Page 101
  1  points at which the pH provides a warning
  2  signal to prevent transfusion of bacteria in
  3  contaminated PCs.  
  4   It would be an exciting
  5  imagination to have an approach which is able
  6  to perform in pH all of the time on the basis
  7  of a mathematical algorithm the bag would
  8  produce a warning signal "you shouldn't use
  9  me."
 10   Coming to the summary of my
 11  presentation, there is a strong need for
 12  clinical studies on rapid bacteria detection
 13  methods. This is the text I found here in the
 14  preparation paper from FDA. "Rapid bacteria
 15  detection devices have not been validated in
 16  clinical settings and, thus, use of this
 17  device during the clinical trial would
 18  validate its clinical sensitivity and possibly
 19  justify its use as a stand-alone quality
 20  control of release test."  
 21   This sentence I would like to
 22  support very intensively. Additionally, I am
          
      Page 102
  1  allowed to inform you that we just had a
  2  meeting last week, last Thursday, in Amsterdam
  3  with Dr. Kofta.  
  4   I'm sure you know him as the head
  5  of a lot of studies in microbial safety of
  6  blood components. He is going to organize a
  7  study comparing early and post-storage
  8  sampling and including rapid bacterial
  9  detection methods, flow cytometry, general
 10  detection of universal bacteria, PCR and
 11  respectively NAT.
 12   I would like to close with a kind
 13  of summary of my presentation. Post-storage
 14  rapid bacteria detection cannot guarantee
 15  sterility of tested PCs and will not prevent
 16  transfusion transmitted bacterial infections.
 17   Why? Let's imagine rapid methods
 18  having a sensitivity of one per sample, or
 19  what I could say is a sample can say nothing
 20  at all about the residual volume of the
 21  platelets. But rapid post-storage bacteria
 22  detection can identify highly contaminated
          
      Page 103
  1  PCs.  
  2   These are the products which are
  3  leading to immediate septic shock and maybe
  4  organ failure within one or two hours and
  5  frequently to death. That means post-storage
  6  rapid bacteria detection in PCs is capable to
  7  prevent fatal cases in transfusion medicine.
  8  That is the best we can do. Thank you for
  9  your attention.
 10   DR. SIEGAL: Thank you very much,
 11  Dr. Montag-Lessing.
 12   We'll next hear from Louis Katz on
 13  the Redesign of the PASSPORT Study for Seven-
 14  day Platelets.
 15   Louis.
 16   DR. KATZ: I guess I want to thank
 17  my colleagues in the PASSPORT working group
 18  for volunteering me to do this presentation.
 19  PASSPORT actually means something. We won't
 20  dwell on it.  
 21   Essentially what I would like to
 22  do today is show you what we know so far about
          
      Page 104
  1  the impact of the reversion from seven to five
  2  days. What we know about results from
  3  PASSPORT to-date, tell you what we think we
  4  did wrong in the original design of PASSPORT
  5  and what we would like to try to do to bring
  6  seven-day platelets back.
  7   The reason we want to bring seven-
  8  day platelets back is the availability issue.
  9  I think everybody sitting around the table
 10  understands that basically we are trying to
 11  get what we think is an important product to
 12  the bedside of patients.  
 13   Seven-day platelets enhanced our
 14  ability to do that and the reversion has cost
 15  us some of the benefit that we saw we had. On
 16  the other side, we have to balance that
 17  against the adequacy issue against the safety
 18  issue that has been well described by the
 19  first two speakers.
 20   This is a summary of PASSPORT.
 21  There was an approved device -- two approved
 22  devices, in fact. The kits are from Gambro,
          
      Page 105
  1  now Caridian BCT and Fenwal. The bags were
  2  approved for storage of platelets for seven
  3  days based essentially on some in vitro
  4  studies and recovery and survival assays in
  5  platelet recipients. We knew we had a bag in
  6  the absence of bugs that allowed us to store
  7  platelets for seven days and they appeared to
  8  work. It's an approved device.
  9   What we needed to know then, and
 10  one of the conditions of dating platelets for
 11  seven days from FDA and by consensus to people
 12  that were interested in seven-day platelets,
 13  was to revisit the issue of the safety vis-a-
 14  vis platelet associated sepsis that had led to
 15  a reversion from seven to five days in the
 16  1980s.  
 17   We had a primary hypothesis that
 18  seven-day platelets would be noninferior to
 19  uncultured five day platelets. That is the
 20  standard of care at the time we were thinking
 21  about this using a microbiologic endpoint.
 22  Perhaps mistake No. 1.
          
      Page 106
  1   We would then evaluate the
  2  performance, sensitivity, specificity,
  3  positive and negative predicted values of a
  4  two-bottle release test. Two bottle meaning
  5  an aerobic and anaerobic bottle in the
  6  BacT/ALERT.
  7   The prevalence of bacterial
  8  contamination for the two-bottle test would be
  9  determined and we would gather information as
 10  well on how useful the anaerobic bottle was
 11  when included in the release test. There was
 12  no comparison of the clinical safety of six
 13  and seven-day stored platelets to younger
 14  platelets.
 15   For those of us distributing
 16  platelets from blood centers, I think we
 17  really were more interested in the clinical
 18  outcome and perhaps should have insisted up
 19  front that we think about that. Water under
 20  the bridge.
 21   These are the participating
 22  organizations. Right in the middle of nowhere
          
      Page 107
  1  is me and the rest. We are pretty nicely
  2  distributed around the country. Accrual
  3  began, I believe, at the New York Blood Center
  4  in the fall of 2005 and then we accrued
  5  subsequent blood centers and a few hospitals
  6  as time rolled on.
  7   We've estimated that in calendar
  8  2008 had we continued to produce seven-day
  9  platelets for the entire 12 month period we
 10  would have produced 400,000 doses of seven-day
 11  apheresis platelets which would be between a
 12  quarter and a third of the platelet doses
 13  being transfused in the U.S. if we extrapolate
 14  from the 2005 National Blood Collection and
 15  Utilization Survey.
 16   So there were a bunch of
 17  assumptions that you've heard about previously
 18  at prior BPACs and a little bit from us that
 19  we used to calculate the sample size we needed
 20  for the primary endpoint which was the
 21  microbiologic status of platelets at outdate.
 22   We have the release test which is
          
      Page 108
  1  two bottles inoculated between 24 and 36 hours
  2  after collection of the platelet, incubated
  3  through the shelf life of the platelet. Then
  4  we had surveillance cultures which is
  5  culturing platelets that didn't get
  6  transfused, culturing the outdated platelets.
  7   Without going through this because
  8  I think you've heard it before, what we needed
  9  to show was less than four in 50,000 outdated
 10  platelets contaminated in order to have 90
 11  percent confidence of a less than 5,000 false
 12  negative risk.  
 13   Again, this is a microbiologic
 14  endpoint and proved to be difficult to meet
 15  because cultures turned positive at a higher
 16  rate than we expected, and also because
 17  getting outdated platelets became very
 18  difficult because the seven-day dating on the
 19  platelets reduced outdate rates very, very
 20  substantially in most of the centers that were
 21  participating.
 22   Here is the basic protocol. As I
          
      Page 109
  1  told you, hold the platelets 24 to 36 hours to
  2  allow bacteria to leave lag phase. Then 4 mL
  3  each in an anaerobic and aerobic BacT/ALERT
  4  culture bottle. Incubate culture 24 hours
  5  before release to allow early detection so we
  6  call it a release test. Continue incubation
  7  through outdate.  
  8   Recall products with positive
  9  culture after release. That is, interdict
 10  platelets hopefully before they are transfused
 11  and we were quite successful. And then the
 12  second surveillance test on some outdated
 13  platelets to estimate the false negative
 14  release test rate.
 15   Well, January 29th. One of my
 16  favorite days in blood banking. Early this
 17  year we got an e-mail from the study sponsors.
 18  I think they were still Gambro then. Does
 19  anybody in the audience -- yes, still Gambro?
 20  They are Caridian now -- and Fenwal based on
 21  primarily data from the surveillance cultures
 22  felt that we were detecting too many and that
          
      Page 110
  1  there might be a clinical risk to patients and
  2  we should stop the study. That was January
  3  29th.
  4   We have attempted to assess the
  5  impact on operations by surveying very
  6  recently the people who were participating.
  7  In late August, in anticipation of the BPAC
  8  meeting we contacted the 33 principal
  9  investigators and asked them a series of
 10  questions. We got responses from 21.
 11  Eighteen of 21 saw increased outdates and I
 12  don't think that's surprising. We knew that
 13  very quickly after the suspension.
 14   A fill rate is the percentage of
 15  orders from a hospital or a hospital system to
 16  a blood collector that are filled with what
 17  was requested. At my center we run fill rates
 18  of 99.5 percent and higher for seven-day
 19  platelets. When a hospital orders from us we
 20  are able to give them what they requested. I
 21  will make no comments on the clinical
 22  appropriateness of platelet transfusions.
          
      Page 111
  1   Four of 21 stopped using the
  2  anaerobic bottle. There is no regulation from
  3  FDA and no standard from ABB that says how to
  4  do this. Eight of 21 reduced the hold time
  5  after culture including my center so we were
  6  required in PASSPORT to hold for 24 to 36 --
  7  excuse me, to hold for 24 hours after the
  8  culture was inoculated. And because of
  9  increased outdates eight of the 21 have
 10  reduced the hold time.  
 11   That is, we will send the bags to
 12  our hospitals a little bit quicker in order to
 13  meet their needs. That ranges from no hold
 14  time after inoculation up to 28 hours. It is
 15  essentially in response to having repeatedly
 16  particularly on Tuesdays and Wednesdays to do
 17  emergency release prior to the nominal
 18  incubation period of cultures rather than
 19  routinely do emergency release.  
 20   A number of us have changed our
 21  protocol. Five of 21 have delayed their TRALI
 22  mitigation plans. I will very briefly a
          
      Page 112
  1  little later on remind you of the risk model
  2  that our group has produced. Three of 21 have
  3  increased their distribution of untested
  4  whole-blood derived platelets. We do not have
  5  data on production of whole-blood derived
  6  platelets by non-participants and what that
  7  might have done.
  8   Okay. So this is the outdate
  9  data. Fifteen of the 18 respondents who noted
 10  increased outdates were able to provide
 11  quantitative data. We have gone from a median
 12  outdate rate at those 15 centers of about 4
 13  percent up to a median of about 10 percent.
 14  You can see the ranges in the box.
 15   If we extrapolate this to the
 16  entire 400,000 estimate of PASSPORT production
 17  during calendar 2008, this is probably 20,000
 18  to 25,000 additional units of apheresis
 19  platelets that will outdate and a total in the
 20  PASSPORT participants of about 40,000
 21  outdates.
 22   One caveat is that these are not
          
      Page 113
  1  normalized -- the estimate is not normalized
  2  for the individual center platelet production
  3  so the number is soft but I'm pretty sure in
  4  the ballpark.
  5   So this is just data, one table
  6  from the risk assessment we showed you in May.
  7  I think the point of this table is that the
  8  degree we think that seven-day platelets are
  9  replaced by uncultured whole-blood derived
 10  platelets we may, in fact, add incremental
 11  risk of septic transfusion reactions to the
 12  blood supply.
 13   As I told you on the last slide,
 14  we have seen at least some modest move to
 15  whole-blood derived platelets. In particular,
 16  at my center we have a number of export
 17  customers purchasing 25 to 30 percent more
 18  whole-blood derived platelets from us than
 19  before the suspension of the study.
 20   We don't really have a good number
 21  nationally to estimate how many more whole-
 22  blood derived platelets are being distributed
          
      Page 114
  1  and the degree to which that is a result of
  2  the suspension of PASSPORT.
  3   We also in that model, and I won't
  4  show you the data, we are concerned about the
  5  implementation of TRALI mitigation strategies
  6  as a result of the suspension of PASSPORT and
  7  increased outdates. As I showed you, I think,
  8  on the last slide there are several centers
  9  that are delaying plans that they had made.
 10   These were the definitions that we
 11  had for release test results, true positive,
 12  false positive, etc., down the line. If they
 13  were not transfused, that is interdicted
 14  platelets had a positive confirmatory test
 15  essentially got the bug out a second time.
 16   Indeterminants, we had a positive
 17  initial but didn't get a confirmatory or data
 18  from the confirmatory was not interpretable
 19  like a different bug for example. Okay, so we
 20  had definitions. In the transfused we had
 21  definitions as well. The data on the
 22  transfused platelets came back from the
          
      Page 115
  1  hospitals to us when we identified a release
  2  test positive and inquired.  
  3   In the case of false negatives if
  4  an event occurred at the hospital and they
  5  reported it to us. The false negatives are a
  6  bit problematic. That is a fairly passive
  7  surveillance system and I think that is a flaw
  8  in the way we did the study.
  9   This is the 320,000 plus single-
 10  donor platelets that were made during the
 11  duration of the study and the microbiologic
 12  results. True positives one in 5,000. A
 13  fairly high incidence of false positives and
 14  indeterminants. In the transfused platelets
 15  you can see here we transfused 13 true
 16  positive, 41 false positives, 200 and some
 17  indeterminants, and two false negatives.
 18   The clinical outcomes available
 19  from these I'll discuss a little more. We had
 20  somewhat over 200 units transfused that we
 21  have some level of concern about plus the two
 22  false negatives. There were no deaths. Paul
          
      Page 116
  1  Ness' estimate published in 2001 but from the
  2  late '90s without bacterial culturing at the
  3  Hopkins estimated a death rate of 15 per
  4  million.
  5   We had 14 transfusion reactions
  6  that were possibly related to bacterial
  7  contamination for a rate of about one in
  8  23,000 or 44 per million and Paul Ness again
  9  suggested 70 per million in the Hopkins data
 10  and I think this comports nicely with what
 11  Mark Brecher showed you. We have actually
 12  come a fair distance with this issue but I
 13  don't think anybody would claim that we are
 14  perfect.
 15   These are the 14 reactions in 21
 16  recipients who received false negative, true
 17  positive, and indeterminant. Once again we
 18  see that not everybody who gets bacteria in a
 19  bag gets sick but some do.
 20   Whether all 14 of these represent
 21  reactions to bacterial contamination is
 22  unclear. Larry Dumont, I believe, showed you
          
      Page 117
  1  the line listings of these reactions at the
  2  May meeting and there is probably a third to
  3  a half of them that from a clinical standpoint
  4  I don't think are imputable to bacterial
  5  contamination. Nevertheless, we count them
  6  because that is what the rules were and this
  7  will bear on our proposal to bring back
  8  PASSPORT.
  9   This is the day of storage at the
 10  time of reactions so of the 14 reactions three
 11  occurred in the critical six and seven days of
 12  extended storage. It begs the question of the
 13  denominator and I will once again plead bad
 14  data regarding our understanding across large
 15  numbers on the day of transfusion during the
 16  dating of the platelets.
 17   This is some data on that issue
 18  that
 19   Mark Brecher has published, Mark
 20  Brecher's group has published. The red
 21  columns are 2005 before seven-day platelets to
 22  green after. As you can see, I added this up
          
      Page 118
  1  and now I can't remember. It looks to me like
  2  seven or eight percent of their seven-day
  3  platelets were transfused on days seven and
  4  eight after the implementation of the PASSPORT
  5  approach.
  6   These are data from my center in
  7  the month prior to the suspension of PASSPORT.
  8  As you can see, here are day six and seven.
  9  This is five or six or our largest platelet
 10  users for just under four weeks. The bottom
 11  line is 23 percent of the platelets we
 12  distributed to these hospitals using a fair
 13  number of platelets were transfused on days
 14  six and seven.  
 15   Peter Tomasulo of Blood Systems
 16  has provided me data that tells us that about
 17  17 percent of the seven-day apheresis
 18  platelets made at Blood Systems in Scottsdale
 19  and most of the western United States were
 20  distributed to their hospitals on days six and
 21  seven.
 22   Here is the surveillance testing
          
      Page 119
  1  results that kind of got us into trouble. We
  2  made 320,000 single donor platelets and we
  3  were able to do surveillance testing on 4,369,
  4  three of which were positive when the data has
  5  been accrued after the closure of the study.
  6  Now, when the study was closed I believe the
  7  number was two out of 2,700 and some.  
  8   While statistically we hadn't
  9  failed the 50,000 benchmark in the study, I
 10  think we all agree that at two in 2,700 or
 11  three in 4,000 and something we were unlikely
 12  to make it to 50,000 with four or fewer
 13  positive surveillance cultures. There was a
 14  single donation that grew staph. aureus.  
 15   This was on day eight and you can
 16  see the timing, that the bottles turned
 17  positive. There was a split donation, a
 18  double collection, that is, that grew a
 19  coagulate negative staph. and then another
 20  single donation that grew viridans strep. The
 21  confirmatory test grew something else. I'm
 22  not sure I know how to interpret that one.
          
      Page 120
  1   Okay. Where was PASSPORT in the
  2  universe of studies that looked at bacterial
  3  contamination? This data was shown to you
  4  previously in May by Larry Dumont. If we look
  5  -- there we go. If we look at it as if we
  6  were using a single bottle test, that is, an
  7  aerobic only, and compare it to the Red Cross'
  8  data where they do that, comparable rates in
  9  the single.  
 10   If we look at two-bottle test in
 11  comparison to the Irish, the Welsh, I think
 12  you'll see that with the two-bottle test the
 13  rates that we found are pretty consistent.
 14  These blues are culturing outdated platelets
 15  and so I think we are detecting in PASSPORT
 16  what other people have been detecting. We're
 17  not an outlier in either direction.
 18   Well, what are our conclusions so
 19  far? First of all, two bottles find more than
 20  one in PASSPORT than in other studies. We
 21  don't know whether that is a particular
 22  characteristic of the anaerobic bottle or the
          
      Page 121
  1  fact that you've doubled your sample size in
  2  order to inoculate two bottles. I think more
  3  likely it's a combination of the two.
  4   A very recent publication in
  5  Transfusion, Leon Su and Peter Tomasulo and
  6  the crowd of blood systems did not find an
  7  advantage for the use of the second bottle.
  8  I think it's kind of three to one in favor of
  9  two bottles and why it might work better we've
 10  talked about.
 11   The surveillance did identify the
 12  three false negative release tests. We know
 13  that bags with bugs may or may not be
 14  associated with symptoms of a septic
 15  transfusion reaction. The relative clinical
 16  risk of day six and seven to day five have not
 17  been established so we need clinical, not
 18  strictly microbiologic outcomes, is the
 19  mindset that those of us making platelets for
 20  the hospitals have come to. Can we get there
 21  is in part up to this group I suppose.
 22   Our first thought was can we
          
      Page 122
  1  implement culture latent storage or more
  2  likely a rapid test near to point of release
  3  in the hospital transfusion service
  4  preferentially before the platelets are
  5  transfused. Certainly in the blood center, in
  6  most blood centers in the U.S., we do not have
  7  the product latent storage. Okay? It's in a
  8  hospital on a platelet shaker awaiting need in
  9  the cardiovascular OR so we don't have control
 10  to do the reculture which would be certainly
 11  the most sensitive assay that we could do.
 12   Our transfusion services as
 13  PASSPORT principal investigators have polled
 14  their hospitals. Unenthusiastic is an attempt
 15  at understatement. It's not universal. There
 16  are hospitals that are perfectly willing to
 17  use the licensed assay with nominally 20 to
 18  30-minute turnaround time for release of
 19  platelets.  
 20   There are a number of strategies
 21  that could be used, testing part of your
 22  inventory at the beginning of the shift, this,
          
      Page 123
  1  that, and the other thing. But nearly
  2  universal refrain is lack of staff and lack of
  3  money so they are really not enthusiastic and
  4  the primary response most of us have gotten
  5  is, you just get us plenty of five-day
  6  platelets and seven-day platelets is your
  7  problem.  
  8   We have big shoulders so we're
  9  trying to figure out how to bring back seven-
 10  day platelets. What we would propose to do is
 11  this kind of laundry list and maybe some other
 12  things that I'm showing you in the bullets.
 13   No. 1, standardize the skin
 14  preparation. Skin prep was not specified in
 15  the PASSPORT protocol and, in fact, there is
 16  literature that tells us maybe we should
 17  mandate that the diversion pouch be
 18  appropriately used in all cases. Actually, I
 19  think that's a standard or in the new
 20  standards of the ABB now so we will have no
 21  choice.
 22   Can we enhance the sensitivity of
          
      Page 124
  1  the release test? Perhaps. Can we provide
  2  clinical surveillance to make a comparison of
  3  day five to say six and seven? In addition,
  4  I think there is interest particularly at FDA
  5  and perhaps among the investigators about
  6  labeling changes on platelets so that people
  7  are more aware of the risk of septic
  8  transfusion reaction. I'm not going to say
  9  much about that.  
 10   So standardized skin prep. There
 11  are data from blood centers, blood banks, and
 12  in the IV device and blood culture literature
 13  that tell us, for example, that tincture of
 14  iodine and/or chlorhexidine may be superior to
 15  povidone.  
 16   The principal investigators would
 17  like to standardize the application of skin
 18  preparation to one of these two methods and
 19  require that all participants use a specified
 20  set of protocols for skin preparation.
 21   Diversion of the initial aliquot
 22  you've heard a lot about. It is helpful. We
          
      Page 125
  1  currently divert, I think, about 35 CCs in my
  2  center and we use that blood to do infectious
  3  disease testing in our viral laboratories so
  4  it works out very nicely. That is the skin
  5  flora. As you've heard, bacteremias with gram
  6  negative rods or bacteremias from a bad tooth
  7  or osteomyelitis are not addressed by
  8  diversion.
  9   Can we enhance the sensitivity of
 10  the release test? The short answer is yes, we
 11  can. There is very substantial literature
 12  that those of us who do infectious diseases or
 13  clinical microbiology are familiar with. Over
 14  a very long period of time we have argued
 15  about how much blood should we put in the
 16  blood culture bottle.
 17   We have blood culture data and
 18  modeling data on platelet contamination that
 19  strongly suggest that volume is critical.
 20  This is some clinical data and if you look at
 21  -- you don't want to look at the endocarditis
 22  patients. They tend to have a continuous
          
      Page 126
  1  relatively high-grade bacteremia and the
  2  advantage of higher volumes and endocarditis
  3  is not so substantial.  
  4   In non-endocarditis patients
  5  doubling the volume from 10 to 20 mL gets you
  6  about a 30 percent increase in sensitivity and
  7  then a variety of other scenarios. The bottom
  8  line is it's a Poisson statistic issue. If
  9  you miss the bug by sampling too little, you
 10  miss the bug. That's pretty straightforward.
 11   These are data from Steve Wagner
 12  and others at the American Red Cross and they
 13  demonstrate that going from 4 mL to 8 mL in
 14  this model gets you 20 to 25 percent increase
 15  in the sensitivity of culturing.  
 16   Pretty consistent with what we saw
 17  in doubling the volume in the clinical setting
 18  so we think that a larger volume may be
 19  something that we should discuss. Then
 20  clinical surveillance and we think this is
 21  probably the most important thing.
 22   The PASSPORT investigators
          
      Page 127
  1  consider in a certain sense that the
  2  microbiologic endpoints we have used
  3  previously are surrogates and if our real aim
  4  is to mitigate the occurrence of septic
  5  transfusion reactions, we should use something
  6  closer to clinical surveillance.
  7   We used microbiologic surveillance
  8  because we thought that's what we could do.
  9  At the end of the day you get what you pay
 10  for. We didn't pay much and we didn't get
 11  much. Because our assumptions about the
 12  sensitivity of the release tests were probably
 13  optimistic, we got ourselves into a trap with
 14  sample size and stopping -- well, it wasn't a
 15  formal stopping rule that was applied but we
 16  made it pretty impossible for this study to
 17  succeed.
 18   So we are asking ourselves can we
 19  do simple clinical surveillance for vital
 20  signs consistent with sepsis during and
 21  following transfusion for a subset of a
 22  hypothetical PASSPORT to study.  
          
      Page 128
  1   Hospitals would return the data
  2  absent protected health information to a
  3  central facility where DSMB blinded to the day
  4  of storage at the time of transfusion would
  5  classify the reported events with prospective
  6  criteria for sort of inputability as a septic
  7  transfusion reaction or no.  
  8   Then we would do the comparison of
  9  seven-day labeled platelets transfused on day
 10  five with those on day six or seven, stopping
 11  rules to be agreed upon. Of course, the
 12  application or the approach that we are
 13  putting out there today is going to be
 14  substantial negotiation amongst the PIs, the
 15  sponsors, and the agency.
 16   This is not real. This is just a
 17  first mock-up of the kind of form that we
 18  would ask the participating hospitals to
 19  provide us on every transfusion that is
 20  labeled seven days in a participating hospital
 21  transfusion service.  
 22   It is very straightforward. I'm
          
      Page 129
  1  not going to go through it but I think it
  2  makes the point of the kind of things that we
  3  would -- the kind of information we would like
  4  to accrue if we are able to reconstruct a
  5  study.
  6   The primary outcome would be non-
  7  inferiority based on the rate of events. We
  8  need to do sample size calculations. If we
  9  get signals that this is an approach that can
 10  be fleshed out further, we need to sit down
 11  and figure out with the hospitals that we
 12  serve what is doable.  
 13   I don't want to go and present a
 14  pig in a poke to my hospitals so we want to be
 15  a little bit further down the line and go and
 16  ask the hospitals, "Are seven-day platelets
 17  worth it to you to provide us with a standard
 18  information set on these platelets?"  
 19   Stopping rules will be agreed on
 20  and then the sponsors to the FDA with a final
 21  proposal. These are the people that are at
 22  this point participating and trying to get
          
      Page 130
  1  this up and running again. I thank you for
  2  your attention.
  3   DR. SIEGAL: Thank you, Louis.
  4   Now we'll hear from the FDA from
  5  Salim Haddad on their perspective on the
  6  reintroduction of seven-day platelets in the
  7  PASSPORT redesign.
  8   MR. HADDAD: Good morning. My
  9  presentation I will be talking first -- I will
 10  give you first an overview of two proposals to
 11  redesign the PASSPORT study. One is FDA's.
 12  The other one is the plan that Dr. Katz just
 13  presented to you. We had an opportunity to
 14  review it at FDA earlier this summer.  
 15   Then I will make an assessment of
 16  the individual measures included in the two
 17  protocol proposals. Then I will present the
 18  overall FDA plan and the modifications
 19  relative to the original PASSPORT study. Then
 20  I will have the questions for the Committee.
 21   This slide and the next one have
 22  side by side the main features of the two
          
      Page 131
  1  proposals. Both proposals pertain to
  2  apheresis platelets and the extension of their
  3  shelf life from five to seven days. Both
  4  proposals will be using a standard skin
  5  preparation and diversion pouch to use the
  6  initial load of bacteria in the collection.
  7   Both proposals will be testing the
  8  platelet products on day one between 24 to 36
  9  hours after collection using the BacT/ALERT
 10  detection device both in aerobic and anaerobic
 11  bottle. It is an 8 mL sampling volume in each
 12  bottle which is double the sampling volume of
 13  the original PASSPORT study.
 14   Both proposals will propose to
 15  reduce the whole period prior to release to
 16  less than 24 hours. The main difference
 17  between the two proposals is that FDA is
 18  proposing to retest the product on day five
 19  and at outdate.
 20   Another difference in the primary
 21  outcomes FDA proposes to use the bacterial
 22  contamination rate and the septic transfusion
          
      Page 132
  1  rate at day seven and compare it to day five
  2  and the criterion of success will be that the
  3  rates would be no higher than day five and the
  4  industry proposals will be using only the
  5  septic transfusion reaction on day six and
  6  seven and again compare them to day five.
  7   In terms of tracking specific
  8  transfusion reactions FDA favors an active
  9  process. When we initially reviewed that
 10  protocol in the summer it was unclear whether
 11  they would be using inactive or a passive
 12  process I think from Dr. Katz' presentation
 13  they will be using an active process with the
 14  form to be dispensed with every product.
 15   However, I think there will be
 16  collecting data on a subset of the transfused
 17  product. Both proposals recommend an enhanced
 18  follow-up on the septic transfusion reactions.
 19  FDA recommends tracking the age of all
 20  transfused products and both proposals
 21  recommend a team of experts to evaluate the
 22  septic transfusion reactions.
          
      Page 133
  1   As Dr. Katz mentioned, the
  2  industry will propose some labeling on the
  3  product to mitigate the risk of transfusion of
  4  seven-day platelets. I will elaborate further
  5  on all those items.
  6   Going through the different
  7  measures nobody would argue against using a
  8  standardized skin preparation and the
  9  diversion pouch because most bacteria
 10  isolated from contaminated blood products are
 11  normal skin flora. Many studies have shown
 12  that disinfection after donor phlebotomy site
 13  may significantly reduce the bacterial
 14  contamination.
 15   However, skin disinfection does
 16  not guarantee sterility due to inaccessibility
 17  to organisms present deep in the skin and the
 18  sebaceous glands and in the hair follicles.
 19  The bacteria may be introduced into the blood
 20  container by means of a skin core during
 21  phlebotomy which may be diverted into the
 22  pouch.
          
      Page 134
  1   Several studies have demonstrated
  2  the effectiveness of diverting a small aliquot
  3  from the initial collection and reducing the
  4  bacterial contamination of the collected
  5  product. As an illustration this is a graph
  6  that Dr. Brecher showed you earlier. This is
  7  a plot of the rate of bacterial contamination
  8  which was analyzed by collection technology.
  9   As Dr. Brecher mentioned, either
 10  the use of a pouch properly placed on the draw
 11  line will significantly reduce the bacterial
 12  contamination in the collected product. This
 13  is data from the Red Cross from their
 14  surveillance between 2004 and 2006.  
 15   At that time they were using a 4
 16  mL sampling volume into a single aerobic
 17  bottle testing on the one only and not at
 18  outdate. The sample was at least 24 hour
 19  vessel collection with a 12-hour hold period
 20  and then the platelets were five-day
 21  platelets.
 22   After the collection the sampling
          
      Page 135
  1  time is important because a pre-sampling
  2  storage period would allow any bacteria
  3  present in the platelet bag to proliferate and
  4  the length of the pre-sampling period depends
  5  on the sensitivity of the device for the
  6  BacT/ALERT. This was set at the minimum of 24
  7  hours using spiking studies.
  8   The role of the sample volume. As
  9  I mentioned earlier both proposals recommend
 10  doubling the sampling volume on testing on day
 11  one with the bacteria load from 4 mL to 8 mL.
 12  We know that at collection the size of the
 13  initial contaminating bacteria inoculum is
 14  small and is subject to sampling error in the
 15  sense that the sample may miss organisms that
 16  are present in the collection which may later
 17  on go on to proliferate in the back.
 18   The presence of bacteria in the
 19  platelet product at collection because it's a
 20  low concentration can be considered a rare
 21  event and is subject to the Poisson
 22  probability distribution. The Red Cross did
          
      Page 136
  1  studies using the Poisson distribution
  2  analysis model to compare the percent bacteria
  3  detection in an 8 mL sample versus a 4 mL
  4  sample.  
  5   This is a graph that Dr. Katz
  6  showed you earlier. Here we have on the Y
  7  axis the percent bacterial protection. On the
  8  X axis this is the concentration range of the
  9  initial contaminating inoculum. The upper two
 10  curves represent the probability curves which
 11  were devised by the model for an 8 mL and a 4
 12  mL sample.  
 13   The bottom curve represents the
 14  absolute increased and improved detection by
 15  going from a 4 mL to an 8 mL sample. You see
 16  at low bacterial concentration an increase in
 17  sampling volume will increase the percent of
 18  bacterial detection.  
 19   However, at the other end when you
 20  have high concentrations the increase in
 21  sampling volume would have a zero impact on
 22  the improved percent detection.  
          
      Page 137
  1   The conclusions from the study is
  2  that the maximum theoretical absolute increase
  3  in sensitivity by going from 4 mL to 8 mL is
  4  about 20, 25 percent over the inoculum range.
  5   However, this is data based on a
  6  model and this is data generating by testing
  7  actual products. This is from the Irish blood
  8  transfusion study which was described earlier.
  9  Here they used a large sampling volume, 7.5 to
 10  10 mL and two bacteria load bottles, one
 11  aerobic and one anaerobic.  
 12   They tested both apheresis
 13  products and single units of whole blood and
 14  platelets. The apheresis were tested at a
 15  mean time of 17 hours after collection and the
 16  pools greater than 36 hours.  
 17   They tested on day one, day four,
 18  and at outdate after day seven and the day one
 19  confirmed positive rate was 320 in a million.
 20  That's 1.6 in 5,000. On the four the
 21  contamination rate was 300 in a million and at
 22  outdate it was 840 per million.
          
      Page 138
  1   So what the data showed is that a
  2  significant number of misses on day one and
  3  the sensitivity of the day one testing as been
  4  calculated to be at about 22 percent. This is
  5  slightly different from the sensitivity that
  6  is in the article by Murphy because to be
  7  consistent with other studies only the
  8  confirmed positive rate was used in this
  9  calculation whereas in the article the
 10  confirmed and the unconfirmed positives were
 11  used.  
 12   So a sensitivity of about 22
 13  percent for the one so that means at about 75
 14  percent of false negative rate for day one
 15  testing using large sample volume 7.5 to 10 in
 16  each bottle.
 17   As a comparison this is the
 18  passport study using the BacT/ALERT 4 mL in
 19  each of two bottles sampling at 24 to 36
 20  hours. The day one testing showed a confirmed
 21  positive rate of 234 per million. The outdate
 22  rate was 680 per million, again showing a high
          
      Page 139
  1  false negative on day one testing. The
  2  sensitivity was calculated to be about 26
  3  percent.
  4   So our conclusion from all those
  5  studies on the impact of the sampling volume
  6  is that doubling the volume doesn't have the
  7  sensitivity of the detection. However, the
  8  increase is limited and doubling the volume is
  9  unlikely to be a sufficient method to ensure
 10  the safety of seven-day platelets and their
 11  introduction to the market.
 12   After sampling there is the whole
 13  period prior to release and this is the time
 14  period between sampling and the release into
 15  inventory and it allows for the bacteria
 16  present in the sample, the proliferate in the
 17  BacT/ALERT bottles. In the PASSPORT study it
 18  was set at 24 hours.
 19   However, we recommend that to
 20  reduce it to 12 hours at the blood centers and
 21  implement a robust notification system to
 22  alert the transfusion services of any positive
          
      Page 140
  1  result.
  2   Actually, due to operation
  3  constraints such as distribution, release into
  4  inventory, the actual time that the product
  5  would be available for transfusion is at least
  6  24 hours after culture incubation. The ARC
  7  study, which I mentioned earlier, demonstrated
  8  the feasibility of this approach because out
  9  of 293 true passive components 292 were
 10  infected.
 11   The advantage of reducing the 12-
 12  hour hold period is a gain of 12 hours in the
 13  product shelf life and this advantage is that
 14  there is a low risk of transfusing a positive
 15  unit.
 16   As I mentioned earlier, one
 17  difference between the two proposals is that
 18  FDA is recommending retesting the product on
 19  day five and the purpose of the day five
 20  testing is two-fold, to improve the safety of
 21  the six and the seven platelets and to
 22  establish a baseline contamination rate at day
          
      Page 141
  1  five as a comparator for day six and seven.
  2   Day five was selected because the
  3  day four testing, as was shown by the Irish
  4  study, was insufficient and limiting to day
  5  seven contamination. And the ARC study showed
  6  that there is a spike in septic reactions on
  7  day five including fatal reactions.
  8   The disadvantages of performing a
  9  day five testing is that it is performed not
 10  in the blood center but in the hospital
 11  setting and that poses logistical and cost
 12  burdens. However, these may be offset by the
 13  extension of the platelet storage which may
 14  lead to better management of inventory and
 15  decrease in outdate.
 16   Also by a projected decrease in
 17  the septic transfusion rate with a
 18  corresponding decrease in hospital resources
 19  and services needed to treat complications
 20  from septic transfusion reactions. Another
 21  disadvantage is that the false positive would
 22  lead to product waste.
          
      Page 142
  1   FDA is recommending testing at
  2  outdate. That would be on the 8th. This was
  3  part of the original PASSPORT study and the
  4  industry proposal does not recommend retesting
  5  at outdate.  
  6   The advantages of retesting at
  7  outdate is that it provides definitive
  8  information on the effectiveness of the
  9  upfront measures and enhances in that sense
 10  the long-term safety of seven-day platelets.
 11   It establishes the contamination
 12  rate at day seven and defines the performance
 13  characteristics of the bacteria detection
 14  device. Disadvantages are the cost and
 15  logistics and it does not contribute to the
 16  immediate safety of the seven-day platelets.
 17   Industry recommends substituting
 18  reporting of septic transfusion reactions for
 19  outdate testing. The advantage is that
 20  obviously it investigates directly with
 21  clinical outcomes in patients. The
 22  disadvantage, it would be a disadvantage if
          
      Page 143
  1  it's a passive process but we heard it's going
  2  to be an active process.
  3   It's a subjective process because
  4  it can be confounded by other causes of sepsis
  5  in patients with underlying conditions. It
  6  may be less sensitive than the contamination
  7  rate and it would require a large operation
  8  study to the lower rates of septic transfusion
  9  reactions compared to bacterial contamination.
 10   The difference between active and
 11  passive surveillance have been highlighted in
 12  this study by Dr. Yomtovian and Dr. Jacobs.
 13  This is a unique study. It is the first long-
 14  term perspective surveillance study of
 15  bacterial contamination of platelets that has
 16  been published in the literature. It started
 17  in '91 and it is still ongoing.
 18   They had both type of
 19  surveillance, active and passive. In passive
 20  the trigger is a report of a transfusion
 21  reaction in a patient and it's followed by
 22  evaluation of the patient and the culture of
          
      Page 144
  1  the remaining content of the bank. The active
  2  surveillance in the study consisted of
  3  culturing all platelets at issue and follow-up
  4  with patients on positive platelet cultures.
  5   From 1991 through the entire study
  6  passive surveillance was performed throughout.
  7  However, it was the only type of surveillance
  8  between 2000 and 2004. After surveillance was
  9  initiated in 1991 it was suspended in 2000 and
 10  then resumed in 2004.  
 11   The bars represent the rates of
 12  bacterial detection in the platelets at issue
 13  per quarters. You notice that there is a gap
 14  here. No contaminated platelets were detected
 15  in the time period where only the passive
 16  surveillance was conducted.
 17   Out of 52 bacterially contaminated
 18  platelets that were detected throughout the
 19  surveillance, 50 were detected by active
 20  surveillance and two by passive surveillance.
 21  More to the point of interest, the septic
 22  transfusion reactions they had 16 detected by
          
      Page 145
  1  active surveillance and two by passive
  2  surveillance for an odds ratio of about 10.
  3   So the study concluded that the
  4  majority of patients who received bacterially
  5  contaminated platelets and developed signs and
  6  symptoms of sepsis were not recognized or
  7  reported to the transfusion service for
  8  further evaluations so there were a number of
  9  septic transfusion reactions that were missed.
 10   Based on this conclusion, FDA
 11  recommends that a one-page case report be
 12  perspectively provided with every issued
 13  product to the transfusion clinic entity to
 14  return after a 24-hour evaluation of the
 15  recipient.
 16   The purpose is that when a patient
 17  have received a transfusion and subsequently
 18  developed signs and symptoms of sepsis, that
 19  would be a heightened awareness to investigate
 20  a potential transfusion related septic
 21  reaction.
 22   Other measures in the proposal
          
      Page 146
  1  would be tracking the age of transfused
  2  platelets. In the original PASSPORT the
  3  implemented risk for septic transfusion
  4  reactions between days five and seven could
  5  not be assessed because the study did not
  6  include the reporting of the age of the
  7  transfused product so there was a numerator
  8  but no denominator.  
  9   So FDA recommends reporting the
 10  age of all transfused product rather than a
 11  subset and the advantage is that it would
 12  establish an accurate correlation between the
 13  incidence of sepsis and the age of the
 14  platelets.
 15   Another measure is the enhanced
 16  follow-up on septic transfusion reaction. The
 17  original PASSPORT protocol included provisions
 18  for clinical follow-up during septic
 19  transfusion reaction investigation and this is
 20  where they are reporting patient symptoms,
 21  blood culture on the patient and also culture
 22  of the bag remaining content.
          
      Page 147
  1   However, these measures were not
  2  always enforced and only in five out of 14
  3  suspected transfusion reactions where patient
  4  and bag culture and the nine remaining cases
  5  remained inconclusive. So the enhanced
  6  follow-up would maximize data collection and
  7  improve analysis.
  8   This panel of expert suggestions
  9  was made initially by industry and we adopted
 10  it. The original PASSPORT protocol lacked a
 11  case definition of septic transfusion
 12  reaction. We believe the panel of experts
 13  would be useful in providing a case definition
 14  and adjudicating cases.
 15   Also, the panel thinks it would be
 16  helpful in defining stopping rules because the
 17  original passport study had no stopping rules
 18  since the high contamination and septic
 19  transfusion rates were not anticipated and
 20  these were related to the unexpectedly low
 21  clinical sensitivity of the day one testing by
 22  BacT/ALERT.
          
      Page 148
  1   Dr. Katz mentioned the labeling
  2  that would be added to the platelet bag and
  3  this would consist of the results of the
  4  bacterial contamination from PASSPORT stating
  5  that testing does not guarantee sterility and
  6  also that the risks increase with the age of
  7  platelets, and making a secondary bacteria
  8  testing at issue.
  9   We believe that the first three
 10  measures would shift the burden and risk onto
 11  the user. Regarding the use of the seven-day
 12  rapid bacteria test at issue, as was mentioned
 13  earlier, this is in reference to Verax PGD
 14  test. This is a rapid bacterial test and it
 15  was cleared by FDA as an adjunctive QC test
 16  based on spiking studies due to its slow
 17  sensitivity.
 18   It may not be used as a substitute
 19  for culture-based tests because it was not
 20  cleared as a release test due to lack of
 21  clinical performance data. The spiking study
 22  provide data on the analytical sensitivity of
          
      Page 149
  1  the device. However, to determine the
  2  clinical performance such as sensitivity,
  3  specificity, and predicted value, data that
  4  are generated by testing actual products need
  5  to be generated.
  6   So in summary the FDA overall plan
  7  to redesign the PASSPORT study will consist of
  8  a new measure at collection which are
  9  standardized collection methodology, diversion
 10  pouch, and standard skin preparation. Then on
 11  day one the sampling volume will be doubled
 12  from 4 to 8 mL.  
 13   The whole period will be reduced
 14  to 12 hours at the end of which positive units
 15  would be discarded and negative units would be
 16  made available for transfusion through day
 17  four. Units that were not transfused would be
 18  retested on the morning of day five and this
 19  is a new measure.  
 20   After a 12-hour hold period
 21  positive units will be discarded and negative
 22  units will be made available for transfusion
          
      Page 150
  1  through the end of day seven. Units that
  2  outdate would be retested again with the
  3  BacT/ALERT system.
  4   Additionally, the septic
  5  transfusion reactions would be tracked
  6  prospectively and the data sampling monitoring
  7  board would evaluate septic transfusion
  8  reactions and define stopping rules. The
  9  outcomes of the study would be the bacterial
 10  contamination rate and septic transfusion rate
 11  at day seven. The concept objective is no
 12  increased risk of day seven compared to day
 13  five.
 14   The objective of our proposal is
 15  to improve the safety of day seven platelets
 16  by reducing the contamination and septic
 17  transfusion rate associated with day five
 18  through day seven platelets to allow the
 19  reintroduction of seven-day platelets into the
 20  market. And to provide definitive information
 21  on the effectiveness of the upfront measures
 22  on the contamination of seven-day platelets.
          
      Page 151
  1   Here are the questions to the
  2  Committee. These questions have been modified
  3  compared to what you previously received. We
  4  restructured them to focus on the main
  5  differences between the two proposals.  
  6   Question 1. Does the Committee
  7  agree with FDA that reporting of sepsis should
  8  be active and not passive?
  9   Question 2. In addition to
 10  reporting of sepsis does the Committee agree
 11  with FDA that:
 12   (a) Additional aerobic and anaerobic
 13  cultures should be performed on day five both
 14  to increase the safety of platelets on day six
 15  and seven and as a baseline measure?
 16   And does the Committee agree with
 17  the FDA that:
 18   (b) Surveillance cultures should be
 19  performed at outdate to provide a
 20  bacteriological endpoint for the study?
 21   This concludes my talk.
 22   DR. SIEGAL: Okay. It's now a
          
      Page 152
  1  little bit overtime. We are scheduled for a
  2  break. Let's take just 10 minutes and be back
  3  in 10 minutes if there are no objections.
  4   DR. KLEIN: Mr. Chairman, will the
  5  Committee have an opportunity to ask questions
  6  of any of these speakers? Is there a time
  7  when the Committee will be able to ask
  8  questions of the speakers or is that not
  9  programmed into our schedule?
 10   DR. SIEGAL: Yes, of course.
 11  We'll do this in the discussion following the
 12  break. Thank you.
 13   (Whereupon, the above-entitled
 14  matter went off the record at 10:57 a.m. and
 15  resumed at 11:12 a.m.)
 16   DR. SIEGAL: Okay. There appear
 17  to be no volunteers for the open public
 18  hearing. If there is anyone who wishes to
 19  speak at the open public hearing at this
 20  point, please indicate that to the Chair.
 21   If not, and if we can have a
 22  little decorum in the room, we should begin
          
      Page 153
  1  with questions from the Committee members to
  2  the speakers from this morning. Are there any
  3  questions?
  4   Dr. Bracey.
  5   DR. BRACEY: Yes. I've got a
  6  question and it's not directed to a specific
  7  presenter but --
  8   DR. SIEGAL: Excuse me for just
  9  one second.
 10   Please, would the people in the
 11  room who are speaking either leave the room or
 12  sit down and be quiet. Thank you.
 13   DR. BRACEY: We know from what
 14  we've heard that these are not sterile
 15  products. We know that we have some
 16  information about colony-forming units and
 17  their likelihood to cause septic reactions,
 18  but I would like to hear a little bit more
 19  about what is thought -- a little bit more
 20  about the threshold for a septic reaction
 21  related to the colony-forming units noting
 22  that, I believe, the figure that was quoted
          
      Page 154
  1  from Jacobs was somewhere around 104 that is
  2  a single institution study.  
  3   What I'm wondering is is there
  4  information from units that have been
  5  interdicted that give us more information
  6  about CFUs and clinical outcomes.
  7   DR. SIEGAL: Is there anyone who
  8  would like to respond to that question from
  9  the speakers?
 10   DR. BRECHER: Can you hear me?
 11  Okay. To be honest, I'm not sure that anyone
 12  really knows what the actual level is. The
 13  active surveillance at the University Hospital
 14  of Cleveland was they set up a culture on all
 15  the platelets going out the door but I believe
 16  it was one tenth of a mL on a plate so their
 17  sensitivity was maximally 10 CFUs per mL.  
 18   If there were cases with lower
 19  levels of bacteria, they would not have been
 20  aware of those cases. The data that I
 21  presented from the NIH Clinical Center was
 22  from the early '70s when apheresis platelets
          
      Page 155
  1  only had a shelf life of one day so I would
  2  not have expected the salmonella to have been
  3  at a high concentration.  
  4   Yet, eight patients became ill on
  5  average eight to nine days after the
  6  transfusion. The data from the heparin
  7  flushes is also very worrisome that small
  8  amounts of bacteria can go into a patient
  9  perhaps colonize a catheter or an organ
 10  somewhere and show up later and we don't know
 11  what the significance of low levels of
 12  bacteria are.
 13   DR. MONTAG-LESSING: Thomas Montag
 14  from Paul-Ehrlich Institute in Germany. The
 15  problem is that usually the bacteria are not
 16  quantified or there is no chance to quantify
 17  that even in cases when there was an analysis
 18  in BacT/ALERT because that is only a
 19  qualitative diagnosis, yes or no.  
 20   What I can tell you our
 21  experiences from analysis up to assessments to
 22  prosecute in fatal cases when I could see the
          
      Page 156
  1  clinical data. In all of these cases in which
  2  I could analyze the platelet bags. Empty or
  3  not, that is not important because we need
  4  microbes for counting bacteria in which within
  5  a few minutes shaking chills happen.  
  6   Within one or two hours mighty
  7  organ failure happened in which I would like
  8  to express it as follows. The poor patient
  9  survived over a few days because of the
 10  intensive care mechanisms. Yet, no chance for
 11  a normal life after that. In all these cases
 12  the bacteria count was very, very high.
 13   That means at least 108 per mL and
 14  you have to multiply with the whole unit.
 15  That means three times 1010 or so.
 16  Additionally toxins were there.  
 17   In the case of the negative
 18  picture it is a very easy to understand. All
 19  of them are releasing endotoxins and the fever
 20  threshold of human beings is 50 picogram per
 21  mL. We found sometimes the 1010 fold of this
 22  fever threshold.
          
      Page 157
  1   This leads to immediate monocyte
  2  activation to a very extensive release of the
  3  pro-inflammatory cytokines, mainly TNF, tumor
  4  necrosis factor, the dangerous cytokine. All
  5  these cases we could analyze haven't been
  6  repeated. Under the circumstances it's a
  7  very, very high count.
  8   You remember my last slide. My
  9  thinking as a microbiologist and I spend 15
 10  years in clinical microbiology at the
 11  University of Berlin too. If you would have
 12  a rapid method and let's imagine the
 13  sensitivity of 103 per mL, then we could say
 14  one mL is less than 103. The whole platelet
 15  unit has less than three times 106.  
 16   This would lead in most of the
 17  cases from my experience from clinical
 18  microbiology and from Paul-Ehrlich Institute
 19  as a federal agency to, lets call it, the
 20  usual hospital infection of the patient which
 21  can be handled in most of the cases.  
 22   That means we can treat the
          
      Page 158
  1  patient with antibiotics and so on and so on
  2  but it's not possible at all when we bring the
  3  patient immediately into the septic shock and
  4  mighty organ failure. That is my opinion
  5  after 10 years or more of thinking about what
  6  can we do.  
  7   What we can do is to save lives.
  8  We are not able to guarantee sterility of
  9  blood components by neither procedure
 10  including pathogen reduction but we can try to
 11  save lives. That's my view.
 12   DR. SIEGAL: Maureen.
 13   DR. FINNEGAN: Actually it was
 14  pretty interesting this morning because I
 15  almost never write anything down and I've got
 16  almost a whole page here. I think in sort of
 17  summarizing it, it looks to me like we are
 18  comparing apples and oranges in two areas.
 19   One is in the bacterial
 20  contamination in that salmonella is going to
 21  make you sick but porphyromonas is probably
 22  not going to make anybody sick. Klebsiella
          
      Page 159
  1  may make certain number of people sick.
  2  Staph. epi. is probably only going to make a
  3  few people sick. I think that is one of the
  4  things we need to work out is how do we figure
  5  out is there a class of bacteria where a
  6  colony count is more important and is there a
  7  class of bacteria where colony count is less
  8  important.
  9   Then I think the other place that
 10  is apples and oranges is the patient
 11  population. Basically the patients that need
 12  platelets are either immune compromised, heme-
 13  onc patients or they are surgery or trauma
 14  patients that are otherwise perfectly healthy.
 15  If you put a small contaminated dose into a
 16  perfectly healthy person, you probably aren't
 17  going to get a problem. Whereas if you put a
 18  small dose into somebody who is significantly
 19  immune compromised you will. I think that is
 20  verified by Dr. Haddad's comment that
 21  substituting reporting of septic transfusion
 22  reactions is a problem because there are lower
          
      Page 160
  1  rates of STRs compared to bacterial
  2  contaminations which would suggest that, in
  3  fact, the contamination is not the problem.
  4  The problem is where does the rubber hit the
  5  road. Where is the contamination of a certain
  6  bacteria in a certain patient population going
  7  to be a problem.
  8   DR. SIEGAL: Any commentary on
  9  that point?
 10   Louis, you look like --
 11   DR. CRYER: I'll just make a
 12  comment on that last one. While I agree with
 13  you that the trauma patients are by in large
 14  perfectly healthy, by the time they need
 15  platelets they are not healthy anymore.
 16   DR. FINNEGAN: Right, but
 17  porphyromonas is probably not going to make
 18  them die.
 19   DR. SIEGAL: Louis.
 20   DR. KATZ: Well, I think just to
 21  reiterate the point of the PASSPORT group is
 22  that we think the important issue is to
          
      Page 161
  1  clinical outcomes. While originally the
  2  original design of PASSPORT we thought it was
  3  impossible. If we really want seven-day
  4  platelets, we think that we have to figure out
  5  a way to get clinical information across a
  6  substantial sample of patients.  
  7   The microbiologic results are
  8  fascinating, of course, but they are surrogate
  9  for what we are really interested in. I think
 10  most of us look at this as impacting very much
 11  on our ability to get enough platelets where
 12  they are supposed to be.
 13   Susan Leitman asked me a question
 14  during the break regarding if you look at the
 15  PASSPORT study and if you look at fatalities
 16  reported in the literature, there is a
 17  preponderance of gram negative organisms.
 18   Larry Dumont and I discussed this
 19  during the break and we didn't miss the gram
 20  negatives in PASSPORT. I think the amount of
 21  progress we've made, particularly with the
 22  most lethal bugs, is quite substantial.
          
      Page 162
  1   DR. SIEGAL: Other commentary or
  2  questions?
  3   DR. ZIMRIN: I think I was
  4  actually very happy to see the comparison of
  5  the two studies and realize that in contrast
  6  to maybe other meetings, as was alluded to
  7  earlier, that there is a lot of things we can
  8  agree on, that the platelet contamination is
  9  a serious problem, that strides have been made
 10  to improve the situation and has been
 11  successful.  
 12   I agree with Dr. Finnegan in terms
 13  of the different patient populations. I would
 14  also like to mention being a clinician and
 15  dealing with the heme-onc patients the
 16  complexity of the clinical situation and maybe
 17  a possible under-appreciation of infection
 18  which is why I would support strongly the use
 19  of active surveillance of reports that would
 20  look for this information rather than waiting
 21  for a response.
 22   Mark Brecher's presentation was a
          
      Page 163
  1  little concerning suggesting that possibly
  2  even with the kind of information we would get
  3  that we still might be missing clinically
  4  important outcomes. On the other hand, the
  5  kind of detailed evaluation that went into the
  6  patients that he described is certainly not
  7  feasible for the vast majority of patients
  8  that are getting platelets.
  9   I guess one question that I have
 10  that Dr. Katz addressed briefly was that the
 11  information that would be obtained in the FDA
 12  study is wonderful information. I would as a
 13  clinician, as a scientist I would love to know
 14  the bacterial culture results at different
 15  time points. I would like to be able to
 16  correlate this with release tests. I think
 17  this would really get us further ahead.
 18   My question, though, is that I
 19  didn't get a sense of how doable any of this
 20  is. We live in resource-constrained times and
 21  I'm certainly not an economist but I know that
 22  hospital administrators have a tendency to
          
      Page 164
  1  look at the bottom line. I didn't get any
  2  kind of feeling for whether this beautifully
  3  designed study that we might recommend can
  4  actually be implemented.  
  5   The other question I didn't really
  6  get a feel for at all, and it might be beyond
  7  the scope of this meeting, is what we are
  8  going to give up when we do that. Again, Dr.
  9  Katz alluded to it but that was just with
 10  regard to his study. I didn't see that kind
 11  of analysis with regard to the FDA study.  
 12   If there is anyone who has given
 13  some thought to this and really has a feeling
 14  for whether this is a study that is going to
 15  be doable and what in a broader sense we'll
 16  give up to do it I would love to hear that.
 17   DR. SIEGAL: There are also some
 18  statistical questions that we need to address
 19  but Louis has a response.
 20   DR. KATZ: The FDA's algorithm and
 21  our algorithm have not been arrived at
 22  independently. We have been talking to the
          
      Page 165
  1  agency, Jaro and his staff, about this right
  2  along.  
  3   We think what we're proposing is
  4  doable and we don't think what they are
  5  proposing in the real world is doable without
  6  a lot of funding which that is another issue
  7  that we will have to talk about later I
  8  suppose. A point of fact. If you're in the
  9  UK the collectors control the product
 10  basically throughout its lifetime and you
 11  could do that sort of thing.  
 12   I have a region that is a
 13  relatively small blood bank that runs from St.
 14  Louis in the south to Minnesota and Wisconsin
 15  in the north and east to west about 300 miles
 16  and our platelets when they're released go to
 17  hospitals where they sit to be used. I don't
 18  have control to ask the hospitals on day five
 19  to then quarantine the product, do the
 20  culture, hold them for 12 hours without
 21  release, etc. That is not going to happen.  
 22   They don't have computer systems
          
      Page 166
  1  or personnel that would allow that to occur so
  2  that is the reason that we arrived largely at
  3  the two different approaches. We just don't
  4  think the way that blood is distributed in
  5  this country, that is from the collector to
  6  the hospital, where it sits during shelf life
  7  waiting for transfusion is feasible.
  8   DR. BRECHER: Mark Brecher. I
  9  would second Louis' comments. Many of the
 10  hospitals where these platelets are sitting at
 11  do not have a BacT/ALERT machine and so how
 12  would they culture a BacT/ALERT platelet at
 13  that time? Logistically I don't see how it
 14  can happen.
 15   DR. SIEGAL: Yes.
 16   DR. VOSTAL: If I could just
 17  address some of these discussion points.
 18  Besides looking at the appropriate design of
 19  these studies our priority is really to make
 20  sure that if you're going to bring seven-day
 21  platelets back on the market that they are
 22  going to have an additional safety feature so
          
      Page 167
  1  we don't have to pull them off the market a
  2  year and a half later like happened this time
  3  around.  
  4   Just looking at what history
  5  taught us, back in 1985 when seven-day
  6  platelets were introduced the first time, BPAC
  7  made the decision a year later, or a year-and-
  8  a-half later, that there were too many septic
  9  reactions associated with them so BPAC pulled
 10  them off the market.  
 11   Then it took us another 20 years
 12  to come up with a device that would assure, or
 13  at least we thought would assure, increased
 14  safety for seven-day platelets.
 15   When we put that into place a
 16  year-and-a-half later we realized it wasn't
 17  sufficient and there were concerns, not
 18  necessarily from the FDA but from the
 19  participants in the study that seven-day
 20  platelets were not safe and that we should
 21  pull them off the market.  
 22   So if you are thinking about
          
      Page 168
  1  reintroducing platelets, seven-day platelets,
  2  we have to come up with sufficient safety
  3  features or safety intervention that will
  4  allow us to feel comfortable that we are
  5  transfusing a safe product so that's a
  6  separate issue from discussing what is the
  7  appropriate design of this study. Also we
  8  should think about what is the appropriate
  9  safety intervention.
 10   DR. SIEGAL: Dr. Klein, I think
 11  you had a comment.
 12   DR. KLEIN: Yes. I wanted to know
 13  -- this is directed, perhaps, towards Dr.
 14  Haddad and Dr. Katz -- can't you have an
 15  active surveillance of patient outcomes, that
 16  is reactions, does that have to be passive?
 17  That might get around the issue of having to
 18  have your culture reflect your active
 19  surveillance. Granted, that is not going to
 20  be easy but it would really give you the data
 21  that you want.
 22   DR. KATZ: Well, I think that's
          
      Page 169
  1  our proposal but our proposal for active
  2  surveillance is on the microbiologic side less
  3  rigorous than the FDA has proposed. I want
  4  everybody to understand that the PASSPORT
  5  group absolutely in the best of all possible
  6  worlds would do what was described by FDA.
  7   The question is the operational
  8  feasibility. We are all pretty sure that in
  9  the real world that complicated approach on
 10  day five is not going to work.
 11   DR. KLEIN: Perhaps the CBER
 12  people would address that because you seem to
 13  suggest that active surveillance is only
 14  microbiologic and it seems to me it could be
 15  clinical.
 16   MR. HADDAD: When I discussed the
 17  active surveillance that pertained to the
 18  follow-up on the septic transfusion reaction
 19  because currently a lot of septic transfusion
 20  reaction are missed. Our proposal by defining
 21  the surveillance as active is to provide the
 22  clinical service along with the platelet
          
      Page 170
  1  product of that form where it would heighten
  2  their awareness that if they observe signs and
  3  symptoms consistent with sepsis, then it would
  4  heighten their awareness to link the platelet
  5  transfusion to the signs and symptoms of
  6  sepsis.
  7   DR. KLEIN: I'm suggesting
  8  surveying all of the transfusions for evidence
  9  of clinical reaction.
 10   MR. HADDAD: In fact, that is what
 11  our proposal is and we believe that all the
 12  transfused product should be surveyed for any
 13  potential reaction.
 14   DR. SIEGAL: Dr. Cryer.
 15   DR. CRYER: I agree with that and
 16  I also have some real trouble with the idea of
 17  a clinical study and a clinical endpoint. It
 18  is analogous really to trying to figure out
 19  whether the age of red blood cells is
 20  attributable to increased infection or
 21  increased infection is attributable to age of
 22  red blood cells.  
          
      Page 171
  1   It requires a tremendous amount of
  2  work to collect all the data to make the
  3  comparisons in a group of patients that
  4  already have a lot of infectious complications
  5  and a lot of risk factors that lead to them
  6  like the trauma patients, the immunosuppressed
  7  patients, and so forth.  
  8   I don't know how you would -- I
  9  don't know how in a surveillance program that
 10  you would be able to say that the platelet led
 11  to the infection or didn't, especially without
 12  doing what you said which is a control group
 13  of some kind with all the platelets that
 14  weren't contaminated.
 15   DR. SKIKNE: I have a question --
 16   DR. SIEGAL: Could you introduce
 17  yourself, please?
 18   DR. SKIKNE: I'm Barry Skikne,
 19  clinical hematologist, University of Kansas
 20  Medical Center. Platelets live 10 days
 21  normally. I think that maybe if you are
 22  giving seven-day-old platelets you have
          
      Page 172
  1  collected platelets that have a medium
  2  survival of about four-and-a-half to five days
  3  at the time of collection. That is your
  4  average survival. When you are giving seven-
  5  day-old platelets what is the value of that?
  6  You are not going to have a lot of platelets
  7  hanging around for too long before they
  8  cleared. That is one comment. Is that the
  9  right thing to do? Then the older platelets
 10  their function is falling off by this time as
 11  well. They are not as good as the younger
 12  platelets. Why even propose using platelets
 13  beyond five days? That's the question I have
 14  for you.
 15   Then the other thing from the
 16  clinical standpoint and what has been brought
 17  up here is that you may have early
 18  contamination and if you give platelets on day
 19  two and there is the contamination and you not
 20  detecting it clearly by whatever devices or
 21  surveillance mechanisms you have, you have to
 22  observe that patient for up to 10 to 14 days
          
      Page 173
  1  before you can be sure that you have not had
  2  a side effect or infection from that platelet
  3  that has been transfused.  
  4   I think that has been brought up
  5  and I think that should be strongly considered
  6  in any study or any surveillance that has been
  7  done. I think a 24-hour report back is not
  8  adequate.
  9   DR. SIEGAL: Thank you. Is there
 10  any response from anybody to that?
 11   DR. KATZ: Well, the issue of is a
 12  seven-day platelet as good as a five-day
 13  platelet as good as a three-day platelet has
 14  been argued repeatedly. At the end of the day
 15  I think that seven-day platelets circulate and
 16  are recoverable and plug holes, so I tend to
 17  agree. All things being equal, I would rather
 18  get a younger than an older platelet but it's
 19  a matter of maintaining robust inventories for
 20  clinicians and hospitals that drags us out to
 21  five and seven days. I'm not going to argue
 22  that, just that it's an issue to some degree
          
      Page 174
  1  of availability.
  2   With regards to do some reactions
  3  occur later rather than earlier, that's true.
  4  If we have a database of the seven-day
  5  platelets that are distributed, a large
  6  number, most of them occur early and we should
  7  be able from a substantial database to answer
  8  the six and seven versus the day-five
  9  question. We have a statistician on the board
 10  that can talk to us a little more about large
 11  numbers. It is not perfect. What we propose
 12  is not perfect and we understand that but we
 13  think that it provides an appropriate safety
 14  margin, the ability to stop if the safety
 15  margin is inappropriate, and to enhance our
 16  ability to supply what it is we supply.
 17   DR. KLEIN: Just so the members of
 18  the Committee who are not involved in platelet
 19  research understand that the platelet lives
 20  nine days in vivo. However, it can be stored
 21  far longer since it is killed off in vivo
 22  doing what it's supposed to do so you can
          
      Page 175
  1  store platelets eight to 10 days, put them
  2  back into a human being and find out that they
  3  have pretty good recovery and survival as well
  4  as function. There is good reason to believe
  5  that these are function and important cells.
  6   I would like to ask if I may,
  7  since this Committee has asked among the
  8  questions, to talk about anaerobic culture,
  9  whether either Dr. Montag-Lessing or Dr.
 10  Brecher who have enormous expertise in various
 11  microorganisms, can try to sort out the
 12  importance of the anaerobic culture versus the
 13  volume issue which we've heard referred to
 14  either on the initial culture, which is 24
 15  hours, or on the proposed second culture at
 16  five days. Is an anaerobic culture really
 17  that important or is it the volume?
 18   DR. MONTAG-LESSING: Considering
 19  the specter of species in general, point zero,
 20  any species can enter a track including blood
 21  components. That is one of our truths. The
 22  general microbiologists are considering 105
          
      Page 176
  1  different species. Okay, I'm coming back to
  2  the specter of what was published in the last
  3  10 to 15 years.
  4   There are only a few number of
  5  species which are obligatory anaerobes. For
  6  instance, propioni, classical candidate,
  7  propioni vopion agonist. Then colstridium
  8  perfringens, it's pronounced in German. I
  9  don't know exactly how in English. As Mark
 10  Brecher mentioned, listeria monocytogenes, for
 11  instance.  
 12   All the other species are mainly
 13  belonging to the so-called facultative
 14  anaerobes. That holds true for
 15  staphylococcus, streptococcus, enterobacter,
 16  E. coli, klebsiella and enterobacter and
 17  whatever you can imagine.  
 18   That means they are able to grow
 19  in both partals. That means using anaerobic
 20  and aerobic bottle in parallel actually, at
 21  least statistically, increases the likelihood
 22  to overcome the sampling error. That is one
          
      Page 177
  1  important point.
  2   Strictly aerobic species are only
  3  the so-called non-fermenters, classically
  4  pseudomonas aeruginosa and acetobacter and
  5  whatever the names are. What is further is
  6  strictly aerobic non-fermenters there is one
  7  other species I forgot currently. That means
  8  the main outcome of using two bottles is a
  9  statistical point.
 10   DR. BENJAMIN: Richard Benjamin,
 11  American Red Cross. Just to address the
 12  anaerobic bottle issue again. Dr. Brecher has
 13  pointed out at least three advantages of the
 14  anaerobic bottle. One being for strict
 15  anaerobes and the bugs we hear about are P.
 16  acnes, clostridium, and new bacterium as
 17  having caused fatal reactions.
 18   To point out that the BacT/ALERT
 19  test has never been validated to pick up new
 20  bacterium and in all published work has never
 21  picked it up in a platelet product. To say
 22  that it can detect new bacterium is not true.
          
      Page 178
  1   For clostridium species,
  2  clostridium is a facultative aerobe. There
  3  are no reports of clinical -- clostridium
  4  perfringens is a facultative aerobe. There
  5  are no reports in literature in BacT/ALERT
  6  picking of clostridium perfringens in a
  7  platelet product in an anaerobic bottle.  
  8   There is at least one report of it
  9  having been picked up in an aerobic bottle.
 10  We are coming down essentially to P. acnes as
 11  the only strict anaerobe that we pick up in
 12  platelet products that may or may not have
 13  clinical relevance.
 14   DR. BRECHER: Mark Brecher. A
 15  couple points. There have been clostridium
 16  perfringens that have caused sepsis, at least
 17  one case from England that only grew in an
 18  anaerobic media but they were not specific as
 19  to how they did it. They didn't say in the
 20  paper as to how they set up their culture.
 21   The other real advantage, I think,
 22  is that the media is different between the
          
      Page 179
  1  aerobic bottle and the anaerobic bottle and
  2  some bacteria will grow better in one set of
  3  media than the other. Clearly volume is a big
  4  issue but I personally think that if you
  5  decided how much volume you could take from a
  6  bag, I would rather split it over two
  7  different media to try to get every advantage
  8  in growth.
  9   DR. DI BISCEGLIE: I just wanted
 10  to pick up some of the earlier discussion.
 11  Essentially it's about endpoints, if I could.
 12  We have this clinical endpoint of septic
 13  transfusion reaction. My question is how well
 14  validated might be this one page case report
 15  that the FDA is proposing? And is it
 16  validated in the absence of a backup bacterial
 17  culture?  
 18   I can imagine these patients are
 19  very sick. Some of them might have their
 20  reaction in the operating room. Some of them
 21  may have their reaction, I'm hearing, 10 to 14
 22  days afterwards. Don't we need the bacterial
          
      Page 180
  1  backup as part of the definition?
  2   Just a couple of other points on
  3  this if I might. The idea of an adjudication
  4  committee I think sounds very appropriate. I
  5  think it would be different than a data set to
  6  a monitoring committee. It should probably
  7  come from a group if investigators themselves
  8  who are expert in this to adjudicate STRs.
  9   Then the definition of active
 10  versus passive reporting was not quite clear
 11  to me. I heard some definitions but I wasn't
 12  sure. I need some clarification on exactly
 13  what the difference is between active and
 14  passive.
 15   DR. KATZ: I think what we
 16  envisioned in PASSPORT II or Son of PASSPORT
 17  is that if you want seven-day platelets you
 18  have to send back the form. You have to agree
 19  to do that. It's really active surveillance
 20  and we would use data that is already, should
 21  be, gathered at the bedside during
 22  transfusion.
          
      Page 181
  1   DR. GLYNN: I think it's certainly
  2  key to discuss the design of a study. I also
  3  would like to get a little bit more
  4  information about what you're going to do with
  5  the data and once you get the data in terms of
  6  stopping guidelines for the proposed study.
  7  Has there been some discussion?
  8   DR. KATZ: There's been
  9  discussion. We haven't settled on anything.
 10  I think what we wanted before we went forward
 11  was a sense that substituting this kind of
 12  clinical surveillance for the kind of
 13  microbiologic surveillance that the FDA has
 14  initially proposed here there was enough
 15  acceptance of that that we could go forward
 16  and design the study to everybody's
 17  satisfaction.
 18   DR. KUEHNERT: I just wondered if
 19  we could get a little more detail on the
 20  active surveillance as far as what sort of
 21  information is going to be collected and in
 22  what time course in relationship to the
          
      Page 182
  1  transfusion?
  2   DR. KATZ: Well, I showed you a
  3  mock-up and it was only a mock-up but it
  4  actually used elements from your study from
  5  BAYCON. That is an example of how we would
  6  like to do it. Certain elements are obvious
  7  in terms of fever and hypothermia and
  8  hypotension and those sorts of things.  
  9   The question is how far do you
 10  want to go. We thought to look at BAYCON and
 11  what Ros Yomtovian has done in her ongoing
 12  study and picked the best elements of those
 13  studies would be the best way to go.
 14   DR. KUEHNERT: So I think
 15  originally it seemed like the emphasis was on
 16  getting vital signs during the transfusion.
 17  Now you have moved to what the BAYCON
 18  definition was which was four hours after. As
 19  I shared with you on looking at the BAYCON
 20  study data which is confirmed cases. This is
 21  really where we were sure that there was
 22  sepsis. About 40 percent of the cases the
          
      Page 183
  1  symptoms only arose after the transfusion and
  2  most of them after 90 minutes and some of them
  3  up to three hours. That four hours it would
  4  have picked up all the BAYCON cases but that
  5  is sort of circular reasoning because our
  6  definition was four yours after.
  7   DR. KATZ: Again, my inquiry to
  8  you about the timing of symptoms in the BAYCON
  9  study was so we could begin to settle on if we
 10  are going to be able to get large numbers how
 11  far out do we have to carry it because that
 12  has to do with operational realities in the
 13  hospital. I don't think we've carved in stone
 14  four hours but I think your data says to us
 15  that turning off the form when they take down
 16  the bag probably is not the optimal way to do
 17  it and four hours kind of comes up because of
 18  your numbers.
 19   DR. KUEHNERT: I would agree with
 20  that. I would also just say that I think
 21  clinical active surveillance is not an exact
 22  science. No matter how active you make it,
          
      Page 184
  1  it's still going to be very, very difficult to
  2  be able to get definitive answers. I think
  3  having a control group would be the Cadillac
  4  way to do it. Whether that is really feasible
  5  is another issue. Maybe what should be
  6  considered is a combination of microbiologic
  7  and clinical endpoints. It's very important
  8  that those endpoints be rigorously defined
  9  because my concern in looking at what stopped
 10  PASSPORT is whether they were really things
 11  that would be defined as a safety issue
 12  because I'm looking at these three cases and
 13  one of them definitely there was a problem.
 14  The other two I'm just sort of wondering what
 15  the real facts were in relationship to the
 16  unit. I think there would have to be a lot of
 17  work put into confirming cases to make sure
 18  that those cases really did represent a
 19  failure of the system.
 20   MR. HADDAD: Yes, and that is what
 21  is being called for by the follow-up on the
 22  septic reaction in the sense of actually
          
      Page 185
  1  retesting any remnants in the blood bag and
  2  culturing the patient and trying to make the
  3  connection.
  4   DR. KUEHNERT: So just one more
  5  point with the clinical endpoints. I mean,
  6  the longer you go out after the transfusion
  7  the more chance that you are going to have a
  8  patient that is going to get sepsis from some
  9  other cause. The longer you go out the more
 10  the need for a control group to make sure that
 11  you're not picking up some line infection or
 12  some other cause.
 13   MR. HADDAD: But probably in that
 14  case the study will go beyond what we are
 15  contemplating right now. It's going to be a
 16  much bigger study. If I may just one comment
 17  on the difference between active and passive.
 18  Passive reporting is what currently happens
 19  right now which is the report -- the
 20  surveillance is initiated by a report of a
 21  transfusion reaction which then goes back to
 22  the transfusion service and then back to the
          
      Page 186
  1  blood center. Our definition of active
  2  surveillance is that along with the
  3  transfusion product that form will go out and
  4  the transfusion service and the clinical
  5  service will have to fill out that report and
  6  turn it back within 24 hours or whatever
  7  period we'll decide. So the clinical service
  8  will have their awareness heightened as to the
  9  occurrence of septic transfusion reactions and
 10  then if they observe such symptom, they can
 11  link them back to the product by using that
 12  form.
 13   DR. SIEGAL: Dr. Bracey.
 14   DR. BRACEY: Question for Dr.
 15  Katz. Given the potential downsides of seven-
 16  day platelets in terms of potential risks to
 17  recipients, what's your sense of -- and
 18  challenges, operational challenges, what's
 19  your sense of the willingness of your clients
 20  to participate? When they are signed on, if
 21  you will, is this done -- how is that exactly
 22  done? Is it done through the transfusion
          
      Page 187
  1  committee? How does the end user sort of
  2  really know what is going on?
  3   DR. KATZ: Well, it seems to me
  4  that the best approach is to distribute with
  5  the seven-day platelet a form that gathers in
  6  a very simple fashion the information that we
  7  are interested in. If you want seven-day
  8  platelets, you have to play the game. There
  9  is already information that should be gathered
 10  at the bedside so from the nursing standpoint
 11  should not be undoable.  
 12   That is different than saying we
 13  have critical mass of hospitals that have
 14  agreed to do it. It seems to me a very simple
 15  approach. I think FDA and the PASSPORT group
 16  are probably pretty much on the same page here
 17  about the distribution of an instrument that
 18  allows us to get what we need with the least
 19  possible pain.  
 20   Then somebody brought up the point
 21  we shouldn't call it DSMV. That is absolutely
 22  correct and I apologize for being obtuse.
          
      Page 188
  1  Then we get back those forms including I don't
  2  see why we can't do a control group. I don't
  3  think we have discussed that explicitly what
  4  percentage of the seven-day platelets going to
  5  these hospitals would be accompanied by that
  6  form.  
  7   Does it have to be all of them?
  8  Let's talk about that. Then we've got our
  9  control group of those who do not react, or do
 10  not appear to react in whatever the time frame
 11  is we've chosen.
 12   DR. SIEGAL: I'd like to ask a
 13  maybe more fundamental question of both the
 14  FDA and industry. It seems to me that the
 15  standard of care for transfusion of platelets
 16  is not to go beyond five days. The rationale
 17  for that is prior experience and the history
 18  of that has been well outlined here.  
 19   Now we are proposing to do, in
 20  effect, a very large study, though perhaps not
 21  large enough, in which the end user isn't a
 22  blood bank, it's the patient. Essentially
          
      Page 189
  1  what we're proposing is a no-consent clinical
  2  trial. I wondered if there shouldn't be some
  3  discussion of that at this point or a little
  4  bit later.
  5   DR. EPSTEIN: Let me comment
  6  because FDA certainly was not unaware of that
  7  concern and spent a fair amount of time
  8  addressing it before the first PASSPORT study.
  9  The way we have looked at it is that we
 10  approve collection containers for seven-day
 11  storage of platelets. They were approved
 12  based on the functionality of the platelets in
 13  vivo.
 14   There is no standard for control
 15  of bacterial contamination of platelets.
 16  There is the standards recommended through
 17  guidance and so forth that just have to do
 18  with a septic procedure of collection, you
 19  know, closed systems, for example, skin
 20  disinfection.  
 21   There was no microbiological
 22  standard for the platelet. The way the FDA
          
      Page 190
  1  was looking at it was that it was a post-
  2  marketing study of a licensed product which
  3  was designed to enhance safety while learning
  4  whether our hypothesis was correct. We didn't
  5  see it as releasing an experimental product
  6  because we had to, in fact, approve the seven-
  7  day storage in the absence of a bacterial
  8  control standard.
  9   Now, I have to say that underlying
 10  all this was the presumption, perhaps naive,
 11  that they would in Phase IV evaluation prove
 12  no less safe than conventional day-five
 13  platelet. In actuality the data have not
 14  ruled it out. It's just that the likelihood
 15  of a successful study became very, very low,
 16  less than 10 percent. Have I confused you?
 17   DR. SIEGAL: Should we have a
 18  consent form?   
 19   DR. EPSTEIN: Pardon?
 20   DR. SIEGAL: Should we have a
 21  consent form for people who are going to get
 22  six and seven-day platelets?
          
      Page 191
  1   DR. EPSTEIN: Well, I mean, we
  2  think that we are in a comparable ballpark of
  3  safety. We just don't know for sure. What we
  4  are trying to do, if you will, is a belt and
  5  suspenders because remember the baseline here
  6  was an uncultured five-day platelet. In 2005
  7  we didn't yet have a predominance of culture.
  8  Mind you we are just talking about apheresis
  9  platelets here. You still have the entire
 10  world of whole blood derived platelets which
 11  are, for the most part, not being cultured.
 12  The question is really what is the baseline
 13  standard.
 14   DR. SIEGAL: Comment in the back.
 15   DR. KLEINMAN: Steve Kleinman from
 16  AABB. I think one point that is being missed
 17  here maybe by the committee in the
 18  presentations is that if we have safety
 19  concerns about six and seven-day platelets, we
 20  should have the same safety concerns about
 21  five-day platelets which are allowable as
 22  transfusable products.
          
      Page 192
  1   Most of the septic transfusion --
  2  there are no data that say seven-day platelets
  3  are less safe than five-day platelets. In
  4  fact, the presentation of the Irish
  5  transfusion study, I think, was a little
  6  misstated in that the outdated -- the culture
  7  at outdate was either from buffy coat
  8  platelets that outdated at five days or from
  9  apheresis platelets that outdated at seven
 10  days.
 11   I've talked with Dr. Murphy who
 12  published that paper and, in fact, no
 13  apheresis platelets outdated in seven days
 14  because they only allowed them to go to seven
 15  days if they knew they had a need. They
 16  recultured them on day four and they
 17  transfused them.  
 18   In fact, most of that data is day-
 19  five data. You clearly see that the rates --
 20  we didn't have confidence intervals for the
 21  rates but the rates at day five are much
 22  higher than the rates at day four, culture
          
      Page 193
  1  positivity rates, and the septic transfusion
  2  reaction rates are higher at day five than at
  3  day four.  
  4   In my opinion, day five -- getting
  5  a day-five platelet is less advantageous for
  6  a patient now than getting a day-four or day-
  7  three platelet but nobody is saying we should
  8  back off and make platelet storage four days
  9  because we have the availability issue and we
 10  have to balance safety against availability.
 11   I think that extending for seven
 12  days is the same issue and I agree with Jay
 13  that it's not an experimental product in that
 14  sense. It's just incremental. Maybe it's a
 15  little less safe. We don't know that. In the
 16  universe of platelet products it's definitely
 17  not the least safe product that you can get.
 18   You can get a whole-blood derived
 19  platelet at day five that was tested with a
 20  dip stick and I think that would be less safe
 21  and, yet, we don't get informed consent other
 22  than beyond the usual one for transfusion. I
          
      Page 194
  1  think, you know, it really is a balance
  2  between safety and availability and the only
  3  reason to go to day seven, in my opinion, is
  4  to increase availability and we have to
  5  determine whether that's a real safety risk to
  6  patients.  
  7   Therefore, obviously we need some
  8  kind of clinical study and how we are going to
  9  resurrect this study is really the question if
 10  we can. I guess my personal opinion is if it
 11  requires culturing on day five, I agree with
 12  Dr. Katz and Mark Brecher that it just isn't
 13  going to happen. Maybe that's okay. Maybe we
 14  shouldn't go to day seven but the only way to
 15  get to day seven is through clinical
 16  endpoints. I mean, that's my perspective for
 17  the Committee.
 18   DR. SIEGAL: Dr. Fleming, were you
 19  going to comment on statistical power?
 20   DR. FLEMING: Many comments if I
 21  just stick to this issue first. It seems to
 22  me that what we are seeing here is definite
          
      Page 195
  1  need for scientific evidence coming from what
  2  I would call a clinical study. In that
  3  context it seems to me that this is not just
  4  clinical care and the consent process seems
  5  appropriate.  
  6   May, it may be entirely
  7  appropriate that it is a community consent
  8  rather than an individual specific consent.
  9  But to argue that this is a nonissue because
 10  this is just clinical care I don't understand
 11  that argument because what we are seeing today
 12  is clear need for scientific evidence to
 13  determine what is the benefit to risk aspects
 14  of having the six to seven-day availability.
 15  I had other comments but I'll stop at this if
 16  there is more discussion about the consent
 17  issue.
 18   DR. SIEGAL: Anybody else?
 19   DR. ZIMRIN: Everyone who gets a
 20  transfusion actually does get consent and is
 21  informed of the risk of platelet transfusion.
 22  At least that's the policy in my hospital. I
          
      Page 196
  1  assume it's a global policy.  
  2   It's not that patients aren't
  3  informed that there's a risk. Sitting down
  4  with patients and giving them a lecture about
  5  the relative merits of the whole blood
  6  apheresis six or seven-day is probably not
  7  going to be beneficial.
  8   DR. KLEIN: There's another
  9  related issue here that I don't know that we
 10  have the data to address it but perhaps Dr.
 11  Katz or someone in the audience can and that
 12  is that we heard about platelet outdating.
 13  That is largely an economic issue. Right?
 14   We also heard about unfilled
 15  orders and that is a surrogate measure for
 16  patients who can't get platelets when their
 17  physician deems that they should. I guess the
 18  real question is how many patients are out
 19  there who get nothing because you don't have
 20  seven-day platelets. Certainly we don't
 21  inform them that your platelet count is 5,000
 22  but you're not getting anything today because
          
      Page 197
  1  we just don't have it.
  2   DR. FLEMING: That's clinical
  3  care. Certainly there is no specific
  4  scientific consent in clinical care. If you
  5  are doing a study that incorporates clinical
  6  care but it's a scientific research study
  7  perspectively addressing issues of benefit to
  8  risk, then I don't understand how that doesn't
  9  require consideration of the consent process.
 10   Now, that could entirely be
 11  adequately addressed by a community consent
 12  rather than a specific individual consent. As
 13  soon as we acknowledge that there is a
 14  critical scientific issue to be addressed, and
 15  I think we are talking about -- both FDA and
 16  advocates of PASSPORT are saying we are going
 17  to have a revived PASSPORT trial.  
 18   We're just discussing how it's
 19  going to be done to address this important
 20  scientific study, this key scientific issue.
 21  It seems to me that it's appropriate this be
 22  a community consent but that issue --
          
      Page 198
  1   Fred, I believe you're right to
  2  raise that issue to indicate that
  3  consideration should be given is this a
  4  setting where our community consent is
  5  acceptable in which case you don't need
  6  individual consent.
  7   DR. SIEGAL: Henry?
  8   DR. CRYER: I don't think this is
  9  a community consent issue here. Just take Dr.
 10  Klein's example. I'm a patient and I've got
 11  a platelet count of five and my hospital
 12  doesn't have any five-day platelets. If they
 13  fill out a form and play the game, they can
 14  get some seven-day platelets and then so could
 15  I.  
 16   So, in my view, you had better ask
 17  me. You had better tell me, "Look, your
 18  platelets are five. We don't have what we
 19  would normally give you but we have some
 20  seven-day.  
 21   We can either--normally we
 22  wouldn't give you anything and we just play it
          
      Page 199
  1  out. We're not sure whether these seven-days
  2  are okay or not." You better let me
  3  individually consent on that issue. Now, if
  4  I'm a trauma patient--not all patients get
  5  consent.  
  6   Trauma patient, 20 units of blood,
  7  middle of an operation, bleeding, they are
  8  going to get platelets without any consent.
  9  Again, I think we run into some ethical issues
 10  that I don't see how you can do a community
 11  consent on that.
 12   DR. ZIMRIN: But you would be okay
 13  giving them whole blood platelets that you
 14  think are very likely to be much worse without
 15  a consent? Let's say you don't want to be
 16  bothered, you're busy, you're in clinic,
 17  you're in the OR and there is no one available
 18  so why don't we just give them the uncultured
 19  platelets rather than the seven-day platelets?
 20  I understand actually on the IRB.  
 21   I understand these issues and I
 22  understand everyone's discomfort with this but
          
      Page 200
  1  I think it needs to be put into context sort
  2  of where the risk falls in terms of general
  3  clinical care.
  4   DR. RENTAS: You know, our most
  5  concern should be with the safety of the blood
  6  supply and I think that's what we're here for.
  7  As painful as I know it will be, what Dr. Katz
  8  and Dr. Brecher have talked about, doing the
  9  five-day culture here, when you look at
 10  PASSPORT and now we are trying to see if we
 11  can get into the Son of PASSPORT, but we never
 12  want to go to a Grandson of PASSPORT. We
 13  don't want to do this all over again. Let's
 14  look at the proposal that the industry have
 15  put together here.  
 16   To me the biggest change will be
 17  that we will be doubling the volume of the
 18  inoculum here. But other than that is that
 19  going to be enough when you have a day six or
 20  a day-seven platelet out there to ensure that
 21  product is still sterile out there so that is
 22  the question I will ask.
          
++