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.
 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
  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.)
      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