the only available
18 system at the
moment. It's consistent with
19 the product label.
It can be done. Whether
20 it's a sufficient
measure within the context
21 of the PASSPORT
study is an open question, and
22 the reason is that
we have reason based on the
Page 267
1 Irish data to
suspect that the contamination
2 levels on day
five are still likely to be low
3 because the
cultures were negative. You had
4 a high miss
rate. You picked up, I forget
5 what it was,
three or four times more on day
6 seven than on
day four. Nothing is positive
7 on day seven
that wasn't contaminated on day
8 four so you
had low sensitivity on day four
9 presumably
because the bacterial titers were
10 low. Why were they
low? Because you are
11 dealing with these
slow growers and they just
12 still hadn't come
up. Our thinking is that
13 knowing that the
available rapid test has
14 significantly lower
sensitivity compared to a
15 culture that we
would not see it as
16 sufficient. We have
no resistance to it being
17 added to the study.
If there is no quarantine
18 hold, perhaps it
has an added value of safety.
19 We just wouldn't
know, but you wouldn't have
20 deteriorated
anything. In other words, you
21 would use a
day-five product based on the
22 addition of having
grabbed a culture and doing
Page 268
1 a release --
I'm sorry, doing a rapid test.
2 DR. CRYER:
Would it be more
3 appropriate if
you are doing this intervention
4 to just say
that after day five that any
5 platelets
given have to be cultured 12 hours
6 before giving
them?
7 DR. EPSTEIN:
That is what FDA is
8 saying.
9 DR. CRYER:
No, it isn't. You're
10 saying five days.
It could be 48 hours later.
11 I'm saying if you
plan on giving one of those
12 platelets, then you
ought to culture them just
13 before you give
them regardless of what day it
14 is. If it's a
seven-day platelet, culture
15 them at seven -- at
six-and-a-half days and
16 not rely on
five-day data.
17 DR. EPSTEIN: Well,
okay. There's
18 a trade-off here.
The closer to time of issue
19 that you grab a
culture, the higher the
20 likelihood that the
contamination rate would
21 be high. Right?
But, on the other hand, you
22 have allowed less
time for organisms to grow
Page 269
1 in the
culture. You just don't know which
2 strategy wins.
We just don't know. I mean,
3 the ideal
thing would be you culture it at all
4 points of
issue and then we would actually
5 find out the
contamination rate based on days
6 of storage.
Nobody thinks that that much
7 culturing is
feasible.
8 DR. DI
BISCEGLIE: Not to try make
9 the question
the way we want it but I'm not
10 sure splitting it
helps. I think what would
11 be clearer in my
mind, in any case, if you
12 just take off the
second half of the question.
13 In the context of
the study should
14 additional aerobic
and anaerobic cultures be
15 performed on day
five, period, without the
16 rationale which is
maybe what I had my
17 questions about
because that rationale might
18 then be used to
make it mandatory
19 subsequently.
20 DR. KULKARNI: May
I make a
21 comment? I just
have trouble with the word
22 "safety." You're
talking about a surrogate
Page 270
1 marker for
culture positivity. I think I
2 agree with
you. I think we take off that last
3 thing about
safety and all that and just stop
4 it at day
five. That would just remove the
5 word "safety."
Safety for people means
6 different
things.
7 DR. SIEGAL:
It also sounds as
8 though
practically speaking day five cultures
9 aren't going
to get done in everybody so
10 you'll have a
built-in control group by
11 neglect.
12 DR. VOSTAL: But I
guess the main
13 question -- one of
the questions that we have
14 is what would it
take to get seven-day
15 platelets back on
the market? Do we have to
16 do anything
different from the way we were
17 doing at the
PASSPORT I study? We thought the
18 PASSPORT I study
was not safe to continue. If
19 we now decide it's
safe to continue, did we
20 actually introduce
anything else besides what
21 we used to
do?
22 DR. ZIMRIN: Well,
I thought the
Page 271
1 outline
suggested that there were several
2 changes. Do we
need to go over that again?
3 DR. VOSTAL:
The changes -- okay.
4 I think
--
5 DR. ZIMRIN:
You said there was 8
6 mL rather than
4.5 mL. There's the diversion
7 pouch. There's
the standardized skin -- I
8 mean, I don't
know, I think we already
9 discussed
this.
10 DR. VOSTAL: Yes.
So the question
11 will be is that a
sufficient safety feature
12 that will make
these seven-day platelets safer
13 than they used to
be.
14 DR. DI BISCEGLIE:
I would say the
15 investigators, I
think, are on the money that
16 they are studying a
clinical outcome. That
17 would be what would
reassure me. That's what
18 you've changed. You
are measuring in an
19 active manner a
clinically important and
20 relevant outcome.
What is the reason for the
21 cultures? It's to
support that. It's not in
22 and of itself
because it already failed in
Page 272
1 PASSPORT I. It
already failed as an outcome
2 but it's to
support the logical and clinically
3 relevant
outcome.
4 DR. VOSTAL:
Right, but we are
5 comparing a
readout of the study as a septic
6 reaction rate.
Any type of intervention we
7 did up front.
So to me I think we have to do
8 something to
increase the safety above what
9 was done in
PASSPORT I. Doubling the volume
10 is a potential
safety feature. However, some
11 studies suggest it
will only get you so far in
12 terms of increasing
sensitivity. Some
13 studies, as Lou
pointed out, that don't see
14 any increase in
sensitivity. So to me, you
15 know, that's why
we're bringing this question
16 to you, whether you
would feel that that's a
17 sufficient safety
intervention.
18 DR. SIEGAL: Dr.
Fleming.
19 DR. FLEMING: I
think it is --
20 DR. KATZ: I just
want to make the
21 point once more
that the bugs that grew on the
22 surveillance
cultures and the bugs that grew
Page 273
1 late in
PASSPORT the investigators, some of us
2 who transfused
a lot of blood over a long
3 period of time
were not sure that they are
4 important to
find. I won't give you anecdotes
5 about triple
platelet products that all grew
6 P. acnes late
and were transfused with
7 absolutely not
one shred of evidence that the
8 patient knew
they were getting P. acnes. That
9 is really the
issue. I think in our
10 estimation the
importance of what we're
11 missing is not as
great as the worst case
12 scenario.
13 DR. VOSTAL: But we
should point
14 out that the bugs
that were captured in the
15 surveillance in the
PASSPORT study were
16 actually not P.
acnes. There were three other
17 bacteria, but it
was not P. acnes.
18 DR. FLEMING: I
think we do have
19 to keep in mind
that there are two separate
20 issues here, two
separate purposes for this
21 assessment on day
five. One of those relates
22 to having a culture
assessment in addition to
Page 274
1 the sepsis
assessment, but a completely
2 separate issue
is whether we believe that the
3 use of day
six/seven needs to be done in the
4 context where
patients are given more
5 protection by
having an assessment at day
6 five. If we
do, then that does continue on in
7 clinical
practice if the study is positive
8 because if
it's positive in that context, you
9 haven't shown
that it's acceptable to use day
10 six and seven
without having done a day five
11 culture to filter
out those that would be
12 clinically
unacceptable.
13 DR. KLEIN: That's
correct, but if
14 you, for example,
didn't have that
15 information, you
did the culture and didn't
16 have the result,
then you would know whether
17 or not the culture
and the clinical results
18 correlated with one
another. Unfortunately it
19 doesn't look like
one can do that kind of
20 study.
21 DR. FLEMING: I
mean, ultimately
22 if resources were
unlimited, an approach here
Page 275
1 would be to
have the day one to five control
2 be compared to
day six/seven use without a
3 culture screen
at day five versus day
4 six/seven use
with a culture screen at day
5 five. That
would give us all three options,
6 but it would
obviously now be a three million
7 rather than
two million person trial.
8 So this
distinction here of the
9 reason we're
doing it we can't lose sight of
10 that fact. If you
believe it's only needed
11 for baseline and
using day six and seven
12 without a screen is
fine, then that is a
13 different
perspective than where you are
14 saying at this
point if we are going to answer
15 one question we
ought to answer the question
16 in the patient's
most protected interest.
17 If you are going
to give day
18 six/seven, are you
going to have a screen in
19 advance at day
five? I'm personally happy to
20 answer question (a)
as it is or to split it
21 into the two parts,
but I don't think we can
22 lose sight of the
fact that there are two
Page 276
1 distinct
issues that you may or may not
2 support as the
reason for the day five
3 culture.
4 MR. JEHN: So
are we voting as one
5 question or
two questions? Are we splitting
6 it?
7 One question.
Okay. So for the
8 2(a) let's see
the hand show for the yeas.
9 Okay, Ms.
Baker, Dr. Ballow, Dr. Bracey, Dr.
10 Cryer, Dr. Di
Bisceglie, Dr. Fleming, Dr.
11 Glynn, Dr.
Kulkarni, Dr. McComas, Dr. Rentas,
12 and Dr.
Siegal.
13 Nays? Dr. Finnegan
and Dr.
14 Zimrin. Are there
some abstains? Okay, Dr.
15 Klein and Dr.
Kuehnert. Dr. Katz, do you have
16 any
comments?
17 DR. KATZ: Only, I
think I've
18 probably said it,
that if this is required we
19 will do due
diligence and we will go to the
20 people that are our
partners in this in the
21 hospitals and we'll
see what can be done. I
22 will guess that
it's going to be a big hurdle,
Page 277
1 but we'll
see.
2 DR. KUEHNERT:
And I just wanted
3 to comment on
my abstention that it's very
4 difficult to
make a judgment whether the
5 cultures
should be done even for the study
6 design when
you don't know exactly what the
7 nature of the
surveillance is going to be or
8 the answer to
2(b).
9 MR. JEHN: So
we had 11 yeas, two
10 nays, and two
abstains.
11 DR. VOSTAL: Okay.
Moving onto
12 question 2(b). So
in addition to reporting of
13 sepsis, does the
Committee agree with the FDA
14 that surveillance
cultures should be performed
15 at outdate to
provide a bacteriological
16 endpoint for the
study?
17 DR. SIEGAL: Any
controversy about
18 this?
19 DR. DI BISCEGLIE:
We had very
20 little discussion
about the need for this, and
21 I guess I would be
interested in just hearing
22 from the agency
again the rationale for doing
Page 278
1 this,
supporting this.
2 DR. VOSTAL:
So here we're talking
3 about
surveillance cultures, and these are the
4 cultures that would be done at outdate of
5 seven-day platelets to confirm the previous
6 interventions that were done in terms of
7 decreasing contamination.
8 So without having that outdate
9 culture, you are really not going to know if
10 your sampling at day five allowed any products
11 to go forward. You can follow septic reaction
12 rates, but you've heard all the issues that go
13 along with looking at septic reactions.
14
Not everybody who gets a platelet
15 product gets a
reaction. People are on
16 antibiotics. It
depends on the bugs. To get
17 the most accurate
assessment of what your
18 intervention is
doing, we feel that having a
19 culture at the end
of day seven is the way to
20 go.
21 DR. CRYER: I'll
just comment that
22 I think that is an
experiment. If I was a
Page 279
1 hospital and I
had a bunch of platelets that
2 I was going to
throw away, I would probably
3 just throw
them away whether you wanted me to
4 culture them
or not.
5 If you really
think that 2(a) is
6 important in
that there is a safety measure
7 involved here,
I like the idea that I had
8 mentioned
earlier that you use the strategy of
9 culturing all
the platelets that you are going
10 to give just before
you give them rather than
11 that approach which
is going to give you the
12 same data that you
are missing rather than
13 this approach that
doesn't make much sense to
14 me.
15 DR. FINNEGAN: I
think you just
16 contradicted
yourself because you said that
17 the patient may not
have a reaction because
18 they are on
antibiotics. They got the
19 platelets they
needed and they are doing well
20 but the culture is
positive. I don't care.
21 I don't care if the
culture is positive. Why
22 would you care if
the culture is positive if
Page 280
1 the patient
did well?
2 DR. VOSTAL:
So this would be a
3 culture on
products that did not get
4 transfused.
That's what we're talking about.
5 Surveillance
cultures.
6 DR. FINNEGAN:
So they went out to
7 day seven and
nobody used them?
8 DR. VOSTAL:
Yes, outdated
9 products.
10 DR. SIEGAL: I
would care,
11 Maureen, because my
immune deficient patients
12 might be much more
susceptible than your
13 surgical
patients.
14 DR. FINNEGAN: No,
no. But what
15 he said was --
okay, I misunderstood because
16 what he said was
the patient got platelets.
17 They were on
antibiotics so it covered up the
18 infection that they
might have gotten and they
19 are doing well,
then that's the endpoint that
20 we want. It's not
what the culture shows.
21 DR. SIEGAL: But we
also want to
22 know the
denominator in effect. You want to
Page 281
1 know what the
real endpoint was at the optimal
2 time for
culture in that seven days.
3 DR. FINNEGAN:
Right, but --
4 DR. SIEGAL:
So it's helpful
5 information.
6 DR. FINNEGAN:
If you do that at
7 outdate, it's
going to be past the time you
8 would give it
to the patient so how valuable
9 is that
information?
10 DR. SIEGAL: Well,
the clinical
11 outcome is
important but it's not the only
12 outcome you're
looking for. It seems to me
13 that knowing that a
seven-day platelet was, in
14 fact, contaminated
is important in and of
15 itself.
16 DR. EPSTEIN: This
may be obvious
17 to many members of
the Committee but perhaps
18 not all. The
underlying concept here is to
19 compare the culture
positive rate of a seven-
20 day platelet
sampled at the end of seven days,
21 let's say on day
eight, to the culture
22 positive rate on
day five. If there is no
Page 282
1 culture on day
five, then there is really no
2 baseline for
comparison.
3 The
underlying point here is to
4 use a
bacteriologic endpoint to determine the
5 safety of the
seven-day versus five-day
6 platelet as
opposed to simply the septic rate
7 as an
endpoint. Again, the reason for asking
8 2(b) is
whether to utilize culture as an
9 endpoint and
it's implicit that it's in
10 comparison to a
day-five culture positive
11 rate.
12 DR. SIEGAL: Tom.
13 DR. FLEMING: I was
actually going
14 to make essentially
the same point. 2(a) is
15 critical because it
is inherent to the
16 clinical strategy
as well as to having an
17 endpoint, a culture
endpoint. 2(b) is just
18 specific to adding
the ability to do a
19 secondary or
co-primary endpoint based on a
20 culture.
21 I'm truly a
clinical endpoint
22 person so I
strongly endorse those that would
Page 283
1 say let's
really focus on the septic
2 transfusion
reaction rate. But what we've
3 noted is it is
unclear how well powered that
4 endpoint is
going to be. It's unclear whether
5 we are going
to get adequate specificity to
6 understand
whether we are focusing on the
7 specific
cause, and there are issues of
8 passive
capture, etc.
9 I am totally
with those that would
10 say I want to focus
on the clinical endpoint,
11 but there are
challenges with that clinical
12 endpoint. The point
is could you get enhanced
13 insight by also
having a culture endpoint as
14 a co-primary, as a
secondary endpoint. That
15 is what I see the
essence of 2(b) is.
16 As FDA says, it
would be used to
17 compare the culture
profile at day five versus
18 day six or seven. I
also agree with FDA
19 ideally day five
versus day seven but the
20 problem with that
may be numbers and power and
21 are we forced to do
day five versus day
22 six/seven to have
more power. That is the
Page 284
1 essence of
what 2(b) is for.
2 DR. SIEGAL:
Well, to be or not to
3 be. That is
the question.
4 MR. JEHN: All
right. Could I see
5 the yeas,
please?
6 DR. KUEHNERT:
Could I just ask
7 one point of
clarification? From what Dr.
8 Epstein said,
should we read into this that it
9 really is
saying if the answer to 2(a) is yes,
10 then should
surveillance cultures be performed
11 at outdate, etc. Is
that correct?
12 DR. FLEMING:
Indeed, because if
13 you weren't doing
it at day five, then there
14 is no comparative
at day six/seven so, yes,
15 it's implicit. If
you said yes to 2(a), then
16 should you go ahead
and do it at day
17 six/seven.
18 MR. JEHN: All
right. So for the
19 way it was worded
for 2(b), let's see a hand
20 for yeas, please.
Ms. Baker, Dr. Ballow, Dr.
21 Bracey, Dr. Cryer,
Dr. Di Bisceglie, Dr.
22 Fleming, Dr. Glynn,
Dr. Kuehnert, Dr.
Page 285
1 Kulkarni, Dr.
McComas, Dr. Rentas, and Dr.
2 Siegal.
3 Nays, Dr.
Finnegan and Dr. Zimrin.
4 DR. ZIMRIN: I
said no to 2(a) so
5 that's
basically why I said no to 2(b) also.
6 MR. JEHN:
Okay. Were there any
7 abstains? No
abstains this time.
8 Dr.
Katz?
9 DR. KATZ: If
we do day-seven
10 cultures, we will
find misses from day one and
11 day five, and we
admit that, and I think it's
12 more complication
and difficult to do, but we
13 will
listen.
14 DR. FINNEGAN: Not
to berate a
15 point but then we
are going to be using
16 laboratory data to
make a decision about
17 materials that we
need for clinical care and
18 that does not make
sense.
19 MR. JEHN: Dr.
Klein, I seemed to
20 have missed you.
What was your vote? Yea.
21 Okay. So the vote
is 13 yeas and two nays.
22 DR. SIEGAL: Okay.
I think there
Page 286
1 are no more
questions so we should adjourn
2 unless there
are objections. 2:15 I'm told.
3 (Whereupon,
the above-entitled
4 matter went
off the record at 1:31 p.m. and
5 resumed at
2:15 p.m.)
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Page 287
1
A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N
2 2:24
p.m.
3 DR. SIEGAL:
All right. Let's
4 start. This
afternoon our topic is iron
5 status in
blood donors. We are first going to
6 hear an
introduction from Dr. Leslie Holness
7 at
FDA.
8 Dr.
Holness.
9 DR. HOLNESS:
Good afternoon. I
10 hope everyone had a
good lunch.
11 The FDA seeks some
scientific
12 assessment from the
Committee on iron status
13 in blood donors.
There is evidence that
14 repeated blood
donation causes tissue iron
15 depletion defined
as low tissue iron stores in
16 the absence of
anemia.
17 In the cases of
iron depletion,
18 donor hemoglobin
may be below the individual's
19 normal level, but
remain above the acceptance
20 level for donation.
It remains unclear
21 whether iron
depletion in blood donors is a
22 public health
concern.
Page 288
1 The
objectives of the current FDA
2 requirements
regarding hemoglobin testing are
3 to protect the
health of the donor. The
4 expected drop
in hemoglobin from the donation
5 will not be
harmful to the donor. Frankly
6 anemic donors
should be deferred and may need
7 medical
evaluation.
8 On the
recipient side, the
9 hemoglobin
requirements ensure the
10 availability of
blood and ensure a minimum
11 product potency.
Later we will see potential
12 concerns related to
donor iron loss and the
13 impact on health
status of donors, especially
14 repeat donors.
There is a particular concern
15 that iron
deficiency in women blood donors of
16 childbearing age
may affect the future
17 development of a
fetus.
18 The current FDA
requirement for
19 minimum hemoglobin
level is found in 21 CFR
20 640.3(b)(3), an
allogeneic donor of either sex
21 must have a blood
hemoglobin level no less
22 than 12.5 grams of
hemoglobin per 100
Page 289
1 milliliters of
blood or hematocrit value of 38
2 percent.
3 In a proposed
rule from the FDA
4 released last
November entitled, "Requirements
5 for human
blood and blood components intended
6 for
transfusion or further manufacturing use,"
7 although the current hemoglobin requirements
8 are left untouched, we are specifically
9 soliciting comments and supporting data on
10 changing the hemoglobin level to 12.0 grams
11 per deciliter of blood or a hematocrit value
12 of 36 percent as an acceptable minimal value
13 for female allogeneic donors.
14 We are also asking
for information
15 on adverse events
in donors, changing the
16 inter-donation
interval, and the use of copper
17 sulfate method to
determine hemoglobin levels.
18 The comments
received are still being
19 evaluated.
20 This slide
reflects the ongoing
21 debate. The
rationale for setting any given
22 hemoglobin
hematocrit cutoff for blood
Page 290
1 donation is
not clear. In 1958 in a final
2 rule the
hemoglobin was 12.5 grams per
3 deciliter as a
lower limit for both sexes.
4 The interval
was not mentioned.
5 In 1963 in a
proposed rule the
6 maximum
donation interval was set at eight per
7 year and that
was not finalized. In 1980 the
8 agency
concluded that the 12.5 gram per
9 deciliter
limit was below the physiologic
10 normal for males so
in a proposed rule they
11 set a minimum of
13.5 grams per deciliter for
12 males and 12.5
grams per deciliter for
13 females. The
interval was set at six
14 donations per year
for both sexes. This is
15 also not
finalized.
16 The rule also
changed the donation
17 interval to four
times a year for females and
18 five times a year
for males. That was not
19 finalized. So we
come to 1989 and currently
20 the hemoglobin is
set at 12.5 grams per
21 deciliter or a
hematocrit of 38 percent for
22 both sexes and the
donation interval was
Page 291
1 changed to six
per year and that was finalized
2 and is the
present regulation.
3 There are
different pre-donation
4 hemoglobin
levels in several countries. For
5 males the
Council of Europe has set a minimum
6 of 13.5 grams
per deciliter for males. The
7 World Health
Organization, Australia, and the
8 UK have a
minimum hemoglobin of 13.0 grams per
9 deciliter.
Similarly for females the Council
10 of Europe has a
minimum hemoglobin of 12.5
11 grams per
deciliter.
12 The World Health
Organization,
13 Australia, and the
UK have 12.0 grams per
14 deciliter. The U.S.
and Health Canada have a
15 minimum of 12.5
grams per deciliter for both
16 sexes, and the
values do not take into account
17 diet, age, or
racial differences.
18 These are
available tests for
19 hemoglobin. These
three methods are
20 relatively cheap,
robust, and can be
21 implemented
rapidly. The gravimetric method
22 using copper
sulfate, spun hematocrit, and
Page 292
1 acid
hemoglobin method using the portable
2 photometer.
These biochemical tests for iron
3 status are
sensitive and better reflect the
4 iron status of
the blood donor.
5 Some of the
tests will be
6 mentioned by
today's speakers, zinc
7 protoporphyrin, mean corpuscular volume, or
8 MCV, serum ferritin, serum transferrin
9 receptor, and the log of the serum transferrin
10 receptor ferritin ratio, the hypochromic
11 percent of mature red cells, and the
12 hemoglobin content of reticulocytes.
13
Hemoglobin measurement is an
14 indirect measure of
iron stores. Blood donors
15 vary in their
baseline hemoglobin levels.
16 Premenopausal women
generally have lower
17 levels than men.
Studies show polymorphisms
18 of the
erythropoietin gene or its receptor may
19 be responsible for
the difference.
20 Dr. Sarah Cusick,
a micronutrient
21 specialist from
CDC, will show us normal
22 levels for
hemoglobin in a number of U.S.
Page 293
1 populations.
Donors may meet the hemoglobin
2 standard and
be iron deficient and donors can
3 fail the
hemoglobin standard and not be iron
4 deficient.
5 Iron levels
may fall with repeat
6 donation.
Sixty percent of the total
7 deferrals in
2002 to 2004 were for an
8 unacceptable
hemoglobin level often in
9 previously
accepted donors. Deferral for
10 hemoglobin level
while not a permanent
11 deferral may result
in the loss of the donor.
12 Side effects from
iron depletion
13 from donation may
include fatigue, Restless
14 Leg Syndrome, pica,
and progression to anemia.
15 Dr. Gary Brittenham
from Columbia University
16 will give us an
overview and information on
17 effects from
depleted iron stores.
18 Iron supplements
may help retain
19 donors who have low
iron levels by preventing
20 a decline of
hemoglobin below the acceptance
21 level.
Nevertheless, aside from some pilot
22 studies, few
establishments have implemented
Page 294
1 iron
supplementation for a variety of reasons.
2 Fear of masking underlying
3 disease,
including giving iron to donors who
4 may
over-absorb iron; a general concern about
5 blood establishments assuming medical care
6 responsibility for the donor; and the toxicity
7 of some iron preparations if inadvertently
8 made accessible to young children. However,
9 a few countries in Europe have ongoing
10 programs for iron supplementation.
11
Dr. Karin Magnussen from the
12 Copenhagen Hospital
will review the data from
13 an iron
supplementation program in Denmark.
14 Dr. Barbara Bryant
of the University of Texas
15 at Galveston will
discuss data gathered from
16 a pilot protocol on
iron supplementation while
17 she was at the NIH.
And Dr. Dan Waxman of the
18 Indiana Blood
Center will discuss the results
19 of a U.S. program
that has been in place for
20 some
time.
21 Several studies in
the U.S.,
22 Germany, Austria
report low hemoglobin as the
Page 295
1 largest single
cause for deferral. Minimum
2 recommended
hemoglobin for blood donation
3 remains
controversial, and there is no
4 universal
international standard. Hemoglobin
5 level does not
correlate well with iron
6 depletion.
7 Dr. Ritchard
Cable from the New
8 England Region
of the American Red Cross will
9 discuss data
from an ongoing REDS-II RISE
10 research study on
iron depletion in blood
11 donors.
12 So, in summary, if
regular repeat
13 donation induces
iron depletion and regular
14 repeat donors are
our safest donors, are we
15 harming our safest
donors? Are we losing
16 otherwise suitable
donors unnecessarily based
17 on a failure to
assess iron stores or to
18 provide iron
supplementation?
19 These are the
questions for the
20 Committee.
21 1. Is iron
depletion in blood
22 donors a
concern?
Page 296
1 2. If so, are
there tests for
2 iron status
that would be practical and
3 appropriate in
the donor setting?
4 Last
question, please discuss the
5 risks and
benefits of alternative strategies
6 to mitigate
iron depletion in donors
7 including:
8 a. iron
supplementation
9 b. dietary
recommendations
10 c. modification of
the inter-donation
11 interval
12 d. changing the
acceptance standard for
13 donor hemoglobin or
hematocrit. That's what I
14 have today. Thank
you.
15 DR. SIEGAL: Thank
you, Dr.
16 Holness.
17 Now we will have a
review of iron
18 metabolism and
impact of iron deficiency on
19 blood donors from
Gary Brittenham of Columbia
20 University.
21 DR. BRITTENHAM:
Good afternoon.
22 I have to begin by
saying that Paul McCurdy,
Page 297
1 an FDA iron
maven, told me that at the very
2 first meeting
of the Blood Products Advisory
3 Committee this
was the same topic so we'll
4 have to see
whether all the progress that's
5 been made in
understanding iron metabolism
6 since that
time has practical consequences.
7 My brief is
to discuss iron
8 metabolism and
how it relates to iron
9 deficiency in
blood donors. Perhaps I'll
10 begin by saying
that from my understanding the
11 principal group
that is affected by this, not
12 exclusively but
principally, are our safest
13 most reliable
donors, that is, women of
14 childbearing age. I
want to explain to you
15 why I believe
that's the case.
16 Let me start out
by just briefly
17 going through what
we understand about iron
18 metabolism. This is
a slide that you'll be
19 tired of by the
time we finish the
20 presentation
because I have tried to fashion
21 it so that it
explains and shows where iron is
22 and how it moves.
The various compartments
Page 298
1 here are shown
first in red. These are the
2 circulating
blood cells that contain most of
3 the iron in
the body.
4 The movement
of iron is from
5 transferrin,
the iron bearing protein, to the
6 erythroid
marrow where the iron is turned into
7 the red cells
to the red cell compartment
8 where it is
then as the red cells reach the
9 end of their
life span they are taken up by
10 specialized
macrophages that then recycle the
11 iron back to
transferrin. Each day there is
12 a circuit of this
of some 30 milligrams of
13 iron moving
around.
14 The exchange
otherwise I have
15 lumped together all
the other tissues in the
16 body here and
muscle and other parenchymal
17 cells. They have
limited iron requirements
18 and these are
basically to replace any iron
19 that is lost in
cells that die and to remove
20 that iron for its
recycling.
21 The liver, as
we'll see, has a
22 specialized role in
iron metabolism. If the
Page 299
1 serum iron
drops and becomes too low the liver
2 can donate
iron. If it's too high, it can
3 take up
iron.
4 The key
factor in iron metabolism
5 is its very
limited exchange with the outside
6 world so the
way the amount of iron in the
7 body is
controlled is by controlling
8 absorption
from the GI tract.
9 If there is
some three or four
10 grams of iron in
the body, 3,000 to 4,000
11 milligrams, then in
an adult male there is
12 only 1 milligram
that exchanges each day, 1
13 milligram that is
absorbed and another
14 milligram that's
lost.
15 What is lost is
simply lost in the
16 cells that are
excreted that are lost from the
17 body, really
physically lost. It's been said
18 that once an atom
of iron enters the body,
19 then it stays there
for 10 years so it shows
20 that this is one of
the most efficient
21 recycling schemes
known.
22 In red we show the
functional
Page 300
1 compounds. I
mentioned the green area here
2 that is the
transport iron. What we are to be
3 concerned with
are the iron that's in storage
4 either in
macrophages or in hepatocytes. This
5 shows the
major iron compartments and how iron
6 moves around
the system.
7 If you'll
bear with me, I'm going
8 to very
briefly take account of what's
9 happened in
our understanding of this. Really
10 in the past decade
there has been an explosion
11 of knowledge so we
have now identified many of
12 the molecular
mechanisms by which each of
13 these things
happen.
14 If we start with
this area, this
15 shows the GI tract
that is lined by
16 enterocytes and
iron enters through here
17 through a metal
transporter, enters the
18 enterocyte, and
then is exported through a
19 protein that is
called ferroportin that we'll
20 hear about into the
systemic circulation.
21 That iron is then
picked up by
22 transferrin and
carried to all the tissues in
++