ATDEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR BIOLOGICS EVALUATION AND RESEARCH

This transcript has not been edited or corrected, but appears as
received from the commercial transcribing service.  Accordingly the food
and Drug Administration makes no representation as to its accuracy.

BLOOD PRODUCTS ADVISORY COMMITTEE MEETING

Thursday, December 12, 2002

8:10 a.m.

Holiday Inn Gaithersburg

Two Montgomery Village Avenue

Gaithersburg, Maryland

PARTICIPANTS

Kenrad E. Nelson, M.D., Chairman

Linda A. Smallwood, Ph.D., Executive Secretary

Members

	James R. Allen, M.D.

	Mary E. Chamberland, M.D.

	Kenneth Davis, Jr., M.D.

	Donna M. DiMichele, M.D.

	Samuel H. Doppelt, M.D.

	Michael G. Fitzpatrick, Ph.D.

	Harvey G. Klein, M.D.

	Judy F. Lew, M.D.

	Paul H. Schmidt, M.D.

	Sherri O. Stuver, Sc.D.

	Lori A. Styles, M.D.

Consumer Representative

	Robert J. Fallat, M.D.

Temporary Voting Member

	Liana Harvath, Ph.D.

Nonvoting Industry Representative (Acting)

	Toby L. Simon, M.D.

C O N T E N T S

PAGE

Welcome, Statement of Conflict of Interest,

Announcements                         	                  5

Committee Updates

     Summary of Workshop on West Nile Virus

     November 4-5, 2002:

	Hira Nakhasi, Ph.D. 	                          9

	Kay Gregory, AABB, ABC, ARC 	33

     Medical Device User Fees (MDUFMA):

	Mary Elizabeth Jacobs, Ph.D.	                 41

     Approval of OraQuick Rapid HIV-1 Antibody

     Test:

	Elliot Cowan, Ph.D.  	                        58

I. Bacterial Contamination

     A. Background and Introduction:

	Alan Williams, Ph.D.    	                     68

     B. Skin Preparation of Phlebotomy:

	John Lee, M.D.             	                  81

     C. Update on the Diversion Pouch:

	Jaro Vostal, M.D., Ph.D.      	              125

     D. Quality Control Approaches for Detection

        of Bacterial Contamination:

	Alan Williams, Ph.D.             	           131

     E. Experience with Plastic Tubing

        and Universal Bacterial Culturing:

	James Aubuchon, M.D.                	        150

     F. Data Presentation:

	Tracy Manlove                          	     178

     G. Design of Clinical Trials for Clearance

        of Devices Intended for Screening of

        Platelet Products Prior to Transfusion:

        Background:

	Steven Wagner, Ph.D.                      	  202

     H. Proposed Study Design:

	Jaro Vostal, M.D., Ph.D.                   	 217

     I. Open Public Hearing

	Kay Gregory, AABB	                           234

	Roger Dodd, M.D., ARC  	                     241

	Celso Bianco, M.D., ABC 	                    244

	Cynthia T. Crosby, Medi-Flex 	               254

     J. Questions for the Committee       	            259

C O N T E N T S(Continued)

PAGE

II. Human Parvovirus B19 NAT Testing

for Whole Blood and Source Plasma

     A. Introduction and Background:

	Mei-ying W. Yu, Ph.D. 	 299

     B. Overview of Parvovirus B19 Infection:

	Kevin Brown, M.D.    	                       308

     C. Industry Data Presentations

	American Red Cross, Susan Stramer, Ph.D.	 353

	Fractionators/PTTA

	  Barbee Whitaker, Ph.D. 	                   308

	  Steve Petteway, Ph.D.     	                357

	  Edward Gomperts, M.D.        	             367

     D. Open Public Hearing

	Kay Gregory, AABB, ABC, ARC       	          383

     E. FDA Perspectives and Questions for

        the Committee:

	Mei-ying W. Yu, Ph.D.           	            384

     F. Committee Discussion                 	         389

P R O C E E D I N G S

Welcome, Statement of Conflict of Interest,

Announcements

	DR. SMALLWOOD:  Good morning.  Welcome to the 75th meeting of the Blood
Products Advisory Committee, the longest running series in the FDA
history.

	I am Linda Smallwood, the Executive Secretary.  At this time, I will
read for you the Conflict of Interest Statement that applies to this
meeting.

	This announcement is part of the public record for the Blood Products
Advisory Committee meeting on December 12th, 2002.

	Pursuant to the authority granted under the Committee Charter, the
Director of FDA's Center for Biologics Evaluation and Research has
appointed Dr. Liana Harvath as a temporary voting member.

	Based on the agenda, it has been determined that there are no products
being approved at this meeting.  The committee participants have been
screened for their financial interests.  To determine if any conflicts
of interest existed, the agency reviewed the agenda and all relevant
financial interests reported by the meeting participants.

	The Food and Drug Administration has prepared general matter waivers
for the special government employees participating in this meeting who
required a waiver under Code 18, Section 208.

	Because general topics impact on so many entities, it is not prudent to
recite all potential conflicts of interest as they apply to each member.
 FDA acknowledges that there may be potential conflicts of interest, but
because of the general nature of the discussions before the committee,
these potential conflicts are mitigated.

	We would like to note for the record that Dr. Toby Simon is
participating in this meeting as the Acting Non-Voting Industry
Representative acting on behalf of regulated industry.

	With regard to FDA's invited guests, the agency has determined that the
services of these guests are essential.  There are interests that are
being made public to allow meeting participants to objectively evaluate
any presentation and/or comments made by the guests.

	For the discussions on bacterial contamination, Dr. James Aubuchon has
reported that he is a researcher on bacterial contamination.  He has
spoken on behalf of Pall Corporation and he is a member of the Medical
Advisory Board for Verax.

	Dr. Stephen Wagner is the Director of Cell Therapy, American Red Cross,
Holland Laboratory.  He also received a research grant from Organon
Technika for the detection of bacteria in platelets.

	In addition, there are speakers making industry presentations and
speakers giving committee updates from regulated industry and other
outside organizations.  These speakers have financial interests
associated with their employer and with other regulated firms.  They
were not screened for these conflicts of interest.

	FDA participants are aware of the need to exclude themselves from the
discussions involving specific products or firms for which they have not
been screened for conflicts of interest.  Their exclusion will be noted
for the public record.

	With respect to all other meeting participants, we ask, in the interest
of fairness, that you state your name, affiliation, and address any
current or previous financial involvement with any firm whose products
you wish to comment upon.

	Waivers are available by written request under the Freedom of
Information Act.

	At this time, I would like to ask are there any declarations that need
to be made before we proceed with this meeting.

	Hearing none, I would just make a brief announcement that outside you
should have found a sheet that listed the tentative dates of the Blood
Products Advisory Committee meetings for the year 2003.

	I hope that you will make a note on your calendar to hold these dates,
but we will advise you when we have confirmed them.

	At this time, I would like to introduce to you the members of the Blood
Products Advisory Committee.  As I call the names of the members, would
you please raise your hand.

	Dr. Kenrad Nelson, Chairman.  Dr. Lori Styles.  Dr. Paul Schmidt.  Dr.
Harvey Klein.  Dr. Liana Harvath.  Dr. James Allen.  Dr. Sherri Stuver. 
Dr. Robert Fallat.  Dr. Toby Simon.  Dr. Donna DiMichele.  Dr. Mary
Chamberland. Dr. Samuel Doppelt.  Dr. Fitzpatrick.  Dr. Judy Lew.

	As you may have noticed, we have a very full agenda today, very full. 
We will try to keep on time and we will ask everyone's cooperation in
doing so.

	At this time, I would like to turn the proceedings of the meeting over
to the Chairman, Dr. Kenrad Nelson.

	DR. NELSON:  Thank you, Dr. Smallwood.

	The first item on the agenda are some committee updates.

	First, Dr. Hira Nakhasi is going to summarize a Workshop on West Nile
Virus that was held in November.

Committee Updates

Summary of Workshop on West Nile Virus

November 4-5, 2002

Hira Nakhasi, Ph.D.

	DR. NAKHASI:  Good morning.  Thank you, Dr. Nelson.  Since Linda said
there is a full schedule today, it will be 6 o'clock is the regular
time, I don't know how long we will be here, but I will try not to
contribute to the delay and go right away into giving my update.

	[Slide.]

	This update is on the workshop which we held on November 4th and 5th,
and many of you attended that workshop, and this was on the Development
of Donor Screening Assays for West Nile Virus.

	This workshop was in response to the recent epidemic in the epidemic
2002, and we wanted to see how we could understand what the epidemic is
and how we can get the methodologies in testing soon developed and to
screen the blood for the West Nile Virus.

	[Slide.]

	The goals of this workshop were as such, as pointed out here, we wanted
to know what is going on with the current status on the West Nile
pathogenesis and epidemiology in the U.S., and wanted to know what are
the methodologies suitable for blood and tissue donor screening, and
wanted to know from the industry perspective are they ready for testing
in a large-scale screening mode.

	Also, we wanted to hear from the manufacturers about the inactivation
process in the blood products.  We also wanted to hear from the proposed
studies on prevalence and donors, and how this test would be licensed
and FDA's expectation from that, and issues relevant to the
implementation of the West Nile Virus.

	It was a two-day full agenda, very interesting. There were a lot of
discussions, but before I go to what we achieved from that meeting, I
just want to give you a little bit of background for the people who may
not know about West Nile Virus.

	[Slide.]

	The West Nile is a mosquito-borne flavivirus.  It has a positive strand
RNA and primarily infects birds, but horses and humans are incidental
hosts.

	About 80 percent of the infected persons remain asymptomatic, and the
rest, 20 percent, develop mild febrile illness, flu-like symptoms.  In
that, approximately 1 in 150 infected people develop meningitis or
encephalitis.

	The viremic period can occur up to two weeks, but it is sometimes a
very short period, but can also last for almost a month.

	[Slide.]

	Blood transmission of West Nile has been confirmed in the recent U.S.
outbreak, and I will go into a little later about the cases.  However,
the magnitude of the risk of West Nile from transfusion is unknown at
this time.

	Again, the problem with this virus is that it is a very low titer virus
compared to other viruses like HIV and HCV, 103 copies/ml.  It is, as I
said, viremia is transient, however, in some of these encephalitis
patients, the viremia can be as high as 106 copies/ml.

	The viremia resolves rapidly after seroconversion to IgM, and IgM can
persist as long as one year.  West Nile infection does not become
chronic.

	[Slide.]

	The current status of West Nile as of last week, what we saw from the
CDC/MMRW report, in 2002, the total number of West Nile cases reported
was 3,775, of which 216 deaths have occurred.

	The whole of the U.S. is practically endemic except in a few states in
the West even though one case was found in L.A., but the majority of the
United States is endemic.

	[Slide.]

	Viremia begins one to five days before the onset of symptoms and can
last an average of six days.  As I said earlier, you can go up to two
weeks or 14 days.

	The estimated risk at this time, Lyle Peterson from CDC had published a
paper this year of one and two infections per 10,000 donations
nationwide, however, in highly endemic regions where the activity is
very high, 16 at the peak of the epidemic, was 16 per 10,000 with a mean
of 68, because as I will say here, it can go from late August to late
September, and that is the range there.

	So far, 47 possible transfusion-transmitted cases have been reported. 
Out of that, 13 have been confirmed, 14 were not transfusion related,
the rest are under investigation still going on at CDC.

	[Slide.]

	Then, people presented, researchers presented data on the methodologies
which are suitable for blood and tissue donor screening.

	Both serological and nucleic acid based tests were discussed. 
Basically, the serological or IgM antibody assay, people have used
recombinant antigen, but these are all research assays at this time, so
mind you that they are not being used in a clinical setting, in a trial
setting.

	Some of these serological assays use recombinant antigen, can
cross-react with other cousins of West Nile, like St. Louis
encephalitis, dengue, and Japanese encephalitis, that is what we heard,
however, this test could be used in a high throughput assay, low
specimen volume, and can be multiplex, short turnaround, and can be
adapted to the platforms which are existing already for serological
testing for other components.

	The nucleic acid tests, there are many PCR based, there are standard
PCR, Taqman PCR, and NASBA, but what came out of the meeting, that
Taqman, real-time PCR is the most sensitive at this time and equal to
NASBA.

	It could be used in the high throughput setting and detection limits
are 15 plaque-forming units/ml to 15,000, however, in some of the cases,
we heard also it can go 0.1 plaque-forming units/ml.

	The caveat here is these tests so far, what we have is the human
viremia is around 18 PFU.  It is basically towards the tail end, and the
lower limit of it, but then we recently heard, which I will maybe talk
about down the road,  that CDC has come up with a much more sensitive
test, which is 10-fold sensitive and can, by making such a modification
concentrate, increase the volume of the sample and also making other
changes in extraction.

	But we heard that minipool NAT, detection rate is only 50 percent, and
need to adapt smaller pools.  Sue Stramer and ARC pointed out that even
smaller, eight pools could be better, but maybe it may go to the usual
NAT also.

	[Slide.]

	Again, there were some other issues.  I don't want to go into detail of
these things basically, because what we were told earlier, what we knew,
that the West Nile Virus, once the virus is resolved, the antibody comes
out, the viremia is resolved, but there are cases where RNA can be
detected in the presence of antibody.

	Again, under the caveat is that West Nile IgM can remain positive for
one year longer without any infective, and whether there is infective,
people do not know.

	It looks like NAT could be the preferred choice for testing, however,
IgM assays have also a role to play, the serological assays may have a
role to play in confirmation of NAT results or seroconversion studies.

	Also, it was discussed that if we screen blood, it will have a strong
impact on tissue and organ donation and screening it.

	Again, there was emphasis, which I will talk a little later, that we
may have to have developed tests which are suitable for cadaveric
samples.  Then, there are some activities going on in the panel
development, I will talk a little bit later.

	[Slide.]

	The industry presented their data, which was basically plan and not
much information.  NGI presented some data where they have a NAT test
which has sensitivity of 100 copies/ml with the range of 10 to 200
copies.

	We heard that they had screened a large number of some samples, and the
prevalence rate was 1 in 8,000, and one of the samples was very high
titer donation and could result in pools of 64 and 512.

	We also heard from GenProbe that they have a test development
validating their tests using synthetic RNA, and the detection was 7.6
copies/ml.  They are still working, we may hear maybe they have some
information during the open discussion, that they are working on
selecting the primers where they can use it.  They are still in that
mode.

	Roche presented some data, which is basically the plan, no data, and
then basically the development of tests, everybody agreed that it will
be IND/BLA mechanisms.  The validation of these tests will be at the
beginning of 2003 and IND by the middle of 2003.

	[Slide.]

	We heard about the virus inactivation process strategies and several
manufacturers presented data on using the currently used inactivation
processes like psoralen, riboflavin, Inactine treatment of various blood
components, and they also tested in West Nile, inactivation in that
process, and they could inactivate more than 4 logs.

	Therefore, on that basis, some people felt that they may not need to
demonstrate West Nile Virus specific inactivation, however, other
people, an equal number have held that it will be having showing West
Nile Virus specific inactivation would also add a layer of safety
similar to like HIV and HCV.

	It is known that whenever there is an agent which we can culture and
show that it can be specifically inactivated, it is FDA's understanding
that we should use and show specific virus, specific inactivation.

	However, there are caveats to these inactivation processes, such as
adverse events which will be due to the products have been treated with
such, such as immunological reactivity, increased sensitivity of blood
cells to other drugs, specificity of inactivation between pathogens and
hosts.

	It was agreed upon that studies are needed to assess the risk of this
inactivation process on blood products.

	[Slide.]

	Then, we heard about the proposed studies.  There were several studies
talked about, and there have been changes going on since we heard about
studies.  Now, we have heard that the ARC is conducting a linked study
of a large number of samples, of 85,000, out of which 7,000 are going to
be tested under the CDC, and will be tested by CDC's sensitive method
which I described just a few moments ago.

	Then, those samples will be tested by GenProbe's test, and these
samples are linked.

	Also, there is a research study under RADAR, which is REDS/TRIPS, but
the samples are small.  This is mostly going to be IgM sero problem
studies, and finding out from that, sero problems in their samples,
those seropositive samples will be tested for NAT using several NATs.

	Then, the other study is the Roche samples.  Roche has a large number
of samples again collected through moderate, low, and high epidemic
areas, and we have not heard anything about what is going on with that.

	But the objective of all these studies was to really see the prevalence
of viremia, compare minipool versus individual NAT, confirm viremia by
IgM and RNA testing of donor follow-up samples, and then develop
analytical-sensitive panels, compare West Nile, RNA, and IgM assays, and
also incidence rate of transfusion-transmission of West Nile, and
exposure to recipients by testing autologous donations for IgM
reactivity.

	These prevalence studies, we were told that it will be done in two
phases, Phase I, where the performance of candidate West Nile RNA assays
will be validated against the benchmark, which is the CDC NAT (50 geq/ml
at 50 percent detection limit), which will be 100 geq/ml at 100 percent
detection levels.  We were told that the completion will be in the first
quarter of 2003, that is to perform validation of these tests.

	Then, Phase II is testing the samples by the middle of 2003 under IND.

	[Slide.]

	Here, we at CBER-FDA also have some effort going on with, first of all,
there is several efforts actually, not some, several efforts going on. 
One is the development of reference panels for lot release testing, and
these we are taking the virus from the CDC, culturing that, and then
spiking into the naive blood and then that panel will be distributed
among different groups and tested to see how these tests will perform.

	Then, we are also developing an in-house Taqman PCR and IgM assays to
basically compare with CDC's, because many times we have to do
investigational tests in-house, too, so we want to have the capability
of the testing in-house, too.

	The objective is basically to study viral dynamics, infection dose,
distribution in the blood components, viral tropism, correlation between
viral strains and infectious outcome.

	[Slide.]

	Then,  we discussed about the regulatory pathway for these assay
developments, and a few of these slides are directly stolen from Jay's
presentation to AABB.  The donor screening and supplemental tests will
be reviewed as biological products under the PHS Act, and will be
through IND/BLA process.

	The instrument part and the software portion of this application will
require separate 510(k) submission.  You have heard in a couple of
BPAC's earlier that a licensed test used for screening donors has been
determined to be a major level of concern, so we need whatever is
necessary for the submission to 510(k) has a major level of concern
which is given in this guidance, has to be part of that.

	Also, last October, we used an FDA guidance, which talks about the
current thinking on management of donors and products.

	[Slide.]

	Obviously, to the audience, I don't have to teach all this, what is
needed for the validation of these tests, and also what are the needs
for the clinical tests, so I don't want to go into detail there.

	[Slide.]

	There has been transmission through organ donations.  There was quite a
bit of discussion about what tests would be needed and how would we
protect the organ donations.  Again, this slide has been taken from
Jay's slides.

	The screening of tissue donors will come under FDA regulation after
publication of a final rule on donor eligibility as proposed FDA rule
would require approved donor screening tests for organ donations, and
therefore, a need exists to show the effectiveness of West Nile Virus
screening in the cadaveric blood samples.

	Even though the solid organs and bone marrow are regulated by HRSA, FDA
approves the tests which are commercially available.

	[Slide.]

	So, FDA's current thinking is to recommend routine use of licensed
donor screening tests to detect donor infections, possible use of donor
screening tests under IND. It would be built on existing platforms,
validation in donor screening environment, adequate sensitivity to
detect low level of viremia, and possible need for individual unit NAT.

	Again, will encourage the technologies, such as viral concentration,
which CDC is doing, because as I told you, the virus load is much, much
lower, so to increase the sensitivity and then the development of
reference panels to standardize different tests.

	[Slide.]

	Then, there was quite a bit of discussion on the implementation,
rightfully so, from industry, how would we implement these tests, and
there are a lot of issues which are relevant to that, logistic issues,
and again, some of these have been taken directly from ARC's, Sue
Stramer's presentation, which she described that there is the need to
SOP modification, process qualification, space is a problem because
there has to be enough room for other tests by medical information
systems, which transfers the information, it is getting overloaded, how
do we do that, and impact on the scheduled release of other tests
because there are also other tests, which you will hear this afternoon,
Parvo B19, Chagas, and other tests, individual NAT, so how are we going
to implement all this on top of the other things.

	The other issues that were discussed, the testing, how will the testing
be done, because we heard this epidemic is during certain period of the
year, and will testing be done seasonal versus year-round, geographical
versus national testing, individual versus minipool, do we need to test
other related viruses because SLE, JE, and other infections have been
also shown to occur and are related, transmitted through the blood, do
we need to test those guys, and what have we learned from the past, for
example, St. Louis encephalitis epidemic and what happened, can we think
about that model and applied to this one.

	Those are all hypothetical questions and we need to think about it and
apply estimated risks to determine the need for donor screening.  So,
these were the issues in the implementation section.

	[Slide.]

	So, the general conclusions obviously were that we need to have
specific tests, we need to determine what is the infectious dose of the
virus.  We need to know what components of the blood transmit this
infectivity.  We also need to know how these infectious agents survive
in blood banking storage conditions.

	We also need to have confirmatory tests because this will be screened,
we need to have a confirmatory test. How does it cross-check with other
flaviviruses, or if there is a cross-check, do we need that type of
test, multiplexing of these tests.

	We also need to find out the estimated risk and then the cost of
implementation.  Obviously, FDA is not obligated regarding the cost, but
obviously, we need to think in that direction, too.

	[Slide.]

	So, the general conclusion was that really, you know, I was very much
impressed by the close cooperation between FDA, PHS, device
manufacturers, and blood organizations, which they came all together in
a very positive way to say that we need to develop NAT screening tests
for the West Nile Virus, whether it is nucleic acid based or whether it
is serological.

	Testing will start under IND by the next West Nile Virus epidemic, I
hope so, and meanwhile, the safety of the blood supply can be ensured in
procedures which are in place in blood banking practices, and currently,
FDA has issued a guidance for current thinking on management of donors
and products.

	Hopefully, we will see an outcome in the middle of next year about this
testing.

	Thank you very much.

	DR. NELSON:  Thank you, Dr. Nakhasi.

	Any questions, comments?  Judy.

	DR. LEW:  I just wanted to ask, can you clarify when you say 1 in 150
infected persons develop meningitis or encephalitis, is that 1 in 150
symptomatic or overall?

	DR. NAKHASI:  Yes, 150 infected people.

	DR. LEW:  Well, infected is different from symptomatic.

	DR. NAKHASI:  I think it's symptomatic, is that correct - no, infected,
yes.

	DR. SIMON:  On the presentation that we heard at the last meeting from
the CDC, they indicated that we were, at that time, thought to be about
halfway through this particular epidemic, and I wonder, is there
consideration that we might be at a point when this test is introduced
that the risk has fallen to a low level, and how do you assess that risk
going forward?

	In other words, we will be introducing the test after the time period
during which it might have been useful.

	DR. NAKHASI:  I am sorry, I didn't get the exact question.

	DR. SIMON:  What I am wondering is by the time the test is introduced,
will we have passed through the period of risk and be at a point where
the risk is so low, that there will be little value to the test.

	DR. NAKHASI:  If we are aiming at around maybe hopefully in June or so,
and I think the epidemic which we had, the peak is between late August
to late September, so I think the test, if it is introduced around that
time, if we have a test available, it will not be past that time, so it
will be before that even though there are some cases as early as in May
sometimes.  I believe that we will have a test which may be before that.

	Jay, do you want to say something?

	DR. EPSTEIN:  Toby, I think you are suggesting that we may have had our
epidemic, but no one can really predict what will happen in the next
mosquito season, but the expectation is that we may see another epidemic
of West Nile Virus with human infections in 2003, so the whole concept
is to try to have a test available at least at the investigational level
prior to or at the onset of that season, but no one can predict what
that season will look like.

	DR. NAKHASI:  I hope that there is not, you know, we will see how the
things are going, but if the predictions are that, you know, since 1999,
the epidemics have ranged in the summer months, so even though in 1999,
it was much more localized in the New York area, but then, 2000 and
2001, it was less, but in 2002, it took off.  Who knows what will
happen?

	Again, that is the reason I suggested that we need to think about from
our past experiences, like SLE epidemics, it was 1977 or 1976, there was
a higher epidemic than the following year, there was very little, so you
are right, we hope, we think that if the trend continues, at least we
have a test available at that time.

	DR. ALLEN:  In your background information, you pointed out, as the CDC
did earlier, in your presentation, that the viremia is fairly low, only
about 103 copies/ml. Then, under the Review of Methodologies with NAT
testing, you noted under the caveats that the average human viremia is
18 plaque-forming units/ml.

	Can you reconcile those?

	DR. NAKHASI:  The plaque-forming units and the copies, you know, the
data is not really very well established at this time, so we and our
laboratory and CDC is also really trying to figure out exactly how one
plaque-forming, how many copies/ml, so the copy numbers we do not know
exactly the numbers yet.

	DR. NELSON:  I guess the point is it is probably too low to just simply
add this test to the current pool, minipools or maxipools.

	DR. NAKHASI:  Exactly.

	DR. NELSON:  Whatever the exact numbers of virus are in the average
case.

	DR. SCHMIDT:  Thank you for your very complete report.  One piece of
information that is not in there, is in a CDC publication, saying that
the incubation period of the disease can be as short as two days.

	When we are dealing with something with an incubation period of two
days and talking about the viremia one day after, we just have to look
at this differently, I think, in our planning from our look-see at other
diseases that we are used to dealing with.

	DR. NAKHASI:  Yes, I just actually in one of the slides, I mentioned it
can range from one to five days, so you are right, I think that is very
important.

	There are a couple of things.  One, the viremia is very low, and the
second, the duration can be short, so it is a very tricky situation.

	DR. KLEIN:  Do we know whether antibody confers long-term protection or
can you be reinfected two years from now with variant viruses?

	DR. NAKHASI:  I don't know.  Any West Nile expert around here?  I don't
know how long the protection is.  Mary?

	DR. CHAMBERLAND:  I think the sense is that there is long-term
protection, that once you are infected, you are likely not susceptible,
but how well that has been studied,  I don't know.

	DR. KLEIN:  And that may have implications for the overall epidemic,
not just the seasonal epidemic.

	DR. NAKHASI:  Also, there have been some reports which I remember that
there is some cross-protection from other infectives, you know, like if
you have some other infections, you may have some cross-protection.

	DR. FITZPATRICK:  You did say in your presentation that the whole of
the U.S. is endemic, but there are states where there is neither animal
or human evidence of West Nile, so I think it might help if you would
clarify those states that are non-endemic or those areas that might be
as opposed to.  That statement might be construed as being a bit
misleading.

	DR. NAKHASI:  Maybe from AABB presentation, you may hear that there are
some states which are non-endemic and which are endemic, but actually
Lyle Peterson's chart, which showed the last time, there were some of
the states which were not, but I think the AABB presentation will
clarify that.

	DR. ALLEN:  I think the problem with trying to clarify is we don't know
what is going to happen in the future.  If we had tried to predict based
on what happened in 1999, what would happen in 2001, we probably would
have been quite wrong.  I think we just haven't looked at the spread
yet.

	So far Arizona hasn't had any cases except imported cases, but we are
absolutely certain that within the next year or two, we definitely will.

	DR. FITZPATRICK:  I agree.  I think it would be more truthful to say it
is most likely that the whole U.S. will become endemic, but it isn't
yet.

	DR. ALLEN:  Right, and what is going to happen in terms of endemicity
five years from now in terms of an established recurrent pattern once
this first burst of it, the epidemic has passed across the nation, I
think is anybody's guess at this point.

	DR. NELSON:  Yes.  I guess that deals a little bit with Toby's concern.
 The St. Louis encephalitis epidemic in '75 was a large epidemic,
equivalent to the current West Nile, but subsequently, there were just
handfuls of cases in the subsequent years, even decades.

	But I think the way West Nile is sort of spreading and the fact that
the West has been spared so far except for one case in Los Angeles and
an isolate I guess from Washington and Montana, the likelihood is that
the West well could have an epidemic next year, but it is hard to
predict.

	We could put all this effort into developing a test and then have 10
cases next year.  This is such a complex disease that it is hard to
predict accurately.

	I think AABB wanted to make a statement.  Kay Gregory.

 AABB, ABC, and ARC

Kay R. Gregory

	MS. GREGORY:  Thank you.  Actually, this is a statement on behalf of
the American Association of Blood Banks, America's Blood Centers, and
the American Red Cross.

	As of December 10, 2002, we know that 13 persons have been identified
who acquired West Nile Virus infection from infected blood components
from eight blood donors. These eight donors resided in states where
mosquito-borne West Nile Virus infections to humans was documented by
surveillance during the 2002 epidemic.

	Transfusions of red blood cells, platelets, and fresh frozen plasma
have been implicated.  Persons with transfusion-associated West Nile
Virus infection were aged 7 to 75 years with a median of 47 years.

	Four persons had hematological or other advanced malignancies; three
had stem cell or organ transplantation; and four persons, all 70 years
or older, received transfusions associated with other medical problems
or a surgical procedure.

	In addition, transfusion-related infection was documented in two women
who received transfusions post-partum, and transmission to a
breast-feeding infant from one of these women was documented.  Nine
patients developed West Nile Virus meningoencephalitis and three died.

	As a result of this information, the American Association of Blood
Banks, America's Blood Centers, the American Red Cross, and the
Department of Defense are recommending a voluntary market withdrawal of
selected frozen transfusable in-date products in inventory in an effort
to mitigate the risk of transmission of West Nile Virus through blood
transfusion.

	The frozen products affected are products that were collected in areas
experiencing mosquito-borne transmission of West Nile Virus to humans in
2002.  This withdrawal includes both products that were in the blood
collect facility and products that have been shipped to hospitals for
transfusion.

	The identified periods at issue will vary from state to state and were
developed in consultation with the Centers for Disease Control and
Prevention after review of the relevant epidemiologic and national
surveillance data. The Food and Drug Administration has been briefed on
this issue and is fully aware of this industry recommendation for the
voluntary withdrawal of these products.

	First, quarantine of frozen products collected during the defined risk
periods.

	Blood Centers and hospitals should immediately quarantine all frozen
products collected during the defined risk period.  The risk period is
generally defined as seven days prior to onset of symptoms of the first
reported meningoencephalitis case and ending with the seventh day after
onset of the symptoms of the last reported meningoencephalitis case in
the respective state.  We are providing a table listing this information
for each state.

	Blood collection facilities will inform their hospital customers of the
applicable defined risk period, including the peak incidence reached and
expiration dates of the products involved.  Blood collection facilities
and hospitals should assess the available supply of frozen products as
soon as possible after the initial notification.

	Next, we considered the replacement of quarantined frozen products.  As
soon as feasible, and consistent with the need to maintain inventories
critical for patient care, blood collection facilities will prioritize
replacement of units collected during the week of peak incidence,
followed by replacement of units collected during the antecedent and
subsequent weeks.

	This voluntary market withdrawal is intended to apply to all at-risk
frozen product inventory collected in 2002, with the exception of frozen
rare red cell products, which are to be handled in conformance with
existing protocols for emergency release and transfusion of red cells.

	Finally, prioritization of use of the quarantined product.  To the
extent that quarantined products must be transfused during this time
period due to medical need,  transfusion services are strongly advised
to manage inventories in a manner that avoids transfusion of blood
products collected during the peak incidence week for each applicable
state.

	If it becomes necessary to transfuse quarantined products, a prudent
strategy would be to use those products that were collected as near as
possible to the beginning or the end of the defined risk period.

	Transfusion services are also advised whenever possible to avoid
transfusion of products collected during the entire risk period for each
relevant state to any of the following groups:

	1.  Immunocompromised patients (particularly organ and stem cell
transplant recipients, patients on immunosuppressive drugs, and patients
with hematological malignancies and myelodysplasia and other advanced
malignancies);

	2.  Patients over 65 years of age; and

	3.  Pregnant, immediate post-partum and breast-feeding women.

	Transfusion services may also want to give special consideration to
neonates.

	Let's talk about supply.  To the extent possible, all blood collection
facilities will make every effort to assure that adequate supplies of
frozen products with lesser or no ascertainable risk are provided to
areas where frozen products are at higher risk for West Nile Virus
transmission through transfusion.

	Under existing regulations, withdrawn plasma prepared from collections
of whole blood may be relabeled as recovered plasma.  Blood centers with
existing short supply agreements may continue to ship recovered plasma
for further manufacture under their existing agreements, provided that
temperature storage requirements are met.

	However, blood collection facilities that wish to convert frozen plasma
collected by apheresis during defined risk periods to recovered plasma
prior to the frozen plasma out-date, must request a variance from the
FDA.  We want to stress that FDA will need to act on these variances
expeditiously.

	It is anticipated that cryoprecipitate, and frozen plasma converted to
recovered plasma, that cannot be shipped for further manufacture under
existing agreements will be destroyed.

	Blood collection facilities have committed to make and stockpile frozen
blood components during non-endemic months to minimize the need to make
these products during defined risk periods for human West Nile Virus,
until such time as a licensed test for West Nile Virus or other
intervention (including testing under IND) is introduced.

	We also have provided a list of states for which product retrieval is
not necessary because West Nile Virus is not considered a problem for
those states.

	Then, we provided a list of all other states that are considered to
have periods of risk for transmission of human West Nile Virus for 2002.

	Thank you.

	DR. NELSON:  Thank you, Kay.

	Comments or questions?

	DR. DiMICHELE:  I was wondering if you had an estimate of what percent
of transfused patients your deferral requirements comprise.  In other
words, prioritization of the use of quarantine product under that, you
have actually prioritized groups of patients who should not receive
these products.

	MS. GREGORY:  That is correct.

	DR. DiMICHELE:  What percentage of patients who are transfused would
this group comprise, do you have any idea?

	MS. GREGORY:  No, I really don't.

	Celso, do you have any idea?

	DR. BIANCO:  No, we don't have an exact idea how many patients will be
affected.  We do not have an idea how much product is still available in
hospitals at the present time, but there was a lot of thought into that
and we were trying to do the best and to predict that in some
situations, we may have to prioritize.  Hopefully, most of it will be
replaced by product outside.

	This will be more difficult for states where the epidemic has been very
intense and very long - Louisiana, Michigan, and in Texas, and we hope
we will be able to replace that product as soon as possible.

	DR. DiMICHELE:  The reason I was asking is that one group that is not
included here is the chronically transfused group of patients, the
patients who are receiving blood every two weeks.

	Is there any reason that they were sort of excluded from this
prioritization list?

	DR. BIANCO:  The prioritization, Donna, was made, they don't seem to be
immunosuppressed, was based on the cases that were observed.  There was
a lot of transfusion during the period, so probably a lot of infected
units were transfused, but those were the cases that were identified and
reported, and that constituted the patient population. There were no
neonates, but it was thought that it was prudent to do that.

	DR. DiMICHELE:  Thank you.

	DR. PAGE:  Peter Page, American Red Cross, Arlington, Virginia.

	Pertinent to your first question, one could say that for every 100
units of whole blood collected, almost 100 red cell units are prepared
and transfused, but only about 20 of them result in a plasma product for
individual transfusion to a patient.  The rest are essentially
fractionated.

	So, on the average, as far as number of units, not number of patients,
it is about 20 percent has got red cells.

	DR. NELSON:  Thank you.

	Next, is Dr. Mary Elizabeth Jacobs talking about medical device user
fees.

Medical Device User Fee and

Modernization Act Of 2002 (MDUFMA)

Mary Elizabeth Jacobs, Ph.D.

	DR. JACOBS:  Thank you, Mr. Chairman, and good morning.

	I am here today to tell you about the Medical Device User Fee and
Modernization Act of 2002, which was signed by the President on October
25th.

	[Slide.]

	I would like to cover an overview of MDUFMA, the law itself and how it
was developed, the user fee provisions, the performance goals that are
related to the user fees, third-party inspections, which is one of the
major provisions in the law under the modernization part of MDUFMA, some
additional provisions, and then implementation, where we are now.

	I have titled this part MDUFMA and CBER because there is one provision
that applies to CBER in particular that I want to mention and another
provision which as a practical matter applies less to us.

	However, I want to emphasize that all of these provisions apply to any
center in FDA that regulates medical devices.  CBER has been very
committed to making this work and we have been involved in all stages of
the analysis and negotiations.

	First of all, we regulate at CBER up to 10 percent of the device
workload in any given year.  That comes primarily under blood-related
devices, such as the blood screening tests which are used to screen
donated blood.

	We also are involved in combination products, which is specifically
mentioned in the law, and combination products are products that have a
combination of a biologic, a drug, and a device, two or three of those. 
For example, there are hemostatic agents which include device components
and thrombin.

	We are in a very active implementation state and you are going to be
able to get information as it is developed, and I want to tell you the
two places you can get that.

	First of all, most of you probably know that our web site is fda.gov. 
You can then go to the Biologics Center or you can go to the Device
Center.  We anticipate having one web site for all FDA centers related
to devices, however, right now, as an interim measure, you can go for
general information to the Devices Center, which is CDRH, and go to
their web site, and you can send general inquiries to them at
mdufma@cdrh.fda.gov.

	For CBER-specific information, you can go to our web site, which is
CBER under the FDA web site, go Devices, and under that, MDUFMA, and you
can send in inquiries, as you always can to us.  For manufacturers, it
is matt@cber.fda.gov.  For consumers and health care professionals, it
is octma@cber.fda.gov.

	[Slide.]

	First, what is the background?  The law was developed in consultation
with the industry, the Congress, FDA, and with input from other
organizations including consumers and patient groups.

	The two major industry groups are AdvaMed, which used to be HEMA, and
MDMA, Medical Device Manufacturers Association, and it had bipartisan
House and Senate support.

	In addition, during the negotiations, we invited in all the people who
belonged to our BPAC mailing list. That includes the AABB, ARC, ABC, and
all the consumer and patient groups, and we had a separate session with
them. Most of them came, and we went through all the provisions with
them.

	The law explicitly recognizes the need for additional medical device
resources, and the basic idea behind user fees is that FDA will commit
to faster review times than we are required to under the law.

	This represents approximately 25 percent improvement in our review
times.  It isn't 25 percent for every single kind of application.  For
example, the expedited, which are very novel products, have a greater
improvement time than some of the ones in which we had better times.

	In exchange for this, the firms agreed to pay user fees, which will
give them greater predictability.  FDA, prior to having this, has
already had 10 years of successful experience with what is called PDUFA,
which is user fees for prescription therapeutic drugs.  So, this MDUFA
is building on that experience although it differs in certain ways.

	It explicitly recognizes the need for additional resources and this has
an appropriations piece.  This law is not just user fees, it also has
appropriated funds from the Congress, and it was signed, as I said, by
October 26th, so the implementation clock is ticking.

	[Slide.]

	What are the key provisions?  First of all, there are medical device
user fees and, as I said, additional appropriations from the Congress. 
It includes third-party establishment inspections, which I will discuss,
and that is covered by approximately 25 percent of the law.

	It has greater oversight of reprocessed single-use devices, and this is
the provision which I mentioned which doesn't, as a practical matter,
come through CBER.  These are primarily surgical instruments which are
manufactured by what we call OEMs, original equipment manufacturers.

	They are labeled for single use.  They then are frequently reprocessed
and resold and redistributed.  As far as we know now, those will go
through CDRH.  It has provisions for supplying labeling electronically.

	It has modular review of PMAs in the law, and we at CBER have already
had modular reviews of PMAs, but that has been a matter of policy, and
this is the first time it has been in the law.

	Then, it has oversight by the Commissioner's Office of the combination
products to which I referred.

	[Slide.]

	Now, what are the user fees?  First of all, they apply to the major
classifications of submissions, but not to all of them.  For example,
for PMAs which are the more novel devices, for the BLAs which would be
the licensed tests for blood screening, and to 510(k)s which, for those
of you who know these, these are in general lower risk products in which
you deal with the substantial equivalence to products which were
originally on the market.

	However, we have committed to maintain our performance on other kinds
of submissions for which there are no user fees.  For example, the
investigational device exemptions in which we have 30 days to tell a
firm no, you cannot start a clinical trial.  We are maintaining our
performance on those.

	The structure is that we anticipate that there would be, beginning in
2003, this is the first year, $25.1 million in fee revenues, rising to
$35 million in FY 2007. Then, there are workload compensations and other
things which I would refer you to the law on those.

	Then, we have $15 million in additional appropriations, to bring the
total by the end of 2007 up to $50 million.

	Now, one of the questions we are asked is do you have that
appropriations passed yet, and the answer is not yet passed, the
Congress will be coming back after the first of the year, however, we
are actively implementing this and assuming it is going to be passed.

	[Slide.]

	The first year fees range from $154,000 for a premarket application, to
$2,187 for a 510(k).  There are reduced fees to protect small
businesses, "small" meaning sales and receipts less than or equal to $30
million.

	For small businesses, the fees are 38 percent of a standard fee for a
PMA, except for 510(k)s, in which case they are 80 percent.  The small
business fee for 510(k) starts in 2004, and it sunsets October 1, 2007.

	The device industry includes a wide range of corporations, some of
which are extremely large global corporations, and some of which are
almost amazingly small corporations.  This was to have a structure that
was appropriate to all of them.

	The sunsetting in five years is similar to PDUFA. That also has had
two, five-year cycles.  We are now in the third cycle PDUFA 3, so this
will sunset in five years unless it is renewed in some form as MDUFA 2.

	[Slide.]

	There are some waivers.  There is no fee if the applicant is a Federal
or State Government, unless it is going to be marketed.  The first
premarket application by a small business is free.

	The first premarket report for a reprocessed device is free.  There is
no fee if there is a third-party review of a 510(k).  We do have a
provision in which certain third parties can review 510(k)s.  That is
actually not very widely used.  They are then submitted to FDA for
concurrence or nonconcurrence.  Finally, there is no fee if the device
is solely for pediatric use.

	[Slide.]

	Now, what are the performance goals?  Overall, we are aiming to improve
our performance by 25 percent.  These goals are defined in a letter from
the Secretary of HHS, Secretary Thompson, to the Congress.

	It differs from PDUFA in this.  We have a combination of cycle goals,
which means a firm sends us a submission and we respond to them.  That
is one cycle.  For PMAs and 510(k)s, decision goals, meaning the time in
which FDA finished its review and telling a firm it is approved, it is
denied, or what it has to do exactly to get approved. That could be a
nonapprovable or approvable decision.

	The goals are measured in FDA days, so they are independent of the time
that it takes for a firm to respond to us.

	[Slide.]

	The performance goals are very detailed, but I want to just go through
with you what is the basic structure on all of these performance goals,
so that when you read the goals letter, it will be more clear.

	First of all, for BLAs, the licensed screening tests, as we said, the
law takes effect October 26, 2002 for Fiscal Year 03.  The first goals
start in 2005, and that is because the initial funds are going to be
used to hire additional people.  That is why the goals are starting in
2005.

	However, we are having an annual report with our stakeholders.  We
fully intend substantial progress on these, but formally, the goals come
in, in 2005.

	Secondly, the goals ramp up from 2005 to 2007, so for BLAs, the goals
would go from 75 percent in 2005, to 90 percent in 2007.

	The third thing is, as I said, we are reducing the time frame which is
in the law for all of these categories. For example, for the initial
submission on a BLA, which we call our response review and act on, which
includes the complete review, the inspection, and going to an advisory
committee if that is necessary, instead of doing those in 12 months, we
will be doing the goals in 10 months.  That is the basic structure.

	Secondly, for BLAs, we now have a new category of resubmission.  That
means after we respond to you from your first cycle, the firm then
responds to us unless they are licensed in the first cycle.

	If their response has a substantial amount of data, for example, a
completely new study, we have six months to respond.  If they have less,
for example, if they are providing updated stability information, we
have two months to respond.

	Those categories have already been used in PDUFA and I would anticipate
the criteria we would use would be very similar to those that are used
for the PDUFA guidance which is on our web site.

	Finally, our manufacturing supplements again will go, instead of being
done in six months, they will be 75 percent to 90 percent of the
manufacturing supplements, that is, after licensing, would be done in
four months.

	[Slide.]

	Let's go to PMAs, which would be used for the HIV diagnostics, which
have been handled by CBER because of all of our work with HIV as it
relates to the blood supply. These are handled by PMAs again.  The
target goals go from 2005 to 2007.  They have cycle goals and they have
the decision goal.

	For example, a cycle goal would mean your first letter would be if you
have major deficiencies, you would get what we call a major deficiency
letter.  Instead of doing that in 180 days, the goal is for 70 to 90
percent, ramping up again from 2005 to 2007, you would get the letter in
150 days.

	For PMAs, the decision goal, when we finished our review, the goal is
that by 2007, we would have 50 percent completed in 180 days.  That is
quite a challenging goal for FDA.

	Another provision in the law, which applies to this, and the next
category I am going to talk to you about, is that because that is such a
challenging goal for us, we have a provision in the law that says we
will notify the Congress following a public meeting in 2006 if we think
that that would be a problem for FDA to meet that goal.

	[Slide.]

	Let me now go to 510(k)s.  The 510(k)s are the ones that are the more
abbreviated kinds of applications which we call "substantially
equivalent."

	Here, we have instead of 90 days for our first letter, which we call
"additional information," it is like a deficiency letter, you would get
that in 75 days instead of 90 days, again changing the statutory
deadline.

	This also has a total time for the decision of 75 percent within 90
days, again a very challenging goal for FDA, and because of that, we
again have the structure of a public meeting and a report to Congress if
we believe that we can't meet that by 2007.

	[Slide.]

	The next provision is for third-party inspections. I mentioned to you
that we do have a provision for third-party reviews for what are called
"510(k)s."

	Third-party inspections, I am only hitting a few of the points, 25
percent of the law covers third-party inspections.  There was interest
by a number of firms which market globally and which have inspections
for other standards, for example, ISO, to be able to schedule all of
their inspections together by paying a fee.

	There was also some interest because at times FDA inspectors, because
of the international situation, cannot go into certain countries
although European inspectors are there.  So, this would potentially
solve some of the problems of companies which are in those countries.

	These have the most complex provisions.  In order to be accredited, the
third party has to have the same conflict of interest provisions as we
do internally at FDA. For example, people who would be third-party
inspectors for medical devices cannot own stock in companies that are
regulated by FDA, for example, food companies.  So, they are just as
stringent as those for FDA employers.  That is only one of those.

	These are all going to be spelled out in guidance to you, but there are
already many of them in the laws.  The inspections are permitted only
for quality systems in GMP. If it is pre-approval, BiMo, which is our
monitoring of studies, and "for cause," those are exclusively for FDA.

	[Slide.]

	We must publish our accreditation criteria by next April.  Those will,
of course, be on the web site.  They will cover establishments that
market in the U.S. and abroad and where the other country accepts FDA
inspection results.

	The most recent FDA inspection must be classified as No Action
Indicated or Voluntary Action Indicated, which means that the firm is
already in good compliance before this happens, and FDA must
periodically inspect, and this is anticipated to be one out of three.

	Again, I have only hit some of the major points in this.  If you are
interested, please refer to the law and to our web site.

	[Slide.]

	Here are some additional provisions which would be of interest to our
group here.

	First, combination products.  Those reviews are going to be coordinated
by a new office in the Office of Commissioner.  This is because firms
were concerned because frequently, one center is the lead, another firm
is very active in consultation.  They want to make sure that there are
adequate tracking systems, so we are going to be having and are
developing new tracking systems for this.

	We already have a courier system between the centers.  We had a meeting
on November 25th, which we call a Part 15 hearing, which means we
solicited input from firms and from other groups, and they made a number
of provisions.

	I already mentioned to you that under some circumstances, there will be
electronic labeling.

	Finally, I want to mention the provision which is specific to CBER, but
could be also for the Center for Drugs if they have device reviews. 
Under Section 205, there will be a one year report to Congress on the
timeliness and effectiveness of premarket reviews by centers other than
the Center for Device and Radiological Health.

	That means CBER will be developing a report which will go to the
Commissioner's Office and to the Department about our timeliness and
effectiveness, and our regulation of these devices.

	[Slide.]

	Next, let's go to implementation.  We are very actively working on this
now.  We are developing the basic reference materials.  You can look on
the web site.  We have a kind of plain language version of the Act and
frequently asked questions.

	We have implementation teams for all of these major provisions and CBER
is very actively involved in those.  I just want to mention to you that
there is one specific to stakeholder education.  We have active
training, for example, tomorrow, we have required training for everyone
involved in reviewing these.

	[Slide.]

	Finally, I want to mention how can you make your views known to FDA.  I
already mentioned that we are opening a docket.  That means you can send
them in there.  There will be annual public meetings starting in FY 2004
to review our progress.

	The law specifically mentions consultation on certain specific policies
including bundling of submissions and modular PMA.

	Please look at our web site, send in your questions, and I would be
happy to address any questions you have.

	Thank you.

	DR. NELSON:  Thank you, Dr. Jacobs.

	Any questions?

	DR. FITZPATRICK:  I just had one.  You exempted State and Federal
agencies.  What about nonprofit corporations?

	DR. JACOBS:  That is a good question.  I would have to go back and
check the law to see if that is in there, and let me bring that to
people's attention.  I am not sure if that has been addressed.

	Thank you.

	DR. NELSON:  Thank you.

	Next, is an update on the approval of the OraQuick Rapid HIV-1 Antibody
Test.

Approval of the OraQuick Rapid HIV-1 Antibody Test

Elliot P. Cowan, Ph.D.

	DR. COWAN:  Thank you, Dr. Nelson.

	[Slide.]

	The purpose of this update this morning is to inform you that on
November 7th of this year, FDA approved the OraQuick Rapid HIV-1
Antibody Test.

	The intended use of the OraQuick Rapid Test is to detect antibodies to
HIV-1 in fingerstick whole blood specimens, as a point-of-care test to
aid in the diagnosis of infection with HIV-1, and this test is intended
to be suitable for use in multi-test algorithms designed for statistical
validation of rapid HIV test results when such algorithms have been
evaluated and approved.

	[Slide.]

	OraQuick is approved as a restricted device.  Sale is restricted to
clinical laboratories, number one, that have an adequate quality
assurance program including planned systematic activities to provide
adequate confidence that requirements for quality will be met; number
two, where there is assurance that operators will receive and use the
instructional materials.

	It is approved for use only by an agent of a clinical laboratory.

	[Slide.]

	The test subjects must receive the "Subject Information" pamphlet prior
to specimen collection and appropriate information when test results are
provided.

	The test is not approved for use to screen blood or tissue donors.

	In addition, a customer letter will be included with all kits that are
shipped, which has the provision that "By purchasing the device, you are
doing so as an agent of a clinical laboratory and agree that you or any
of your consignees will abide by the...restrictions on the sale,
distribution, and use of the device."

	[Slide.]

	What I would like to do now is just run through the device, to describe
it for you and how the test is performed.

	It consists of several components including the main device itself, as
well as a vial of buffer solution, the stand to hold the buffer
solution, and a specimen collection loop.

	[Slide.]

	The first step in the procedure is to provide the test subject with a
Subject Information pamphlet.  This information pamphlet, it is a
multi-page pamphlet containing such information items as what are HIV
and AIDS, how does someone get HIV, to what is the OraQuick device, to
the interpretation of the results, to where can I get more information
about HIV and AIDS.

	[Slide.]

	A fingerstick is performed and the sample is collected within the
specimen collection loop.

	[Slide.]

	That is then added to the vial that contains the test developer
solution.  The sample is mixed in the vial.

	[Slide.]

	The device is then inserted into the vial and then a time period of 20
to 60 minutes later, a result is read.

	[Slide.]

	The last step of the procedure calls for following CDC guidelines to
inform the test subject of the test result and its interpretation.

	Let me just show you what some of these results look like.  Before I do
that, let me just point out that there are two lines that could appear
on this test.  Number one, there is a line at the C position, which is
the control, and at the T position, which is the test.

	The C position will detect antibodies to human immunoglobulin. 
Therefore, this serves as a procedural control to ensure, number one,
the specimen has been added, and, number two, that all the components of
the test are working properly.  All valid tests will have a line at the
C position.

	The T position, on the other hand, contains peptides to HIV-1, and a
line here will indicate a reactive result.  So, in this case, I am
showing you a nonreactive result which is interpreted as negative for
anybody as to HIV-1.

	[Slide.]

	Here are some examples of reactive results.  The intensity of the lines
may vary relative to one another, but any appearance of color at the T
position is considered to be a reactive result.  Reactive results are
interpreted as preliminary positive according to CDC guidelines.

	[Slide.]

	Finally an invalid result will occur if there is no line at the C
position for the control.  Even in the presence of a line at the T
position, this would be considered an invalid result also.

	This is invalid because of high background and the inability to see
lines, this is considered invalid because the line does not appear in
the proper position.  Invalid test results should be repeated.

	[Slide.]

	OraQuick kit controls consist of a negative and a positive sample.  The
positive is low reactive.  These are provided separately as an accessory
to the kit.

	In the product package insert, it is stated that kit controls should be
run under several situations, number one, by each new operator, prior to
performing testing on patient specimens, whenever a new lot of OraQuick
is used for the first time, if there is a change in the conditions of
testing, for example, new location, lighting, temperature, that sort of
thing, and also a periodic interval specified by the quality assurance
program of the laboratory doing the testing.

	[Slide.]

	I would like to now run through some of the clinical trial data used to
support the approval of this test.  For sensitivity, there were three
groups of specimens that were studied, AIDS, known HIV-1 positives, and
high risk specimens, a total of 1,146 specimens, of which 538 of those
were determined to be true positives.

	The OraQuick correctly identified 536 of these. Two specimens from
known HIV-1 positive patients were not detected.

	The sensitivity in these studies, it was therefore determined to be
99.6 percent with a 95 percent confidence interval of 98.5 percent to
99.9 percent.  I would like to point out that this is within our minimal
performance criteria for the performance of a rapid HIV test for
sensitivity, which is 98 percent as lower bound of the 95 percent
confidence interval, and that criterion was discussed at several BPAC
sessions.

	[Slide.]

	Also, in support of sensitivity, analytical sensitivity studies were
done looking at 11 seroconversion panels and 2 low titer panels.  The
performance of OraQuick was similar to licensed EIAs for the specimens.

	In addition, unrelated medical conditions and interfering substances
specimens were spiked with an HIV-1 positive specimen to give low
positive reactivity.  Again, in this case, all spiked specimens gave
reactive results.

	[Slide.]

	For the specificity, a total of 1,250 low risk specimens were looked
at, as well as non-reactive specimens from the high risk study, making a
total of 1,856 true negative specimens that were examined.  OraQuick
correctly identified all of them.  There were no false positive
specimens in this study.

	So, again, in these studies, specificity was determined to be 100
percent with a 95 percent confidence interval of 99.7 percent to 100
percent.  Again, these are in line with our minimal criteria for
performance for specificity as discussed at BPAC for a rapid HIV test,
which is also 98 percent is the lower bound of the 95 percent confidence
interval.

	[Slide.]

	Also, in support of specificity, unrelated medical conditions were
examined, a total of 321 specimens, as well as 119 specimens with
interfering substances.  There were a few specimens that gave false
positive results in this case, but the caveat here is that all of these
specimens or most of these specimens were frozen repository specimens.

	I would like to remind you that the intended use specimen type for
OraQuick is a fresh fingerstick whole blood specimen.  If anything, a
repeatedly frozen and thawed specimen would be expected to give a false
positive result if there is a problem at all.

	[Slide.]

	For reproducibility, reproducibility studies involved three sites,
three lots, three different days, and three operators per site, making a
total of nine operators who examined the blind-coded panel of five
contrived whole blood specimens.  Four of these were anti-HIV-1-positive
and one was anti-HIV-1 antibody-negative.

	The results for the 20-minute read time were 99.8 percent agreement,
and at 55 to 60 minute read time, 100 percent agreement.

	[Slide.]

	I would also like to touch just very briefly on CLIA issues since this
is something that we have talked about at BPAC before.

	This test on approval was categorized as moderate complexity.  The
Company stated publicly on September 11th of this year that they will
apply for CLIA waiver.

	On November the 7th, at the time of the approval, Secretary Thompson
made a statement at the OraQuick approval press conference, "I strongly
encourage OraQuick to ask the FDA for a CLIA waiver... if the company's
data prove that the OraQuick test is safe and easy to use, it can get a
CLIA waiver."

	[Slide.]

	Finally, I would like to point out that there are a number of things on
the CBER web site.  The approval letter for OraQuick is listed, as well
as the package insert, the summary basis of approval, and an FDA
talkpaper.  I have listed the web site for you here.  That could give
you some more detailed information.

	I would like to close by saying that we are continuing to actively work
with additional manufacturers to approve more rapid HIV tests, so that
we can move toward multi-test algorithms.

	Thank you very much.

	DR. NELSON:  Thank you.

	Questions or comments?

	I noticed it is not approved for use in blood banks at this time, and
it may not be terribly useful in that setting in the U.S., but I think
in many developing countries where it is very difficult to follow and
recontact donors once they leave the blood banking system, I know that
it can be a real horrendous problem.

	I can see where in some settings and at some blood banks, a rapid test
could be very useful.

	DR. COWAN:  We have actually approached the company to provide us with
data to support the use of OraQuick in this country as an emergency
blood donor screen. In the absence of any data, though, we couldn't do
that at this time.

	DR. NELSON:  I imagine that maybe Celso or somebody knows that there is
probably very close to 100 percent success in finding a positive donor
once all the tests results are available in this country, but it may not
be 100 percent everywhere.

	When that result goes down to, as in Northern Thailand, maybe 50 or 60
percent, that can be a real problem.

	Thanks very much.

	The next topic is Bacterial Contamination.  We will start with Dr. Alan
Williams.

	DR. SMALLWOOD:  I would just like to inform you that there is an
electrical problem in the surrounding area, and Pepco is working on
that, so we may have some intermittent interruptions, but I hope it
won't be permanent.

I. Bacterial Contamination

A.  Background and Introduction

Alan Williams, Ph.D.

	DR. WILLIAMS:  Thank you and good morning.

	What I would like to do with this introduction is give a very brief
overview of what is admittedly a very complex topic, and in the course
of that, try to emphasize some of the key points that are in need of
discussion and deliberation and emphasize those that are the topics for
this meeting and others that might be appropriate for future discussion
just to help provide focus.

	I will then finish up by outlining the list of speakers for this
session and the questions that are being posed to the committee.

	[Slide.]

	The first slide deals with the frequency and importance of bacterial
contamination in the transfusion setting.  Sepsis is, in fact, the
second leading cause of transfusion-related fatalities.  It follows a
group in type and compatibility fatalities.  It is the second cause in
transfusion-associated acute lung injury is the third cause.

	There are actually five to nine recognized fatalities per year
associated with sepsis.

	The most comprehensive study documenting clinical cases is the
CDC-sponsored multi-center bacterial contamination BACON study published
recently which documented that for single-donor platelets, clinical case
rates were 9.98, close to 10 per million, of which 1.94 per million were
fatal.

	Among random donor platelets, the rate is just slightly higher, 10.64
cases reported per million random donor platelets, 2.2 per million
fatal.

	Among red blood cells, refrigerator temperatures, case rates were 0.21
per million, of which 0.13 were fatal.

	A conclusion was made that among the fatalities, most of them appeared
to be related to gram-negative organisms and also related to units
containing high levels of endotoxin.

	Those are clinical cases.  A different consideration is the amount of
contamination in units that may result in a spectrum of outcomes in the
recipient from no effect up to fatality.

	The generally accepted figure for platelet units which are stored at
room temperature is between 1 in 1,000 and 1 in 2,000 contamination
risks per unit, but reports vary widely among institutions and among
different studies that are published.

	I think relevant to this is a study by Dr. Leiby with the Red Cross
which looked at outdated platelets, studying close to 5,000 units.  They
found 4 to be positive, for a prevalence of 0.08 percent in that
published study.

	[Slide.]

	Where does the contamination come from?  In many cases, it simply isn't
known, but due to the nature of the organisms and other criteria, it is
known that skin contamination logically is the source of much of the
product contamination.

	This can occur by bacteria that are on the surface and are incompletely
disinfected by the pre-phlebotomy decontamination process or because
blood units are drawn with a large needle, there can, in fact, be a
tissue plug that is caught up in the needle and makes its way into the
collected blood product.

	There also can be occult bacteremia in a donor who appears otherwise
healthy, but may, in fact, be circulating bacteria in the blood.  As
mentioned, the contamination prevalence and incidence as measured by
patient outcomes varies by site.

	I think it is important that because this does vary, there may be room
to control some of these extrinsic control points.  One would expect
that bacteremia in a donor for the most part might be a static level and
that extrinsic contamination may contribute to some of the different
levels of contamination that have been reported.

	[Slide.]

	 A brief description of platelet components.  Apheresis components,
this is where the donor is hooked up to a machine for a period of time
and one or more components is removed.  It is also known as single donor
platelets.

	These products may, in fact, be split and the split units are then
counted to make sure that they contain a minimum platelet count.  In the
country, about 6 million units per year are transfused, and these
products had a five-day shelf life.

	The other class of platelet products are pooled random donor platelets.
 There are the products derived from whole blood collections. 
Approximately, 3 million units per year are transfused, and these are
pooled together from individual platelet concentrates derived from the
whole blood units in different quantities, typically 4 to 6 platelet
concentrates from allogeneic donors are pooled to make a dose of random
donor platelets.

	This pooling procedure, because it involves connecting up to individual
platelet concentrates, currently requires a four-hour hold after the
pooling procedure, and this is typically done in the transfusion
service.

	[Slide.]

	Detection methods that are currently available. Clearly, the most
sensitive and most widely available at this point is culturing, and
there are variables related to the time the culture is taken, the
volume, and source of the sample, how long the culture is incubated, and
what type of detection system is used to monitor the culture.  You will
be hearing a lot more of that in the course of the session.

	There are other techniques available.  For the most part, without going
into great detail, they tend to be considerably less sensitive than the
culture mechanism, but may serve as a very quick read on an individual
unit of platelets prior to transfusion as to whether there is moderate
to gross contamination.

	These other methods include urine dipsticks to measure pH less than 7
or a glucose level, Gram or other types of stains, swirling, and a
technique that shows promise in terms of sensitivity, but needs further
development is actually doing polymerase chain reaction looking at RNA
content of different bacteria.

	[Slide.]

	Now, two automated or semi-automated culture devices were recently
cleared by the FDA for quality control purposes, and I think a major
issue that we will be dealing with today is that they are cleared for
quality control, not for pre-release testing.  This is not a screening
test which allows labeling of the product.

	These two tests are the Biomerieux BacT/ALERT System, which is cleared
for quality control of leukoreduced apheresis platelets, and there is a
specific labeling indication in there that it is not designed for
pre-release testing.

	This system detects both aerobes and anaerobes although the latter
appear to be infrequent causes of clinical sepsis in recipients.  The
system produces continuous monitoring and it is a calorimetric sensor.

	The second system, made by Pall Medical Corporation, is a bacteria
detection system, or BDS.  It is cleared for the quality control of both
leukoreduced random and apheresis platelets.  The system detects
primarily aerobic bacteria, and the sample could be taken as early as 24
hours after the platelet unit collection.

	[Slide.]

	Now, two issues are going to be recurring throughout the discussions
today, but I want to point out that these are areas where there are data
needed, and FDA is very interested in reconsidering the issues based on
available data, but not specific topics for today's question
consideration to the committee.

	The first is the four-hour hold for pool random donor platelets, should
that be extended and particularly should it be extended if we have
procedures for culturing these units to determine sterility.

	This is actually a regulation CFR 606 122.  It raises a concern in
terms of platelet pools in terms of trying to culture or do quality
control because of the four-hour hold with the pool, it provides
insufficient time for sampling that pool and developing a culture result
before the pool actually would be transfused.

	The techniques used for creating pooled random donor platelets are
typically sterile dock welding of the individual platelet concentrates. 
There has been a lot of experience with this procedure.  As you will
hear today, there is one paper published in Europe in I think 1997,
which called into question the sterility of the tube welds and whether,
in fact, when the sterile docking device is used to create pooled
platelets, whether sterility of the final product could be compromised. 
There will be specific discussion about that.

	Also, FDA feels that to extend the four-hour hold, it would also create
considerations beyond contamination, such as measuring in vitro platelet
function, in vivo efficiency, and concerning the fact that mixed
leukocyte response to the set may take place when allogeneic units are
pooled may stimulate cytokine release.

	The second issue is extending the five-day platelet storage.  This is
based on an FDA memo issued in June of 1986.  Clearly, extending
platelet storage would be dependent on an approved pre-release bacterial
detection system, not upon the QC systems currently cleared, and also
would require data related to platelet efficacy when held seven days.

	[Slide.]

	Now, the four key elements that we are going to focus on in today's
session regarding reduction of bacterial contamination risk is effective
arm preparation, an update on the diversion pouch.  This would be an
integral pouch that would potentially capture the first 30 ml or so of
blood, as well as any skin plug and hopefully pull off any bacteria that
might be associated with that early volume.

	It would be a discussion of FDA current thinking in terms of quality
control mechanisms and data that might be needed to establish
pre-release screening approval.

	[Slide.]

	On the arm preparation subject, which will be next, there will be a
background review of the literature by John Lee in our Division of Blood
Applications.

	There is a question associated with this, and I will note that all of
the questions are going to come at the end of the session because of the
need to integrate the public hearing.

	The first question is:  Do available scientific data support
preferential use of an isopropanol/tincture of iodine procedure for
preparation of the donor's phlebotomy site?

	[Slide.]

	The second subsection will be an update from Dr. Jaro Vostal of the FDA
on the diversion pouch and current FDA thinking.  There is no question
associated with this.

	[Slide.]

	The third subsection is the discussion of quality control.  I am going
to introduce some concepts, as well as current FDA thinking, on a
quality control approach.  There will be a data presentation by Dr.
James Aubuchon from Dartmouth University on sterility of plastic tubing
welds, as well as transfusion service experience with universal
bacterial culturing of apheresis platelet units.  We are representing
published experience with these two procedures.

	[Slide.]

	Thirdly, data on the sterility of the plastic tubing weld by Tracy
Manlove with Terumo Medical Corporation.

	[Slide.]

	Questions related to quality control.

	Question No. 2.  Do available data on the sterility of the sterile
connecting device procedure support the use of this procedure to collect
samples for bacterial detection from in-date platelet products?

	Question No. 3.  Does the committee concur with FDA's proposed
statistical approach to providing quality control for platelet
contamination?

	[Slide.]

	The final subsection, a discussion of data that might support
pre-release screening.  There will be a presentation by Dr. Steve Wagner
with the American Red Cross Holland Laboratory on design of clinical
trials for clearance of devices intended for screening of platelet
products prior to transfusion, so it will be a background talk.

	Again, Dr. Jaro Vostal will then present FDA's current thinking about a
study design.

	[Slide.]

	A question related to this.

	Question No. 4.  Does the committee concur that data derived from FDA's
proposed clinical trial design would be appropriate to support clearance
of devices for pre-release screening of platelet products for
transfusion?  Yes or No.

	So, it will be a packed session.  I tried to establish some context for
you.  You will hear more about each of these topics as the session
proceeds, but with that in mind, I would be happy to entertain
questions.

	DR. NELSON:  Toby.

	DR. SIMON:  As some people may be aware, there is two prevalent systems
in the laboratory industry for doing these kinds of cultures with rapid
results - one, the BacT/ALERT, which you have approved for quality
control of platelet screening, and the other, the Bactec system
manufactured by Becton Dickinson.

	My understanding is that as a result of litigation between the two
companies, BD is unable or has agreed not to sell the Bactec system to
blood centers.

	Is the fact that you have cleared two devices for the quality control
of platelets mean that non-cleared systems, such as the Bactec, could
not be used on any FDA scheme for quality control by licensed and
registered organizations?

	DR. WILLIAMS:  I am going to defer the answer on that to someone else
if I can.

	DR. EPSTEIN:  Well, use of those systems would be off-label use because
they are not indicated for quality control monitoring of platelets, so
it becomes an issue of enforcement discretion.  I can't tell you that
that is permitted.

	DR. NELSON:  Any other questions?

	DR. FITZPATRICK:  Alan, it might be a little confusing.  You used the
term "four-hour hold" for the pooled random donor platelets.  Actually,
it is a four-hour expiration after pooling meaning that they have to be
transfused within that four-hour period, right?

	DR. WILLIAMS:  That is correct.  I think "four-hour hold" is kind of a
term in common use, but that is correct, they do expire after four
hours.

	DR. NELSON:  The next presentation is by Dr. John Lee on Skin
Preparation of Phlebotomy.

B.  Skin Preparation of Phlebotomy

John Lee, M.D.

	DR. LEE:  Thank you, Dr. Nelson, and good morning.

	As Dr. Williams pointed out, effective donor arm preparation is a key
step in preventing bacterial contamination of blood, particularly
platelets.

	[Slide.]

	Up to this point, whenever we mention blood safety, we have typically
meant viral safety, and much of our effort has been directed at
improving viral safety towards zero risk.  Bacterial safety has been in
relation somewhat neglected, but nonetheless, a very important area for
further study and concentration.

	As Dr. Williams pointed out, platelet transfusion and particularly
platelets have been implicated in many transfusion-associated problems,
even fatalities, and contamination rate in platelet concentrates has
been generally accepted to be about 1 unit in 1 to 2 per 1,000 units
collected, and the actual rate of contamination depends on storage
duration.

	As you all know, platelets are stored at room temperature and the
longer you store at room temperature, the higher the risk,  and this is
due to the fact that bacteria, unlike viruses, multiply within the
collected blood after a brief lag phase.

	When these causative organisms are identified, they turn out typically
to be--well, I wouldn't say typically--but they often turn out to be a
member of the skin flora.  So, it is a reasonable conclusion that
inadequate donor skin antisepsis is a major contributor to bacterial
contamination of blood.

	[Slide.]

	The current most widely used technique in the U.S. in preparing the
donor arm is the method outlined in the AABB Technical Manual.  This is
a two-stage procedure where at least an 8 cm diameter area is selected
for phlebotomy.

	As a Stage 1 procedure, that area is first scrubbed with a 0.7 percent
iodophor preparation for at least 30 seconds.  That area might be wet. 
It is actually written in the manual that you need not wait for it to
dry and move on to the second step, where the second step consists of
applying a 10 percent povidone-iodine, which has a 1 percent
availability of free iodine.

	This is to be applied beginning with the site of phlebotomy, the needle
entry point, and move outward in a concentric spiral.  After covering
all of the at least 8 cm diameter area, that area should be allowed to
stand for a minimum of 30 seconds.

	[Slide.]

	Now, this method has been in use at least in the U.S. widely for many
years.  There has been a recent challenger to that method, and that
method has been described by two authors, the studies by those authors I
will describe in a few minutes.

	This is a method I believe widely used currently in Canada and also in
UK.  I will refer to this as the IPA/TI method, isopropyl alcohol,
tincture of iodine method,  but in the literature, it is more commonly
referred to as the "Medi-Flex" method, because it comes in as a kit
manufactured by a company as a Medi-Flex kit.

	That also is a two-stage procedure where the first stage consists of
applying 70 percent isopropyl alcohol in an up and down motion.  The
second stage is to apply 2 percent tincture of iodine again starting at
the point of needle entry and moving outward in a concentric spiral. 
This should also be allowed to let stand for an adequate time for
drying.

	You notice that in both methods, two stages are involved.  This is
consistent with sort of the general accepted thinking in the surgical
literature where if you apply two different antiseptics, they may work
by two different mechanisms and therefore have a complementary, if not
synergistic effect, in knocking out the bacterial flora on the skin.

	Although that has been a general thinking, first of all, it is not
clear whether that is true, and secondly, it is not clear whether that
thinking applicable to patient care is necessarily applicable to
phlebotomy at blood collection.

	The second point that I would just like to insert at this point is that
tincture of iodine itself is an alcohol solution.  By "tincture," what
we mean is iodine dissolved in alcohol, and to increase solubility of
iodine, an iodine salt is added, something like potassium iodide, and it
is suspended in roughly a 50 percent alcohol solution.  Most typically,
it is the ethyl alcohol for increased solubility.

	[Slide.]

	Now, this method has been a recent challenger and the reason for that
is because of two out of the three available studies on donor arm
preparation for blood collection.

	The first of these studies was performed by Goldman et al.  It is
entitled, "Evaluation of Donor Skin Disinfection Methods," and it
appeared in Transfusion in 1997.

	The second of these studies was performed by McDonald et al in the UK. 
The Goldman study was performed in Canada.  McDonald's study is
entitled, "Evaluation of Donor Arm Disinfection Techniques," a very
similar title.  It appears in Vox Sanguinis in 2001.  Both of these
studies focused on the IPA/TI method, in other words, the Medi-Flex
method.

	The third study did not address the effectiveness of the Medi-Flex
method, however, it is a study of a somewhat larger scope and relevant
to this discussion, and again only the third available study in this
area, so I included it here.

	That study is entitled, "Impact of Donor Arm Skin Disinfection on the
Bacterial Contamination Rate of Platelet Concentrates."  It also
appeared in Vox Sanguinis in 2002.

	Each of these studies recognized the previous study.  Dr. McDonald
built on the results produced by Dr. Goldman, and Dr. Lee built on
results obtained by Goldman and McDonald although he did not study the
Medi-Flex kit per se.  Dr. Lee's study was performed in Hong Kong.

	[Slide.]

	To describe these studies in a little more detail,  Dr. Goldman
compared four methods in three paired experiments.

	The povidone method, which is the AABB method,  that method was used in
all three experiments as the common comparator to which the next three
methods were compared, the first being the Medi-Flex IPA/TI, the second
being a sponge followed by an ampule application, both of those
applications involving 0.5 percent chlorhexidine in 70 percent
isopropanol, and the last method being green soap followed by 70 percent
isopropanol.

	I might as well just point out that green soap is a method recognized
in the Technical Manual by the AABB at this point as a method to use if
donor proves to be allergic to iodine.

	Dr. Goldman transferred the residual skin bacteria after arm
preparation to culture plates by direct skin contact, so this was not a
study about actual units collected or any kind of a clinical study.  She
simply enumerated bacteria in colonies appearing on culture plates after
that culture plate has been directly pressed onto the donor's skin after
arm preparation.

	[Slide.]

	These are the results that she obtained.  In the first of these
studies, a comparison between the AABB method, povidone-iodine, and the
Medi-Flex method, IPA/tincture of iodine.

	She did not produce a quantitative estimate of the relative
effectiveness, but more of a qualitative result in that the Medi-Flex
method resulted in a significantly higher number of procedures where the
bacterial colonies, residual bacteria as measured by colony count was
reduced either to zero or 1 to 10 in a significantly higher proportion
of donors than with povidone-iodine method, the AABB method.

	Conversely, the percent of donors with a high residual bacterial count
was associated with the AABB method in a higher percentage of donors
than with the Medi-Flex kit.  So, this gave you some indication in a
qualitative sense that the Medi-Flex kit method might be more effective
than the currently used AABB method in the U.S.

	[Slide.]

	In a similar way, she compared the other two agents to the povidone
method, and in somewhat smaller studies the comparison between AABB
method and the Medi-Flex method was performed in 126 subjects with a
high statistical significance.

	The povidone method was compared to the green soap, in other words, the
AABB standard method was compared to the AABB back-up method, and that
also indicated that the standard method is more effective than the
back-up method with a high level of significance.

	When the chlorhexidine method was compared to the povidone method,
statistical significance was not achieved, and she concluded that the
two methods are about comparable.

	[Slide.]

	So, based on these results, Dr. McDonald performed the next study,
which also concentrated on the Medi-Flex kit.  In this study, five
techniques were compared, actually, five techniques were expanded to 12
variations.

	The isopropanol/tincture of iodine method had four variations
associated with it, and then also the standard AABB method was compared.
 The povidone-iodine method followed by 70 percent isopropanol was
compared to it.  A Cliniswab Alcohol method, which is a one-step method
involving 70 percent isopropanol, that was studied.

	Then, the North London method, which happened to be the prevailing
method up to the point of performing this study, was also studied, and
that involved applying 0.5 percent chlorhexidine in a 70 percent alcohol
solution.

	After arm preparation, instead of direct skin contact plating of
culture plates, the investigator used moist saline swabs to transfer the
residual skin bacteria from the prepared donor arm to the culture
plates.

	Which of these enumeration methods are better, it is difficult to say.

	[Slide.]

	These are the results that were obtained by that study.  The Medi-Flex
adapted method, and by "adapted," it is adapted because the second stage
application of tincture of iodine is applied in a straight up and down
motion rather than a concentric spiral, and that proved to be or at
least appeared to be the most effective, resulting in 99.8 percent
reduction in 29 subjects.

	The next best was the povidone method or the AABB method which had an
effectiveness of about 90 percent reduction, and this was comparable to
iodophor application followed by alcohol of again approximately 90
percent effectiveness.

	The Cliniswab alcohol method was close behind with 87.4.

	Now, it is difficult to say if the differences among these results are
all that significant, particularly among the last three.

	[Slide.]

	Based on that initial study, a higher number of subjects were selected
for a more in-depth study, comparing the existing North London method at
that UK blood center to two variations of the Medi-Flex method which
appeared to be promising.

	One is the adapted method where the tincture of iodine is applied in
straight up and down motion rather than spiral as the kit originally
intended, and another variation being the IPA/TI Medi-Flex method as a
two, double alcohol application variation where the isopropanol
component is applied twice rather than once.  You would expect that to
be a high performer, as well.

	The results actually indicated that applying the alcohol once is as
good as applying it twice, maybe even better, and applying the tincture
of iodine suspended in 50 percent alcohol is applied in a straight up
and down motion rather than in a concentric spiral.

	Again, the significance of these results is difficult to make a
statement about, but it seems clear that both of these variations are
more effective than the then existing North London method, which only
reduced to about 78.5 percent of the donor skin flora.

	[Slide.]

	So, these are very suggestive results and basically, that is all there
is.  Because of the lack of data in the blood collection literature,
that became a major challenger to the existing AABB method, but several
points need to be further considered before we widely accept that as the
replacement method.

	First of all, in the ways that the investigators counted the amount of
residual skin flora, the way that they collected the sample either by
moist saline swab or direct skin contact of the culture plate, certainly
it was targeted at identifying the surface, but not the resident
bacterial skin flora.

	It is well accepted that the skin flora consists of basically two
components, a transient component which resides in the skin surface
which can easily be removed by simple hygiene and washing methods, and a
more deeper resident flora which is down in between epithelial cells,
which is difficult to remove mechanically, but has to be sterilized by
an antiseptic method.

	So, the enumeration method is targeted at the surface, a transient
flora only.  Further, it has been a concern that with every phlebotomy
at blood collection,

there is a small core of skin that may be generated by the needle that
is difficult to remove and just stays with the blood component.

	Certainly, that core of skin will contain the resident flora in deeper
layers of the skin which has not in any way been measured by these two
studies.

	[Slide.]

	Even if these counts truly reflected the skin flora levels, it is
difficult to say what it means in the clinical arena.  This is a
laboratory study using culture plates.  What it means in terms of
contamination of the actual platelet units is only to be speculated
about.

	Even if the correlation were to prove to be present with the
correlation between the laboratory results and the clinical outcome, one
has to bear in mind that neither of these studies, at least to my
knowledge, through a close reading of the published articles, has been
blinded in any way, and you might suspect that the care with which the
arm was prepared with a particular agent might have great result on the
results of obtained, as well as the care in setting up the cultures with
either agent.

	So, even though both investigators concluded a high statistical
significance with the results obtained, it is difficult to say whether
the studies had been set up in a way to allow an interpretation of high
statistical significance, so it is not clear how to interpret the
results despite the p-values obtained.

	[Slide.]

	Also, even if all of these prove to be concerns only, and not real
worries to ponder over, you have to keep in mind that there are other
intervention mechanisms being considered, such as the diversion pouch,
such as the QC system, and such as the unit release testing system.

	So, in the context of a multi-pronged approach to reducing bacterial
contamination, what changing from one agent to the other agent means,
that is difficult to know.

	On the other side of evaluating a potential switch from one arm
preparation method to another, what about the down side of things.  The
currently available method seems to be well accepted by donors.  Will
the tincture of iodine be similarly well accepted?

	It causes more of a skin irritation and staining than does the
povidone.  In terms of tincture of iodine, it is more readily available
for absorption into the system at circulation.  Now, where this is
probably not much of a concern in the donor setting, it has been a
concern in the patient setting.

	So, all of these, how will it translate to donor acceptance and
ultimately the effect on donor availability or blood availability, that
has not been addressed at all.

	[Slide.]

	Further, the multiple variations of the Medi-Flex method that Dr.
McDonald studied, it is difficult to say much about the small
differences obtained in the results.

	Furthermore, the tincture of iodine component is a high alcohol
solution to begin with and what kind of results would you obtain with
the same study if you were to include one more antiseptic solution of
simply applying 70 percent isopropanol multiple times?  That is another
question to think about.

	Lastly, the results obtained in these two blood collection studies
conflict with results, parallel results available in the patient care
literature where two investigators, Little and Wilson, studied that
particular antiseptic kit, the Medi-Flex, consisting of the same
components, 70 percent isopropanol and 2 percent tincture of iodine.

	They compared that to the povidone methods and other methods, but
povidone being one of them, for their effectiveness in reducing
contamination of blood cultures at patient sepsis workup, and either the
results were only marginally better with the Medi-Flex kit or actually
comparable as studied in the patient care sector.

	So, these seemingly conflicting results have to be at least reconciled.

	[Slide.]

	Now, those are some critiques on these two studies supporting a
transition over to the isopropanol/tincture of iodine method.

	A third study is available in the blood collection literature, as I
mentioned earlier, and this involved a comparison between 0.5 percent
cetrimide and 0.5 percent chlorhexidine followed by 70 percent
isopropanol application.

	That method was compared to a povidone-iodine method, which is not the
AABB method, but it is a method of applying povidone-iodine with 1
percent available iodine, which is like the prep solution of the AABB
method, but that is used more as a scrub rather than a prep, and the
actual prep solution in this case was the 70 percent isopropanol.

	What they did was just a study, not performed as an experiment, but
tacked onto the actual transfusion service requirements.  Over two,
10-month periods, these two studies were compared one after the other.

	Over these two, 10-month periods, over 170,000 platelet units were
cultured. The samples from these platelet units were injected into an
aerobic bottle of the BacT/ALERT system culture bottle, and results were
obtained that way.

	So, although this does not address the Medi-Flex kit, currently, the
one that has raised the whole point about potentially switching to a
different solution, it does point out that it is possible, readily
doable to generate data that is more applicable to a clinical
interpretation.

	As obtained by Lee et al, the cetrimide- chlorhexidine followed by
isopropanol method resulted in 0.072 percent contamination rate.

	When that was switched to povidone-iodine and isopropanol, in other
words, basically, the scrub component of the method was switched from
cetrimide-chlorhexidine to povidone-iodine, obtaining the same prep
solution component isopropanol, when that switch was made, the
contamination rate fell to 0.042 percent, for approximately a 42 percent
reduction.  But again, I present this only to point out the possibility
of performing more studies in this area.

	[Slide.]

	At this point, I would like to present five points to consider and show
you some examples of supporting evidence from the clinical care
literature.

	Point No. 1 is that the following antiseptics listed here may be
comparable in their ability to reduce skin bacterial flora - 10 percent
povidone-iodine, 2 percent iodine tincture, 2 percent chlorhexidine, 70
percent isopropanol alcohol, or any combination of these, one step or
two step.

	[Slide.]

	Why would this be?  This is based on my reading of the literature.  A
study performed by Calfee et al, the article entitled, "Comparison of
Four Antiseptics in a Randomized Trial," published in the Journal of
Clinical Medical Biology in 2002, a very recent large-scale clinical
study, where four antiseptics were compared in over 12,000 blood
cultures - 10 percent povidone-iodine, 2 percent tincture of iodine, 70
percent isopropanol and povidone- iodine, and 70 percent ethyl alcohol
or the kit called Persist.

	No significant differences were seen in the contamination rates of the
obtained blood cultures using these four different arm preparation
methods.

	The contamination rates ranged from 2.5 percent to 2.9 percent, and the
preparations that contained an alcohol component tended to be more
effective although no statistical significance could be derived.

	[Slide.]

	I list three additional studies here.  The Calfee study I just
described is listed in the first slide, a randomized study in over
12,000 blood cultures, but there are other smaller studies which support
a similar conclusion - Trautner's study in 2002 where 2 percent tincture
of iodine was compared with 2 percent alcoholic chlorhexidine or 2
percent chlorhexidine suspended in a high concentration of alcohol.  In
a blinded, 215 paired blood culture study, there was no significant
difference between the two preparations.

	Wilson's study in 2000, an iodophor/alcohol method was compared to the
Medi-Flex method, isopropanol followed by a 2 percent tincture of iodine
in over 12,000 blood cultures.  Again, no significant difference.

	An interesting study by Shahar in 1990 where this investigator was not
convinced that the arm preparation methods of any kind made much of a
difference, and he compared 70 percent isopropanol followed by 10
percent povidone-iodine, and this is sort of the state-of-the-art at the
time accepted method of collecting blood cultures or maybe you might
call it good clinical practice before obtaining a blood culture sample.

	That method was compared to a method used for just obtaining a blood
sample for laboratory measurement, such as obtaining a CBC, where you
take an alcohol swab, just briefly wipe the patient arm one or two
times, blow on it a bit, and go right to needle insertion.

	He compared the results, after comparing these two studies, in 181
paired blood cultures, and there was no significant difference. 
Something to think about.

	A second point to consider.  Washing with soaps is effective in
removing transient surface skin flora, but has little effect on reducing
the resident flora in the deeper layers of skin.  In fact, the soap
residue, if not completely removed at washing, may interfere with the
activity of subsequently applied antiseptic.

	[Slide.]

	This is based largely on a close reading of an article published by
Lilly, et al, where the article is entitled,  "Detergents Compared with
Each Other and with Antiseptics as Skin Degerming Agents."

	The authors used the word "degerming" to indicate removal of both
surface bacteria and removal of the more deeper layers of skin bacteria
since the surface bacteria is felt to be more of a mechanical removal
and the deeper layer removal being more of an actual microbial killing.

	This was published in the Journal of Hygiene in 1979.  Basically, this
was an enumeration of hand bacteria before and after hand washing with a
particular degerming agent in six subjects.

	The way they measured the residual bacteria on hands was much more
elaborate than was performed by either Goldman or McDonald, where what
was described as a standard method was used.

	A hand was first washed in some kind of basically a saline solution. 
The washing was performed and then also, subsequent to that, a hand
washing was performed in the identical solution and an aliquot of that
solution is then incorporated into a culture medium, into pour-in
culture plates.  Then, the bacterial colonies are enumerated.

	So, it was more than a simple transfer of saline swab, more than a
direct skin contact plating, was much more elaborate and felt to be
superior in that it better measures the overall skin flora, not just the
surface.

	In any case, basically, the investigators compared three types of
degerming procedures - one using a combination of antiseptic and a
detergent, which the one used here was Hibiscrub, which is 4 percent
chlorhexidine gluconate in a detergent base.

	That was compared to detergent alone.  For that, only the detergent
base of the Hibiscrub solution was used, without the 4 percent
chlorhexidine gluconate.  Those two were also compared to an antiseptic,
which was 0.5 percent chlorhexidine gluconate in 95 percent ethanol.

	There are all kinds of critical comments that can be made about the
enumeration methods, and so forth, but the results are rather largely
spread out in that the antiseptic method, which she achieved 96 percent
reduction--I think I switched the numbers here--the antiseptic method
achieved 96.0 percent reduction, whereas, the detergent alone achieved
4.6 percent.  A combination of antiseptic and detergent achieved 81.2
percent.  So, the first two figures under the column Percent Reduction
is switched.  I apologize for that.

	Basically, it tells you that antiseptic is much more effective than
soap, and that is consistent with common sense, but what is somewhat
surprising is that if you were to use a combination of soap plus
antiseptic, it may be no better, and, in fact, probably worse than
applying the antiseptic alone.

	[Slide.]

	This Lilly study in 1979 is consistent with the results obtained by the
Goldman study, which I described earlier, where the results obtained
with the green soap method as compared to the standard AABB method of
using two different concentrations of povidone-iodine, when that was
compared with the green soap followed by isopropanol alcohol, it was
clear that the povidone-iodine method, which did not involve a
detergent, was superior to the green soap method, green soap being a
soap.

	[Slide.]

	A third point to consider is the following.  In the donor setting,
iodine and chlorhexidine may not offer an advantage over isopropyl
alcohol.  These are the reasons why.  Iodine and chlorhexidine is felt
to be advantageous in the clinical care setting because, although it
achieves antisepsis more slowly than does alcohol, it maintains it for a
longer duration.

	So, for instance, if you are concerned about catheter sepsis where the
catheter will remain in the patient for prolonged periods of time, it is
much more important how well maintained the initial site is.  It is just
as important to maintain it as to achieve antisepsis to begin with.

	Of course, the same applies for any surgical procedure.  So,
maintenance of antisepsis is important in clinical care, but in the
donor setting for blood collection, rapid antisepsis is probably much
more important than maintenance of antisepsis since phlebotomy is
initiated and terminated quickly and there is no reason to really
maintain antisepsis, and donors are probably not willing to stick around
for long periods of time, and probably being able to achieve antisepsis
rapidly is probably much more important.

	Most of all, isopropanol is readily available, is inexpensive, and is
well accepted by donors and patients alike.  The same cannot be said
about tincture of iodine.

	[Slide.]

	Two more points to consider.  Repeat application of 70 percent
isopropanol may be more effective than a single application.  This is
contrary to the result obtained by McDonald, et al, where the double
alcohol variation of the Medi-Flex method proved no better and
suggestively slightly worse than the standard method, than the adapted
method where the tincture of iodine was applied in a concentric circle,
in a straight up and down method rather than a concentric circle.

	The second of these two points is that two-page antisepsis is not
necessarily more effective than a single-step procedure.  The reason for
that comes from the following.

	[Slide.]

	Lilly et al also performed a second study entitled, "Limits to
Progressive Reduction of Resident Skin Bacteria by Disinfection," that
appeared in the Journal of Clinical Pathology in 1979, where two
experiments were performed under this study.

	First, the effect of repeat applications of an antiseptic and an effect
on limits to progressive disinfection, where 12 hand disinfections were
performed over four days, bacterial counts were measured after each
handwashing and 4 different solutions were compared - soap, Hibiscrub,
0.3 percent chlorocresol, and 95 percent ethanol.

	Also, a second experiment for evaluating the effect of a two-phase
disinfection.  Six hand disinfections were performed over two days with
95 percent ethanol as the first agent, and then an immediate seventh
disinfection was performed after the sixth with a phase 2 agent.

	The agents compared there were Hibiscrub base and Hibiscrub, 95 percent
ethanol alone, and 0.5 percent chlorhexidine in 95 percent ethanol.  So,
95 percent ethanol was the first scrub, first phase in all of these
methods followed by different second phase methods.

	[Slide.]

	These are the results obtained.  This is for repeat application of an
antiseptic or what the author described as "progressive limit" to
disinfection.

	With 0.3 percent chlorocresol, there was some initial reduction in
bacterial count.  That continued to be true to about seven or eight
washings, but then it leveled off and no further benefit was derived
from continuously repeatedly washing with this solution.

	Hibiscrub, which is again a chlorhexidine/detergent combination,
performed better than that.  There was a more rapid initial reduction
after first hand washing and good additional benefit was obtained by
subsequent hand washing to about six procedures, but then it also began
to level off and no benefit was derived beyond about seven or eight
washings.

	With 95 percent ethanol, achieved the greatest benefit with the first
washing.  Then, although there are some blips there, you generally get
the sense that you got additional benefit from each hand washing, all
the way down to 12 washings.  It is not clear whether the benefits stop
there.  It is possible that additional washings could even produce more
favorable results.

	So, by this experiment, it appears that at least with repeated washing,
95 percent ethanol works best.  Now, the reason for the 70 percent
isopropanol being the most commonly used alcohol based antiseptic rather
than a higher concentration, is that it is a balance between
concentration and volatility, so the higher concentration, the better
antisepsis, but it evaporates on the skin quicker than at lower
concentration and the duration of contact is important for antisepsis,
as well as the concentration itself.

	So, 70 percent concentration appears to be the best compromise between
strength and volatility.  As a single application agent, 70 percent is
most appropriate or most effective, but if you are evaluating multiple
washings, then, a higher concentration could also be used.

	[Slide.]

	The second experiment, which evaluated the role of the two-phase
method, which was generally accepted in the patient care arena, two
phase more effective than one phase presumably because it used different
mechanisms of pathogen reduction.

	Whether or not that is true was looked at in the following way.  When a
Phase 1 solution was used, 95 percent ethanol, you got a certain
reduction.  Then, when it was immediately switched to a second phase
agent, presumably having different mechanism of action, the results
obtained were rather surprising.

	With a Hibiscrub base, which is actually the detergent-only component
of the Hibiscrub solution, the reduction obtained by 95 percent ethanol
was reversed in that the bacterial count actually rose.

	With Hibiscrub or the detergent/antiseptic combination, the results
were largely maintained, but were not improved.  That was also true for
0.5 percent chlorhexidine and 90 percent ethanol. It was maintained, but
not substantially improved.  With 95 percent ethanol alone, you seem to
get a further reduction.

	So, the differences between Hibiscrub, chlorhexidine and ethanol, and
ethanol, they are rather small, so again it is difficult to say much
about that, but it seems clear that a soap is not a good thing to use
after using an antiseptic.

	The authors made the following comments about these results.  They
postulate that there is kind of a balance when you remove the surface
bacteria, the bacteria residing in deeper layers of skin somehow make
its way up to the top, and with repeated washing, you continuously knock
that off.

	An effective solution will knock it off every time used to the full
potential of that particular solution, so if 95 percent ethanol is able
to reduce it to a certain level, it will keep knocking it toward that
level with repeated washing.

	But if you stop using the most effective solution, but use a less
effective solution, then, the emerging flora from the deeper layers of
skin are now faced with a less effective solution, so it is able to
maintain a slightly higher count on the surface.  This seems to be a
reasonable postulate, but whether or not that is true is debatable, but
at least these were the results obtained.

	[Slide.]

	As a summary, I have these five points to consider in evaluating
whether or not the isopropanol/tincture of iodine method is more
effective than the double povidone- iodine method, and/or points to
consider in possibly selecting a most effective, most practical agent
method to use for donor arm preparation and perhaps points to consider
in designing further experiments to arrive at that information.

	First, the major antiseptics appear to be rather comparable.  Second,
the use of soaps may interfere with antisepsis.  Third, in the donor
setting, iodine and chlorhexidine may not necessarily offer an advantage
over simple isopropyl alcohol.

	Fourth, repeat application of 70 percent isopropanol may be more
effective than a single application. Lastly, two-phase antisepsis is not
necessarily more effective than a single-step procedure.

	[Slide.]

	So, with those points in mind, I will present you with this question to
be voted on and discussed about.

	Do available scientific data support preferential use of an
isopropanol/tincture of iodine skin preparation procedure for
preparation of the donor's phlebotomy site?

	I thank you for your attention.

	DR. NELSON:  Thank you, Dr. Lee.

	Questions or comments?

	DR. STYLES:  You mentioned early in your report that the Medi-Flex
system is already being used in Canada and the United Kingdom.

	Is there any data on their relative contaminated unit incidence
compared to ours, and/or is there any data--they must have changed at
some point--if the change for them resulted in any reduction in
contaminated units in their blood supplies?

	DR. LEE:  If it exists, I am not aware of it.  There may be someone in
the room that may be more current on that topic than I am.  It hasn't
been that long that the switch was made in their centers, so the data is
probably accruing, but I don't think those studies have been performed
or published.

	DR. STYLES:  I would just imagine that they would have that same kind
of surveillance data.  I would hope that they would have that, just like
we would, so that you might be able to get some indication in a "real
world" setting of, you know, what sort of benefit that is going to give
you.

	DR. LEE:  I agree.

	DR. NELSON:  Mary.

	DR. CHAMBERLAND:  Do you even know if these data are being collected in
these countries, because I don't think we can assume that it necessarily
is?

	DR. LEE:  That is quite true, no, I don't know that.

	DR. SCHMIDT:  Certainly, some of the complexity of this has to do with
the human element, and you were only able to mention the care in arm
preparation, and investigators of these studies I think don't see.

	It has certainly been my experience in inspecting many, many blood
collections to see often, frequently, no numbers, that the phlebotomist
performs the correct preparation of the site and then, at the last
minute, the index finger goes out to make sure that the vein is still
there.

	Now, these people are doing repetitive things, over and over again, but
each one is kind of different because each site is different, each vein
is different, et cetera.

	If you ask the person who has been doing this for two weeks or two
months why they did it, they will say, well, I just touched the edge of
the site.  If you ask the person who has been doing this for maybe two
years or 20 years, they would say I didn't do it.  In their minds, they
didn't do it because they actually don't know that they are doing it.

	Some of this I think is influenced now by the fact that they may be
wearing gloves, and there is this instinctive thing that although they
are not wearing the gloves for that purpose, that it is giving some
protection.

	I think that those of you who have hospital samples drawn for friends
or yourself, if you pay attention to that, it is a different setting and
they are drawing the sample for a different person, but in the hospital,
this happens very frequently in the outpatient lab.

	I have talked to hospital pathologists about it, and they think it is
fine.  Maybe it's fine, but we are not talking about that.

	The second point I would like to make is with the tincture of iodine, I
think nowadays it comes in prepared individual cellophane-wrapped swabs.
 The old problem, which I would think would exist in developed countries
and maybe some undeveloped states, is the jar of 50 percent
alcohol/iodine sitting out there for hours and days, and sometimes a
week, is certainly no longer 50 percent.  That was one of the reasons
everybody was so happy to see the other things come along.

	The final point I would like to make, because maybe we won't come back
to it, is the problem of pooling the platelet concentrates, of course,
happens because of the timing in the hospital setting, the hospital
blood bank or hospital transfusion service, and those people I don't
think are exposed to many of the things we in this room are exposed to,
and the care and attention given the training of the hospital
technologist person in performing the pooling, how much space, what the
facility is that the hospital administrator has given them to do this
properly is something we don't see, but I can assure you it is not
managed with the same care as we look at the preparation of the drugs,
as we call them.

	Thank you.

	DR. ALLEN:  I know we have got a very packed day, but let me ask one
question and than make a couple of comments.

	I was intrigued as I went through the papers that you provided at the
difference between the observed rate of platelet contamination which
generally is well below 1 percent and the false positive blood culture
rates in the studies cited, which often were somewhere in the range of 2
to 4 percent.

	Do you have any explanation?  I mean I have got several hypotheses as
to why that may be.

	DR. LEE:  I think it depends on the care with which you perform the
procedure, and blood cultures are likely to be obtained by multiple
people whereas, in blood collection, the same trained staff is
repeatedly doing the same thing.

	What is measured is different.  In blood cultures, you are measuring
blood cultures, but in other studies, at least the Goldman and McDonald
studies, you know, they measured something quite different, so the
contamination rates are not really transferrable from one area to the
other.

	DR. ALLEN:  Similar types of hypotheses.

	It has been a number of years since I have looked at this kind of
literature although at an earlier point in my career, I looked at it
fairly intensively.  I am a little surprised at the relative paucity of
data just looking at actual skin culture results.

	There was a little bit presented here and there, but I think that this
difference that you talked about between the transient or the surface
bacteria, which are relatively easily removed and killed, versus the
residual flora is extremely important.

	Another of the issues that hasn't been adequately discussed is the
extent of the scrubbing.  As I remember from earlier literature, too
intensive scrubbing may actually promote the release of some of the
residual flora shortly after the completion of the cleansing process,
which goes on for perhaps a minute or so, and that if one were to sample
immediately after the antiseptic has dried and then 10 to 15 later you
would find actually a very sudden rebound of the release of some of the
deeper residual flora that is there, that if you don't have a residual
antiseptic agent, may not then be killed at that point.

	So, I disagree a little bit with the point that you made that donation
is a fairly short-term process and you may not need a residual agent. 
Certainly, you don't need it to the extent that you do when you have an
intravascular device, but certainly, donation can go on for 15 to 30
minutes or so, and I would just question whether perhaps having some
degree of residual activity may not be preferable.

	DR. LEE:  Actually, apheresis donations go on for hours, so your point
is well taken.

	DR. ALLEN:  I also would have liked to have seen much more study of 2
percent chlorhexidine and 70 percent isopropanol.  It was mentioned in
three of the papers, Calfee, Trautner, and Mackey, and to me, that
certainly seems like a very promising alternative combination.

	Finally, I think we do need some information about donor acceptability
of iodine and chlorhexidine in the donation process, and that hasn't
really been addressed at all.

	DR. LEE:  I don't know if there is much information other than
experience type anecdotal information.  I don't know if there is any
published information about donor acceptance of those agents.

	DR. DAVIS:  In my own practice, which involves indwelling catheters, we
prepare the skin using alcohol first, then, the povidone-iodine, and
then we wipe off the iodine with alcohol again.  That may address some
of the donor concerns about residual iodine on the skin.  It works very
well for us.

	DR. FITZPATRICK:  John, first, I want to thank you for a really
extensive review.  We still seem to be needing to know what the source
of the contamination is, whether it is the skin plug or the residual
flora on the skin.  You addressed the flora on the skin.  Previously, we
have heard about the skin plug as the source.

	In your opinion, which do you think is better to address?

 	DR. LEE:  I tend to look at skin plug as an extension of the flora
depending upon how you define flora. It is well accepted that there is a
surface flora and a deeper flora, and if you make the deeper flora go
pretty deep, then, you have a skin plug.

	So, I think it is probably a continuum and not a separate skin plug
issue to consider.

 	DR. FITZPATRICK:  Just one other.  Back in ancient history, we were
required to culture a number of prep sites monthly in order to meet
quality control criteria, and the recommendations were that you culture
individual technicians, so that you could identify those techs that
prepared the site properly as opposed to those that did not.

	Personally, I found that very beneficial in my sites because we were
able to identify individual phlebotomists who were the core of the
problem and not doing the prep properly.

	I think if we go to root cause, maybe we should relook at what we were
trying to determine, which is who is doing the prep properly and who
isn't, and maybe more of a recommendation to examine the technique of
each phlebotomist might even be beneficial.

	DR. CHAMBERLAND:  I just had a couple of sort of historical background
questions.  As I understand it, currently, what is out there is an
industry standard, namely, the AABB standard, extra skin prep?

 	DR. LEE:  That is my understanding, too.

	DR. CHAMBERLAND:  So, there is no FDA-related guidance in this area?

 	DR. LEE:  True.

	DR. CHAMBERLAND:  With the release of these two studies and with the
Medi-Flex procedure in two countries, adaptation or adoption of these
methodologies, has there been any utilization of this methodology in the
United States or are collection agencies pretty much committed to the
AABB standard?

	DR. LEE:  I see some hands in the audience.  I think Dr. Dodd might be
more--

	DR. CHAMBERLAND:  I guess what I am trying to get is an appreciation of
what is driving this, posing this question to the committee.  Maybe in a
subsequent presentation, this will become clearer, but is AABB, is FDA
signaling that you are going to try and have a role in this vis-a-vis
guidance development?

	I guess I am just trying to get a better understanding of these data
are out there, so what is driving the question to the committee and what
might happen.

	DR. LEE:  I think I understand your question.

	I think it occurred like this.  There has been a number of national and
international meetings about bacterial safety of collected blood.  In
every workshop/conference, skin site preparation is an issue, and in
every one of those settings, these two studies by Goldman and McDonald
are described either directly or by the investigators themselves.

	There has been little critique at these workshops as to why not to
adopt or switch to something that looks better, and there doesn't seem
to be a huge down side.  So, if there isn't a huge down side to this,
and there is some reason to believe that it is more effective, then,
shouldn't we move ahead and adopt this on a precautionary principle
rather than wait until all data are available.

	That is the concern that was raised within the FDA.  We are hearing
information that this is better, we are not hearing much information
about why not to do it, then, shouldn't we move ahead.

	DR. DODD:  Thank you very much.  Roger Dodd speaking right now as AABB
president.

	In fact, the method that is being discussed is not an AABB standard. 
It appears in the Technical Manual, which is recommendations.  But, in
fact, as you will hear later, the AABB is proposing a standard to move
to a tincture of iodine approach with chlorhexidine as a backup.

	At that point, it would become mandatory on the membership to employ
that method, so that may muddy the waters, but it is important to
recognize that povidone-iodine is a commonly used procedure which is
recommended, but isn't listed as a standard.

	DR. NELSON:  We are going to discuss this further in the questions for
the committee.  If there aren't any other questions for Dr. Lee, I
wonder if we could take a break now.

	Other questions or burning comments?  We will come back to this I guess
at the end when we consider the questions.

	Let's take a 20-minute break until 11:20.

	[Recess.]

	DR. SMALLWOOD:  We have a very full agenda and we are grossly behind. 
That is an overstatement.  However, we are going to try to do the best
that we can to move forward quickly, and we are going to be enforcing
the time frames for speakers.

	We know that a lot of you are here because you wanted to participate in
this meeting fully and particularly this afternoon's presentation on
parvo.  I also am aware that some of the committee members will have to
leave, so we will have to make an adjustment and sacrifice, and I will
ask your cooperation.

	Thank you very much.

	DR. NELSON:  Continuing on Bacterial Contamination, Dr. Vostal.

C.  Update on the Diversion Pouch

Jaro Vostal, M.D., Ph.D.

	DR. VOSTAL:  Thank you very much.  I will try my best to sacrifice
myself.

	I will just give you a very brief update on an issue that was discussed
with BPAC about a year and a half ago, and that is the issue of having a
diversion pouch in the blood collection sets.

	[Slide.]

	As has been already mentioned a couple times during the day, the needle
cutting through skin can make a skin plug, and this skin plug could be
contaminated due to a poor skin prep or due to passing through a pocket
of bacteria that is hidden under scar tissue.

	The thought has been that if you could take the skin plug that probably
is in the first couple cc's of the blood that is coming through and
divert it away from the main product bag, you might be able to reduce
some of the contamination rate.

	[Slide.]

	When we were here the last time, we presented two studies that sort of
addressed this option.  One study was done by Steve Wagner, and this was
a model of how a diversion pouch or diversion concept would work.

	What he did was he intentionally contaminated a sample site coupler of
a blood bag.  He then sampled that with a large bore needle and
collected the 7 ml fractions and looked at the bacterial content of
these fractions.

	He observed that within the first 21 ml, he was able to recover about
88.5 percent of the bacteria, and if he collected up to 40 ml, he was
able to recover 95 percent of the total bacteria that he collected.

	So, this in vitro model demonstrated that this concept would be
possible.

	[Slide.]

	The second study we talked about last time was a clinical study done by
Dr. Bruneau.  They were collecting actual blood samples and then they
had a special collection set which had two small pouches, each one
holding 15 cc, and they diverted the initial blood, first, 15 cc in the
first one, and then the second one, and then collected the main product.

	They measured the contamination rate in the bag 1 and bag 2.  They
found out that in 76 out of 3,300 donations, that both bags were
positive, either bag 1 or bag 2 were positive, and that was a
contamination rate of about 2.2 percent.

	They also noticed that the first sample was positive and the second
sample was negative in 55 out of the 76,  so a potential reduction of
contamination of 1.6 percent.  So, they argued that you could reduce the
contamination rate from 2.2 percent to 0.6 percent.

	[Slide.]

	Those were the studies that we discussed last time.  This is a study
that has been published since then,  and it is a study by Dr. de Korte,
and he actually measured contamination rate in standard collection sets.
 He has a relatively large number of units collected.

	He compared the standard collection to a collection where the first 10
cc of the blood was diverted away from the final container.  Under these
conditions, he had 7,000 collections.

	They observed a reduction from 0.35 percent contamination rate down to
0.21 percent, a reduction of about a third was achieved using only 10 cc
diversion.  Of interest was specifically contamination by Staph species.
 In the standard collection, they had a 0.14 percent contamination with
Staph, and if they diverted the 10 ml, they had 0.03 percent
contamination due to Staph.

	This is an actual clinical study that shows that using this diversion
approach, you can actually reduce the contamination rate of the final
product.

	[Slide.]

	So, when we were here last time, we talked about the kind of design
that we would like to see for a product that is coming to the U.S.
market.  We stressed that it should be a closed system, that the
diverted blood is separated from the final blood product by a
unidirectional flow, and this would be usually achieved by kind of a
breakaway closure.

	First, the blood would flow into the bag.  This would then be sealed
permanently.  Then, this would be opened, so the blood can flow into the
final bag.  Finally,  that the volume of diverted blood would be
sufficient to achieve the potential benefits that were sort of suggested
by those clinical trials.

	[Slide.]

	In summary, what we discussed last time, we came to the conclusion that
there do not appear to be any negative aspects of using a diversion
system to collect blood.  The preliminary trial suggested that a
diversion of a small volume of blood away from the main storage bag may
be beneficial in decreasing the contamination rate.

	An additional benefit could be of using the diverted blood for testing,
and this could save units that are lost to inadequate sample collection
at the end of phlebotomy.  For example, if you collect a full unit and
you lose the venous access at the end and you cannot collect the testing
samples, that unit will usually be discarded, so collecting those
samples upfront may be able to avoid this problem.

	[Slide.]

	The question that the FDA asked the BPAC Committee back then was
whether manufacturers could claim a significant reduction in bacterial
contamination of the blood product if the diversion pouch was included
in the collection set.

	The committee concluded that the available data did not support such a
labeling claim, however, the committee supported the FDA position that a
diversion pouch would be beneficial because of the potential reduction
in bacterial contamination and reduction of lost products due to
inadequate sample collection.

	[Slide.]

	So, to bring you up to date where we are right now, several
manufacturers have submitted to the FDA blood collection sets.  They are
modified with a diversion pouch. These submissions are currently under
review and we are hoping to clear these in the near future.

	Thank you very much.

	DR. NELSON:  Thank you.

	Questions?

	DR. ALLEN:  Of the manufacturers that have submitted products that are
under review, do they provide clinical data that document the reduction
in bacterial contamination similar to the studies that you presented, or
is that proprietary information?

	DR. VOSTAL:  They actually do not provide clinical data.  They provide
the design that we suggested, and we have taken that as being
sufficient.

	DR. NELSON:  Thank you.

	Dr. Williams is next.  Quality Control Approaches for Detection of
Bacterial Contamination.

D.  Quality Control Approaches for Detection

of Bacterial Contamination

Alan Williams, Ph.D.

	DR. WILLIAMS:  Thank you.

	[Slide.]

	I think in understanding this situation, there needs to be some
understanding of the current environment. As mentioned, bacterial sepsis
is recognized as a second leading cause of transfusion-related
fatalities.

	Now, with the availability of automated culture devices which are
cleared for quality control testing as early as 24 hours after
collection, there is sort of a tug between trying to put out the safest
products possible, as well as stay within the labeling associated with
those cleared products.

	It has been well publicized that there are industry initiatives to, in
fact, test all in-date platelet products for evidence of contamination.

	With the apheresis platelets, the cleared products will allow this to
be done with the semi-automated systems as soon as 24 hours after the
product is collected and have a culture result possibly before issue, so
as not to hold up the product and make it done on a pre-release type
situation, there potentially are mechanisms for retrieving that product
should a problem be found with it.

	As mentioned earlier, random donor platelets are a little more
difficult or a lot more difficult because of the pooling procedure which
takes place at the transfusion service and necessarily any monitoring of
these products needs to be done by the less sensitive methods.

	[Slide.]

	In considering the first two issues, the first is just to ensure that
no harm is done by undertaking quality control schemes that may be
statistically based or might be done universally on all products.

	Sampling of in-date platelet components for culture requires use of
either a closed system, i.e., an integrated satellite bag which one can
then clamp off and use to collect the sterile sample or a tubing weld
made through the use of a sterile connection device.

	Sterile connecting devices are considered functionally closed systems
and obviously, with the platelet component, if you draw the sample, you
want to be able to maintain a five-day shelf life.

	[Slide.]

	Now, the sterile connection procedure itself, there are a lot of data,
particularly in the U.S. supporting the sterility of that procedure when
the weld is, in fact, intact.  This normally would include as part of
the blood center's SOPs a visual inspection for leakage of the weld
joint.

	The data that exist include the original data submitted for the device
review.  A study, which will be described by Dr. Aubuchon in a few
moments, published in Transfusion, and I think an observation, although
not specifically reflecting data collection, an acknowledgment that the
sterile docking procedure is, in fact, commonly used on platelet
products because when a platelet product is split, a sample needs to be
drawn to actually count the content of the splits.

	It would be potentially feasible to look at contamination for split
platelet, apheresis platelet products versus non-split products, but I
have not seen data that looks at that, but, in fact, the procedure is in
place for a room temperature product.

	[Slide.]

	The reason for bringing the subject to the attention and a vote is, in
fact, there is a single study of pooled platelet concentrates that
reported 15 of 1,105 contaminated units among pools that were connected
by tubing welds.  As a control, they looked at cultures on 378 apheresis
platelet concentrates.

	The 15 contaminated units, in fact, they went back to the buffy coats
of those products and did not find evidence of contamination, and the
study concluded that quite possibly those contaminations were due to the
sterile docking device itself.  This was published in 1997 from the
Belgian Red Cross.

	I think subsequent speakers including Dr. Aubuchon and probably the
Terumo speaker will have further comments about that study.

	I think the bottom line in terms of policy development is that any
extrinsic contamination rate of this magnitude would clearly negate any
benefit of large-scale culturing.

	[Slide.]

	The other aspect of quality control is to identify strategies that
facilitate the direct reduction of bacterial risk given that no current
detection systems are approved for pre-release testing, and balance that
with what is typically considered quality control, reducing bacterial
risk by assuring that blood collection and processing procedures conform
to a defined standard of some sort.

	[Slide.]

	There is a proposal currently under consideration by the Council of
Europe, and I believe public comments have been received, and
modification of this represents FDA's proposed current thinking on a
statistical quality control procedure.

	What this would involve, would be at least 5 percent or depending on
facility size, a minimum of 1,500 platelet products annually are subject
to quality control testing for bacterial contamination at 24 hours or
later when the product meets a labeling criterion for which one of the
automated devices could be used, that is a possibility for random donor
platelets or products that could not be used within the labeling
requirements, it might be reasonable to use outdated products or other
criteria to produce the culturing to meet this quality control method.

	The second portion of this is that standard statistical methods should
be used to identify significant deviations from a baseline contamination
rate, and we are proposing that baseline contamination should not exceed
0.2 percent.  So, in other words, 0.2 percent is the standard. The
implementing facility needs to predefine a scheme which will establish a
trigger point when that standard is surpassed on a statistical basis.

	[Slide.]

	The chosen method should be based, as I mentioned, on a predetermined
level of confidence to exclude a maximum tolerated contamination rate,
and an action limit should be established.

	Now, there is an example in the handout that you received.  I am not
going to go into that in detail, but basically, the scheme that is laid
out is that this represents the activities of a small collection
facility that is doing the 1,500 cultures per year.

	In the course of the year, if they realize three cultures for 400
attempts at culturing, it would, in fact, put the background prevalence
of contamination over the action level, but the statistics wouldn't be
supportive of the fact that this sample accurately represents that level
of contamination in the entire process.

	However, if that site designated that within the annual sampling
period, they found 7 positive units, this would not only constitute an
action level that was exceeded simply based on the prevalence of the
contamination, but also statistically within defined confidence and
power limits would be shown to exceed that 0.2 percent standard.

	This is based on a binomial distribution.  Other statistical
considerations may apply, but I think this is one fairly straightforward
way to reach that sort of statistical control scheme.

	So, the bottom line take-home message is a facility would be required
to test either 5 percent or 1,500, whichever was greater, assign
inaction limits on a statistical basis that would call for investigation
and revalidation if that limit was exceeded.

	[Slide.]

	A second element of that is as with any good quality control system,
any instances of a positive culture should be investigated promptly to
facilitate identification of a correctable cause.  As discussed earlier,
this could be related to arm preparation procedures, an individual
phlebotomist who had unacceptable technique, or things that might be
occurring in the component preparation laboratory.

	Simply by trending some of these results, one might get a clue as to
what problem exists in the processing procedures.

	There are other actions which are not specifically part of current
thinking, but, for instance, if culturing is being done and a product is
released, there are issues related to whether the individual
contaminating bacteria should be identified, whether sensitivity testing
should be done, and whether the recipient physician should be notified.

	Clearly, this has implications if the product has been received.  Some
of this is easier to decide than if it is to be done on a routine basis
without knowing the actual disposition of the final product.

	[Slide.]

	The FDA recommendation regarding quality control would incorporate
examples of what would constitute an acceptable quality control
strategy, such as what was mentioned earlier for a small facility, could
also give examples for large facilities which would be done on a similar
statistical basis.

	The current thinking is that the FDA recommendations would not inhibit
what might be occurring at the industry level, but would actually serve
as a minimal standard and that industry standards may well be defined to
be more stringent.  As you will hear, some of this is currently being
proposed by the American Association of Blood Banks for culturing at a
somewhat higher level.

	So, FDA's thought is just to institute a minimal standard that can be
exceeded.

	That basically outlines the quality control issues.  Again, you will
hear more about the European study and the sterile connecting device
from the next two speakers.

	DR. NELSON:  Questions or comments?

	DR. ALLEN:  Since the committee will be considering the question about
the proposed statistical approach, do you have any other information
from blood centers or other published literature that suggests that that
kind of quality control mechanism is useful in terms of keeping the
contamination rate low, that it is a good check on procedures in use at
all steps of the collection and processing?

	DR. WILLIAMS:  Within the blood center environment and specifically
culturing related to contamination, I am not aware of practical data.

	I know to suggest a statistical approach to quality control, this is
the first attempt to do this was with respect to leukoreduction and
residual white cell content, and it has, in fact, been a rather
difficult scheme to reach balance between what is practically possible
in an individual collection site and what would meet a statistical
criteria.  So,  I think those are considerations.

	As far as the actual practicality of a sampling approach, I am not
aware in a blood center situation.

	DR. ALLEN:  Going back to the other use in terms of the leukoreduction,
how easily has it been adapted by QC staff in blood collection centers
as you have worked with them to implement this?

	I am saying this simply because for most physicians and others who
aren't coming out of a manufacturing background, this kind of
statistical quality control, it is a foreign concept, and is part of the
issue one of education and training, do current QC managers at blood
collection centers have that kind of skill and facility, or what are
some of the obstacles here?

	DR. WILLIAMS:  I think my observation would be in most circumstances,
probably not.  The best way to approach it is in the context of FDA
guidance to propose schemes that the FDA would consider appropriate, and
if a center has more sophistication, can make other quality control
approaches that might be distribution dependent, for instance, they can
propose those and have them reviewed for prior approval.

	But it appears that the best way to approach the issue is to be as
prescriptive as possible in defining the simplest system to be put into
place and work with that as a minimal system.

	DR. NELSON:  Did you want to make a comment?

	DR. KUNERT:  Yes.  Matt Kunert at the CDC.

	First, I just wanted to say I think any step forward is a good step.  I
do have a question about what essentially looks to be a benchmark of 0.2
percent contamination was determined.

	In my experience with health care facilities and looking at, say,
nosocomial infections, usually, benchmark is determined after looking at
what the overall rate is among a group of facilities, for instance, and
this, I think is more based on previous studies, but I am just a little
concerned that it seems high, 1 in 500 seems a little higher than,
although there is variation, what I have seen in the literature, in some
of the real-time data, for instance,  some data forwarded on from Japan
where they looked at 10,000 units and got 1 positive.

	Aubuchon had a paper where they had a bunch of false positives, at
least that is what they have determined them to be.  But I didn't have
any positives in two years, so I am just a little worried that this
might be a little high to use as a benchmark.

	The other comment I had is considering about those positives that you
might see, might not be those that are pathogenic, and how that will be
addressed, because you mentioned about clinician notification.

	You have something like a corynebacterium species that is very
different from a Klebsiella species, say, in terms of clinical
significance, and whether you might want to consider having different
benchmarks for different clinical significance as the numbers of
organisms that have questionable clinical significance are likely to
overwhelm those that are really significant, and those are the ones you
really want to prevent, like the gram negatives.

	DR. WILLIAMS:  I agree.  Again, I think there is a balance to be
reached between working up those that are clinically significant and
those which, while maybe not clinically significant, do represent some
breach in procedure, and that has value in the workup itself, but
clearly, you need to reach an approach that is realistic and feasible.

	As far as the 0.2 percent, Jay may wish to comment further since this
is part of the European standard.  From my approach, I think it is felt
to be a standard that can be reached and reasonably approximated by a
statistical approach.  If you get too much more ambitious than that, you
simply can't get there with a sampling approach.

	I guess one final comment is that statistical approaches in sampling
really don't come under consideration if you are doing universal
culturing because clearly, then you have the results, 0.2 is the
standard and optimally, you would want to stay within that.

	DR. EPSTEIN:  First, let me just reiterate that the number was based on
a literature review of current practices in competent centers, but also
let me emphasize that if you set a standard of 0.2, a center would need
to achieve something lower than that in order to be able to repeatedly
produce a statistical assessment showing it was no greater.

	So, it, in fact, implies a more stringent actual performance.

	DR. KUNERT:  If I might ask a point of clarification.  Are you going to
be discussing later as far as when, at what time, either during storage
or at out-date, that the products are going to be sampled, because I had
a comment on that, as well, and I can wait on it if that is going to be
discussed later.

	DR. WILLIAMS:  I think it will be probably addressed in some of the
subsequent talks, but it is not a specific decisional issue for the
meeting.

	The devices that are cleared for quality control are cleared for as
soon as 24 hours after collection.  Clearly, if you are simply doing
quality control, not using the product, the best time to sample it is
probably after 48 hours or ideally at out-date.

	Basically, we are looking for harmony between what the labeling permits
and what would suffice as a quality control program.

	DR. KUNERT:  I was going to just make the comment, because it does
relate to the statistical methodology, I wanted to clarify, on the BACON
data, there was a reference in the Aubuchon paper that was based on an
abstract, and I just wanted to clarify that those cases that were
associated with serious clinical sepsis were associated actually not
necessarily with long storage times, I mean they were associated at day
2, day 3, infrequently day 4, overall, true, day 4 or day 5, but the
ones that caused the greatest clinical impact were the ones that
actually grew very quickly.

	So, I would urge that there was a strategy or an option was to look at
culturing at release, that that would be preferable to those done at
out-date because you want to get as many units as possible because the
incidence of these fatalities and these very serious events are very,
very low, so it is important to get as many as possible at the time of
release, I think.  That was my third comment.

	Thank you.

	DR. FITZPATRICK:  I am trying to put this in the perspective of how I
manage from a practical standpoint when I see that I might have a trend
or I have a couple positive units, I am moving toward or above the 0.2
percent range.

	I have data, but now I have to go back and figure out the root cause in
order to make an intervention.  So, now I have to from what I have
heard, I am going to have to culture my welds, I am going to have to
culture my arm preps, I am going to have to look at the training and
methods used by my phlebotomists and my technologists in that.

	I also have to look at the nursing staff and administration.  That also
is going to take time and effort and a lot of work, and I am not sure
that you are going to find a root cause to intervene and make an
improvement on.

	The end outcome that we want here is to try to reduce that 5 to 9
fatalities per year.  I am worried about the focus of the data
collection.  I think the data collection should be focused on an attempt
to find the root cause of the problem and I am not sure we are doing
that.

	I don't have an answer for that, but I think we need to look at the
work and effort involved and is there a way to focus the data collection
on attempting to identify the root cause as to just confirming what we
know, which is that we have a problem with platelet contamination and
that there needs to be a way to intervene.  That is my concern overall.

	The other question is if you look at applying the QC effort to current
practices, and if you evaluate the fatalities that have occurred over
the past few years, one, have you looked at that, and, two, do you think
that there is an impact that would be made on those fatalities by
applying the QC method?

	DR. WILLIAMS:  I think it is a good point.  It would be an extensive
undertaking to look at each of those cases in that context, so it is a
fair question, but, no, it hasn't been looked at specifically.

	DR. SIMON:  Maybe following up on that, asking that a different way,
let us assume you did this, every blood center in the country met this
requirement, is it more likely than not that the 5 to 9 fatalities would
continue.

	DR. WILLIAMS:  Well, I guess I would again go back to the differing
prevalence of contamination between sites. I mean what is being argued
is there is a certain low level of contamination that you can't identify
a cause, and is constant, and you simply need to culture to identify
those units.

	I think where the quality control gets you is where there are
variations between prevalence of contamination between sites and there
are extrinsic factors, that you allow some control over those factors,
so that you can identify them, eliminate them, and reapproach that
baseline level.

	DR. NELSON:  My guess is that if we had data on all blood collection
centers, that they wouldn't be totally homogeneous, that there might be
outliers, and that might be useful.

	DR. SIMON:  Historically, the discovery of this problem came from
well-known academic centers, Johns Hopkins, Cleveland--

	DR. NELSON:  We are probably an outlier.

	DR. SIMON:  Yes, which presumably had no history of poor technique
problems like this.  It appears to me that this problem as it has
appeared around the country is not related to the kinds of root causes
that Dr. Fitzpatrick would be looking for.

	DR. KLEIN:  Since we know that up to half of these are actually from
the donor, circulating in the blood, are not going to be corrected by
correcting the arm prep and probably not effectively by diverting blood
either.

	This is an approach that I believe we are trying to take to do
something, and I am not against it.  What we would really like is a
release criterion.  The more we culture, the more that we will eliminate
because they are positive, but what we are left with is an in-process
test which is not ideal, and I think we simply have to recognize that
and move on.

	DR. EPSTEIN:  I think a distinction has to be made here.  The proposal
for a quality control strategy is not a proposed method for eliminating
the residual risk.  It is a strategy designed to ensure that all centers
are operating in compliance with current standard procedures.

	What we are reacting to is the finding that if you start culturing,
there have been reports in the literature of as much as 2 and 3 percent
positive cultures, and we think that if all appropriate procedures are
followed, that shouldn't be happening.

	So, we are really not suggesting that this is the cure for the current,
you know, residual rate of sepsis and fatality, but we want to at least
assure that all centers are able to demonstrate that they are operating
to current standards.  So, they are separable issues is what I am trying
to argue.

	DR. NELSON:  Thank you.

	Next, is Dr. Aubuchon, Experience with Plastic Tubing and Universal
Bacterial Culturing.

E.  Experience with Plastic Tubing and Universal

Bacterial Culturing

James Aubuchon, M.D.

	DR. AUBUCHON:  Thank you very much and I appreciate the opportunity to
address the committee again on this subject.  This is an area that we
have been involved with, with various research projects over a number of
years.

	[Slide.]

	Clearly, as you have heard this morning, there are many uses for
sterile connecting devices in conjunction with platelet units,
particularly for QC sampling, to dock on filters in order to leukoreduce
the unit, to pool either before storage, as is done in Europe, or after
storage, as is done in this country, to sample for bacterial detection,
and to remove an aliquot for transfusion to a neonate, for example.

	[Slide.]

	Units can become contaminated in a number of ways and certainly the
welding or the sterile connections that are conducted on the unit is a
potential site for contamination.

	[Slide.]

	For those of you on the committee who have not ever used one of these
devices, let me just walk you through briefly how they operate.  This is
my attempt to explain how it operates, and I don't understand all the
inner workings.  Possibly someone from Terumo can give you the details.

	But the two pieces of tubing that are to be welded are placed in a
metal chuck adjacent to each other.  Beneath that chuck is a copper
wafer.  It looks like just a piece of copper, but there is actually an
integrated circuit inside that piece of copper.

	Before the instrument can be used, the prior wafer has to be ejected,
so you are using a new wafer each time. The wafer then heats and comes
up and slices through by melting the tubing.  As the two pieces of
tubing have been melted, seen here from on top, then, the chuck moves. 
The chuck is actually in two halves, and this part of the chuck moves
backwards, so it drags this piece of tubing and aligns it with this
piece of tubing.

	The outcome then is a new piece of tubing that is connected together in
two pieces which are discarded.  The weld then has to be opened by
squeezing it with your finger after you remove it from the chuck.

	[Slide.]

	There is another device on the market by Haemonetics.  I have not used
it, I am not familiar with how it works.  I presume it is something
similar, but I don't know the details of that.

	The device which we have been using most widely in this country and our
laboratory has experience with was initially marketed by duPont and is
now marketed by Terumo.

	[Slide.]

	To show you some pictures of how this works, you see the two-piece
chuck opened here with the two pieces of tubing.  Here is a platelet
unit with tubing coming across the chuck, an aliquot bag over here that
we are going to weld sitting in the other set of slots.

	[Slide.]

	First, the wafer from the previous weld is ejected by moving this
handle forward.  Here, it is coming out.  You remove that, and a new
wafer automatically comes into position at that point from the cartridge
of wafers.  The wafers are designed to be used only once.

	[Slide.]

	The wafer then heats, comes up, cuts the tubing. You see here that this
part of the chuck has moved backward, now aligning this platelet unit
tubing with the aliquot tubing and opening it up.  You see the two are
now connected.

	[Slide.]

	We became interested in this several years ago while doing some
research in bacterial contamination.  In the spirit of good
manufacturing practices, good laboratory practices, we sought
information to validate that the sterile connecting device did what it
was supposed to do sterilely.

	We did find one piece of information in the literature using spore
contamination.  We proceeded then to do some testing with some real life
stressors, and I will also comment on the European study that was
mentioned in comparing the various culture results.

	[Slide.]

	What we found in the literature was presented at the American Society
of Microbiology in 1983, and it took a fair amount of hunting to find
this.  What these investigators did was to contaminate a segment of
tubing with spores of Bacillus circulans and then connect that piece of
tubing with another segment of tubing that had been filled with
trypticase soy broth.

	The weld was opened and the broth was allowed to cover the area of the
weld, and it was maintained at 35 degrees for four days and then
cultured in order to see if any of the spores had gotten into the
interior of the tubing.

	They conducted 758 welds on 17 different instruments and all of these
experiments showed the contents of the tubing after welding to be
sterile.

	[Slide.]

	We conducted a study in three different phases.  I should note that the
study was not funded by the sponsor.  We did this in our own laboratory
just to show that the technique did indeed work.

	In the first phase, we used a leukocyte-reduced single donor platelet
product aliquotted 4 ml each into 64 small bags.  In the second phase,
we used trypticase soy broth aliquotted at 5 ml into 80 small bags.

	These bags were then joined with other empty bags after having dipped
the tubing for both halves of the welding sides in a liquid suspension
of either Staph epi, Flavobacterium odoratum, or E. coli, and note that
the concentration of bacteria in this aqueous suspension was 40,000 to 3
million bacteria per ml.  This wasn't just a small amount of
contamination, this was heavy contamination.

	A total of 10 ml between the two bags was then created.  The weld was
opened and the contents were moved back and forth several times in an
attempt to pick up any bacteria that may have gotten into the weld.  A
culture was conducted immediately and also after seven days.

	[Slide.]

	The tubing itself was not always dry as the manufacturer would expect
the instrument to be used.  In some cases, the tubing was allowed to be
wet and in other cases the contamination was allowed to dry before the
dock was conducted.

	In some cases, the tubing was filled with the trypticase soy broth or
the platelet unit, in other cases it was empty, so we had multiple
combinations of approaches here in these different phases.

	[Slide.]

	In the third phase of the study, we used a trypticase soy broth bag
that we docked repetitively to a series of empty bags, 100 times in all
using I believe 10 different initial bags here.

	Each time we docked on another small bag, we were lengthening this
tubing and the broth was moving through successive weld sites in order
to again stress the system and to pick up any contamination that may
have occurred.

	Here, we used the Staph epi or enterobacter solution of bacterium at
100,000 again to 3 million per ml as the contaminant and using either
wet or dry tubing.  So, we felt that this was really a stress of the
system far beyond what would be seen in normal practice.

	[Slide.]

	The results are shown here.  In the first phase, we performed 64 welds,
63 of them were evaluable.  We found that we had actually contaminated
one of the bags with a different organism.  This organism did not come
from the weld, so we had 63 units that could be evaluated.  All were
sterile.

	In Phase 2 of the trypticase soy broth, two of the welds were
incomplete.  That is, when we took it took out of the chuck and went to
open it, we could clearly see that the weld was defective and was not
complete.  That is not surprising when you are trying to weld wet
tubing.  It doesn't always melt and reseal properly.

	So, of the 78 evaluable, complete welds, all 78 were sterile.  In Phase
3 with the multiple lengthening of the tubing, trypticase soy broth, all
100 welds were complete, evaluable, and were sterile.

	So, we had a total of 241 evaluable welds in all were sterile.

	[Slide.]

	What does this really mean in terms of safety?  We would have loved to
have done 10,000 welds, but my techs were about ready to hang me after
doing 240.

	Looking at this statistically, with 241 negative welds, we can say that
this implies that the rate of positivity is not greater than 0.004, or
if you combine these 241 observations with the 758 published previously,
it means that the rate of the weld not being sterile is not greater than
1 in 1,000.  So, we are able to document then that the rate, if you have
a successful weld, the rate of contamination does not exceed 1 in 1,000.

	[Slide.]

	Just to tell you some experience about how frequently welds may not be
complete, I pulled data from the first 10 months in our Transfusion
Service of this year.  We performed 5,636 welds.  Each one of these is
documented by unit number and documented that the tech has looked to see
that the weld is indeed complete and not leaking.

	There were 4 failures in those 5,600 welds or about 1 failure every
1,400 welds, so this has not occurred very frequently, but it does occur
often enough that we do indeed need to look at it, as the FDA guideline
indicates that we should.

	In a Belgian study we will be talking about in a minute, they noted
that the failure rate of welds was about 1 in 3,000, so it is not a
common problem and it is important then to look at the weld carefully
when you go to open it and to check for leaks.

	[Slide.]

	Now, we have been using this welding technique in many ways, but
certainly as part of our ongoing study of using bacterial culturing
routinely on all of our units of platelets.

	We use apheresis products at our center and on day 2 we perform a
sterile weld with a small aliquot bag and move some of the platelet unit
into this bag, which is then removed by heat sealing and entered by
syringe and needle, transferring 5 ml into an aerobic culture bottle of
the bacT/ALERT system, and placed in the bacT/ALERT cabinet.

	The units are available for release at any time that they are needed
for transfusion unless or until we receive a report from the
microbiology laboratory that indicates that there is something growing
in that unit, and then obviously we quarantine it.

	[Slide.]

	In the first three years of doing this, we are now at about 3 1/2
years, but in the first three years we cultured almost 4,000 units in
this manner.  We found 23 that were initially positive, 14 could not be
confirmed on repeat culture, and 5, we didn't have anything retained to
culture.  That was early on in the protocol.

	It is important to note that all of these which we believe are false
positive occurred shortly after a new tech learned the procedure and
began doing it.

	[Slide.]

	This procedure in our laboratory is performed in the open Transfusion
Service laboratory by all techs in rotation, and there clearly is a
training curve even beyond initially showing that the technologist knows
what needs to be done, and you can see the falling rate of false
positivity over time.

	Dr. Mark Brecher at the University of North Carolina has been using
this technique in his laboratory since late February of this year.  He
is performing it more in a research setting using a biological safety
cabinet that is a laminar flow hood, and he tells me that in at least
2,000 units that he cultured, he did not have any false positives.

	So, it would appear that if you take additional efforts to prevent
contamination at the time of inoculating the individual bottles, you may
reduce the false positive rate.

	We did have four confirmed positives in the first three years or a rate
of positivity at about 1 in 1,000, or in this era of low viral risk
where we are expressing risk now as occurrences per million, I would
just note that that is 1,000 per million while we are talking about HIV
and HCV risk where we use fractions per million, both a much larger
risk.

	[Slide.]

	Now, how did we actually determine that some of these were false
positives rather than true positives?  This was done through repeat
culture of the unit or a retained aliquot from that unit.  So, of the 23
units that have initial growth, we did have something that we could
reculture in 18 of those.

	In 14 of those 18 recultures, the units had no growth on repeat
culture.  Four units, we were able to detect the same organism a second
time, and we assume those were the true positives while we called the
other 14 false positives.  One could argue with the attribution, but we
feel this is approximately correct.

	It is important to note also that the time to a positive culture being
reported was very different between these two groups.

	If we could find the same organism again in that unit or an aliquot
from that unit, the initial report was received by our laboratory in
under 14 hours, where in those cases that we feel were false positive,
where we could not identify the organism on repeat culture from that
same unit, we were receiving a report at greater than 24 hours and often
greater than 30 hours.

	So, that would imply again that we are dealing with very different
situations biologically.

	[Slide.]

	Where is this false positive contamination coming from?  We feel it is
most likely coming from either taking the sample from the small aliquot
bag by needle or placing the 5 ml actually in the culture bottle.  Of
course, this is despite cleansing the septum before placing the needle
through it.

	Again, if this were done in a laminar flow hood, we might reduce the
probability of this happening.

	[Slide.]

	Now, it was mentioned earlier that there is a report from the late
1990s, from Europe, suggesting that welding potentially caused bacterial
contamination.  In this study, they looked at 1,100 buffy coat pools,
which were created from 6,100 welds of individual units.

	They found 15 positive cultures from those pooled 1,100 buffy coat
pools, and they went back to the individual units that were involved in
each one of those pools and cultured them.

	In 10 of those 15 occurrences, they identified the same organism in one
of the units, and they concluded that in those 10 circumstances, the
pool was indeed contaminated, they have a contaminated unit as part of
the pool.

	On 5 occasions, they were not able to grow the organism on reculturing
each one of the individual units.  Now, where could those 5 come from? 
It is an important number because that is 1 in 1,200, and this is
approximately the same rate of positivity that we are seeing overall,
and are we indeed contaminating as frequently as we are detecting true
positives.

	Well, I would question where these 5 incidences of growth came from. 
They could have come from a contaminated weld as these authors contend. 
It could also have come from a contaminated culture, a possibility they
did not consider.  It could be that on repeat culture of the bag when
they went back to the initial units, they may have missed the organism.

	[Slide.]

	You might say, well, that is not likely, but let me share a case with
you that occurred several months ago in our laboratory.  We cultured a
unit on day 2, as we usually did, and in 9 hours, we had reported to us
growth in that bottle.  The short time to detection implied to us that
this was a true positive.

	However, we went back to the same unit, which was still in inventory
the next day, and on day 3, we cultured it again, and that culture
remained sterile out to a week after beginning the culture.  So, we were
not able to find any organism on growing it and culturing it again on
day 3.

	This raised a question, was this day 2 culture then a false positive. 
We cultured the unit again on day 6 and we got growth, and importantly,
we found the identical organism.  It was a Staph epi which happened to
have an unusual antibiotic sensitivity pattern, therefore, we were
comfortable in identifying the two organisms found at day 2 and day 6
were indeed the same organism.

	So, with the European study not finding an organism on going back to
the bags, it may have been that they just missed it, and they didn't
happen to take the right milliliters that happened to have the bacterium
present.

	[Slide.]

	I think it is more important that we and the authors of this paper
focus on the 10 pools that had contamination that were noted.  One out
of 110 transfusions that would otherwise have been given were cultured
positive and contaminated with bacteria.  That is a very high number
that even exceeds the statistical rate that Dr. Williams was mentioning
previously.

	[Slide.]

	So, is it safer to weld in culture, which I feel is the question that
is being proposed here.

	The current standard in the U.S. is not to do any welding to do any
culturing because we don't routinely do cultures.  Only a few centers
are beginning that or considering it.

	So, if you were to transfuse, for example, a million units of platelets
or performing a million platelet transfusions, 1,000 units of those
million would carry along bacterial contamination.  That is the current
state of practice in the United States today.

	If culturing were performed with perfect welds, with welds that never
contaminated either the culture or contaminated the unit, and if that
culturing were 90 percent sensitive, which I feel culturing is probably
greater than 90 percent sensitive, but even if you only assume 90
percent sensitivity, the million units with 1,000 of them being
contaminated, the contamination would be detected in 900 of them, and
100 units only would be transfused with the contaminating bacteria.

	In order for the sterile connecting process to decrease safety, if we
were to go ahead and weld and culture, the sterile connecting process
would have to cause contamination at a rate of 900 per million or 1 in
1,100, and we already have data that shows that the rate of
contamination in welding, even in very stressful circumstances, is less
than 1 in 1,000.

	[Slide.]

	So, I think the answer to the question is yes, it is safer to weld and
culture than not to weld at all, and I will not quote Shakespeare.

	[Slide.]

	So, there are several alternatives that I think could be considered in
approaching this problem.  The European practice is to pool, what they
use usually is buffy coat platelets, but to pool the platelets on day 1
to leukoreduce them at that time by attaching a filter, and to draw a
culture at that point, and to put the units into inventory.

	[Slide.]

	Another approach would be to culture on day 1 or day 2, sometime after
an initial period to allow the small inoculum to grow up to be
detectable.  We are now talking about the common American practice of
using individual platelet units from whole blood units, thinking of 6
units that would be transfused to a patient, we would have to aliquot 6
individual units and create 6 individual cultures, and to store those
units individually because the FDA does not allow us to pool and then
store the units for a prolonged period of time.

	This would be very expensive, because the culturing cost would be
multiplied by 6 and you would have to keep those 6 units together to
make sure that they were going to the same pool ultimately, I would
think, as well.

	Another approach would be to take aliquots from 6 units and put them in
one culture and then store them as separate units, another possibility.

	In any of these cases, you are going to be taking a substantial volume
from the individual platelet concentrate units.  If one needs to take a
volume to culture, it needs to be an appropriate volume to detect the
bacteria, so we are probably looking at, at least 2 ml, if not 5 ml from
each one of the bags.  These bags are, in general, about 50 ml, so that
is 10 percent reduction in efficacy of the platelet transfusions.

	Of course, you could perform these cultures using sterile connecting
devices or have a pre-attached sampling bag on a platelet pooling bag,
and most blood bankers I think in this country would be very happy to be
able to pool and then store.

	It would take the pooling out of the hands of the hospital, allow it to
be conducted in a standardized fashion, matching essentially the
European practice, but possibly the idea of using an additional
pre-attached sampling bag might be another approach that manufacturers
might want to consider.

	[Slide.]

	So, in my way of looking at the world, I think this problem indeed
warrants intervention although it is relatively infrequent, say, 1 in
1,000, it really is too infrequent to be appropriately addressed by a
statistical quality control approach.

	It is certainly large enough to warrant us doing something about it,
and I believe that bacterial detection can be accomplished with an
overall reduction of the risk of platelet transfusion.

	Thank you very much.

	DR. NELSON:  Thanks, Dr. Aubuchon.

	Toby.

	DR. SIMON:  I assume with the culture on day 2, so that is at 48 hours,
is that right?  You can assume the day of draw day zero?

	DR. AUBUCHON:  The day of draw is day zero.  We usually culture
mid-morning, so I would say that the shortest time period is probably
about 40 hours, and some units may be out to 50 hours by that point.

	DR. SIMON:  I wanted to just make sure I understand the data.  It is
still true that you have a higher rate of false positives than you do
true positives, and those units would be interdicted in your system and
not transfused.

	DR. AUBUCHON:  That is correct.

	DR. SIMON:  And you also have instances where you examine the welds and
find them to be defective, and those units would not be transfused.

	DR. AUBUCHON:  I don't know if you saw the technique that we use at our
institution where we clamp off the tubing, and we do not release the
clamps on either side of the new weld until we document that the weld is
a good weld.

	DR. SIMON:  Okay.  So, if the weld is defective, you do not lose the
units.

	DR. AUBUCHON:  That is correct.  You seal it off and do another weld.

	DR. SIMON:  So, you just have to have a good examination procedure. 
Have you calculated the loss to the country in units from the false
positivity?  It would be greater than the ones we are interdicting for
true positivity, right?

	DR. AUBUCHON:  Yes.  The predictive value of a positive is not very
high because we do have a number of false positives.  We are running now
a false positive rate of about 1 in 500, and the true positive rate of
about 1 in 1,000.

	DR. SIMON:  So, for every true positive, you would have two--

	DR. AUBUCHON:  I have two false positives, that is correct.

 	DR. SIMON:  Thank you.

	DR. FITZPATRICK:  Testing for pre-release is, of course, what we want
to do, and that is what you are doing. The question I have is, has
anybody looked at the impact on the whole of the inventory in the
country on wastage if we are holding products until the third or fourth
day to transfuse?

	DR. AUBUCHON:  Certainly, the longer that you hold the platelet before
distributing it to the hospital, the shorter its useful life span and
the potential increases for outdating, that is certainly true.

	I believe that most facilities are not labeling until sometime late on
day 1 anyway because it takes that long to get the nucleic acid testing
results.

	So, with the current approach as approved by the FDA for use of the
bacT/ALERT or the Pall BDS, where 24 hours after collection has to
elapse before drawing the sample, that could probably be obtained
without any delay to preclude release.

	Now, a unit could be released before a result was obtained in the Pall
BDS system 30 hours later or certainly would have to occur before the
final culture was reported out five or seven days later from the
microbiology laboratory as being negative.

	Blood centers develop systems for notifying hospitals quickly in case
of NAT positivity in Phase I of the NAT IND clinical trial, and that is
the common practice in Europe, as well, where as soon as a positive is
found, the hospital is contacted.

	It doesn't happen very often obviously even at a false positive rate of
1 in 500, so it is not like the blood center is going to be calling up
two dozen hospitals every day to recall units of platelets, but a
retrieval mechanism probably would have to be instituted in order to
allow platelets to be distributed at the normal time in order to prevent
an increase in the outdating.

	Of course, if we are able to convince the agency of the wisdom of
culturing and then extending the storage period to seven days, something
this committee considered at a previous meeting, then, if we were to use
one of those two additional days by holding the unit in the blood center
until we got the final result, it would simplify the system without
causing an increase in outdating.  Sorry to editorialize.

	DR. CHAMBERLAND:  In the Merten's paper, looking at their methodology,
they apparently used both new and reused welding wafers, which is a
difference compared to the methodology that you used.

	Now, this was published in 1997, so maybe over time these wafers, it
has been determined that they should be single use, and they indicated
that of the 15 contaminated pools, six were made with new and nine were
made with reused welding wafers, and this was not a statistically
significant difference.

	I am quickly skimming, but they don't indicate in the five contaminated
units if a reused wafer was used for those five units, and I was curious
as to your thoughts about the role that reuse of wafers might
potentially have played here.

	DR. AUBUCHON:  I don't have any data directly addressing that.  I
certainly do know that reuse of wafers is not according to
manufacturers' directions, and the manufacturer has always stipulated
that the wafers are to be used once and once only.

	There are other blood centers outside the United States that do use
them more than once.  They clean them and reload the little cartridge
packs, but that does not meet the manufacturers' requirements.

	DR. NELSON:  A brief comment because we are way behind.

	DR. KUNERT:  Okay.  Matt Kunert, CDC.

	You had some data here that wasn't in your excellent paper.  I guess
you went for two years and didn't have a true positive, and then in the
third year, had four positives.  I wondered if there was a difference
between the organisms in the true positives and the false positives.

	My other question is I didn't quite get whether all of these were
stopped from being transfused or whether any were transfused, and if
there were, sort of what the results were.

	DR. AUBUCHON:  All of the true positives are Staph epis.  The first
true positive unit was actually a split unit, which happened to be
positive in the 25th month, right after the end of the second year.

	I believe we have had two instances in which units have had the culture
turn positive after the time of transfusion.  One of those occurred very
early on when we had not retained any aliquots for later culture, and
that prompted us to do exactly that, so we could resolve the question.

	In that case, the patient was cultured extensively, nothing was ever
grown from the patient's bloodstream, and the patient had no ill
effects.  On that basis, we concluded that it was a false positive.

	The second occasion, we did have an aliquot.  We recultured it and it
was negative.  We did culture the patient, as well, and as you would
expect, the patient did not have any blood culture positivity.

	DR. KUNERT:  My final question is, of the four, the Staph epi cases
when you did root cause analysis, did you have any revelations?

	DR. AUBUCHON:  No, we did not because we do keep track of our
phlebotomists.  We collect almost all of our own platelets, and we do
keep track of our phlebotomists' technique, and periodically audit that,
and we had not seen any deviation from the way that they were preparing
the arms.  We were using tincture of iodine and alcohol actually on the
skin at that time in any case, and, of course, the welds were all
complete and good welds, and we had the documentation for that.

	So, I would agree with the comment before, that doing a root cause
analysis is not likely to identify the source of the problem.

 	DR. VAUGHN:  Evise Vaughn [ph].

	Just one question.  Seeing as the direction is to move towards testing
for bacteria, is it not possible to use the pre-donation sampling bag to
take the sample and grow from that instead of at the later stage?

	DR. AUBUCHON:  You certainly could.  The difficulty is that there could
be some concentration of the bacteria as the component is made.  In
addition, there is no guarantee that any particular aliquot is going to
have the bacteria in it, and that is the reason, of course, that we wait
for two days or at least one day to allow the culture to grow up to a
point where we can take a small aliquot and culture it reliably.

	I would be a little concerned about only culturing the small diversion
segment.  We would probably get many more positives and probably would
end up throwing out some platelet units that we didn't need to throw out
because the platelet unit itself was not contaminated.

	Certainly, it appears that the rate of positivity is much higher in
those diversion bags than in the culturing of the final product.

	If we are going to culture it as a release criterion, I would rather
actually culture what is being held and then going to be transfused.

	The next speaker is Tracy Manlove from Terumo Corporation.

	We are running very far behind.  I notice you have got quite a few
slides.  I wonder if you could do it in 15 minutes?

	MS. MANLOVE:  I will do my best.

F.  Data Presentation

Ms. Tracy Manlove

	MS. MANLOVE:  I guess I would like to begin by saying good afternoon
since we have reached that point in the day and thank the FDA for the
invitation and the opportunity to discuss this very important topic.

	I do have a number of slides, but Dr. Aubuchon has provided a great
introductory to this, so we may be able to speed through quite a lot of
them.

	[Slide.]

	I am Tracy Manlove and I am speaking on behalf of Terumo Medical
Corporation.  We are the manufacturer of the sterile tubing welders.

	[Slide.]

	I would like to begin by reviewing some of the terminology that we have
been using.  The sterile tubing welders, STWs, are also known and
commonly referred to in the literature as SCDs or sterile tubing
connection devices, and they are all referring to the same pieces of
equipment.

	The outline of my discussion was to go over a brief background and
history, the cleared uses of the devices, and then to review the
description of operation, weld integrity inspections, and general
considerations in making welds, the device release testing criteria,
some supporting data, which is I think what we are really interested in,
and then to summarize, as well.

	[Slide.]

	The sterile tubing welders were originally developed by the duPont
Company in the early 1980's.  The device was conceived to join two
pieces of polyvinyl chloride tubing while maintaining the sterile fluid
pathway.

	[Slide.]

	The original SCD device was developed for use in the dialysis patients
that were on home continuous ambulatory peritoneal dialysis or CAPD.

	In the traditional CAPD procedure, the patient was required to
aseptically connect an indwelling catheter to a bag of dialysis
solution, and they had to do this four or five times daily.  This
contributed to an increased incidence of peritonitis in this patient
population.

	With the advent of the SCD, the risk of contamination was eliminated.

	[Slide.]

	Presently, the device product line consists of three devices - the
SCD312, the TSCD, and the SCD IIB.

	[Slide.]

	The SCD312 that you saw pictures in Dr. Aubuchon's lab and the TSCD are
utilized in the blood bank and transfusion medicine industries.

	[Slide.]

	While the SCD IIB is utilized in the biotechnology industry. 
Applications include cell culturing, fermentation systems, and
bioreactors.

	[Slide.]

	The sterile tubing welders have been in use in these industries for
over 15 years, and all devices within the product line function under
the same principle of operation which Dr. Aubuchon has already
described, and I will, as well, a little bit later in the presentation.

	The field application differences of the devices are only due to the
size of the tubing that the device can accommodate.

	[Slide.]

	Quickly, there are eight cleared uses for the sterile tubing welders,
which are published in the FDA guidance, which I have referenced here. 
This guidance was originally published in 1994 and was recently updated
and reissued in November of 2000.

	Uses include adding a new or smaller needle to a blood collection set,
uses in component preparations, such as adding a third storage container
to a plateletpheresis harness and connecting additive solutions to red
blood cells, of special interest to today's conversation, the pooling of
blood products, and I have included the verbiage in your handout that is
directly from the guidance document.

	[Slide.]

	Also, to prepare aliquots for pediatric use and divided units, and this
is particularly important in minimizing donor exposure in the pediatric
population, as well as maintaining blood inventories.

	Other uses and then, finally, with the removal of samples from blood
product containers for testing, such as QC testing, and as we have been
discussing, is currently done is Europe to remove a sample for the
bacterial culture, and as is currently done in the U.S., to obtain
platelet counts for the split apheresis products that Dr. Williams had
mentioned and others earlier.

	[Slide.]

	Dr. Aubuchon already briefly described the device operation, but I
would like to reiterate and emphasize what we feel, as the manufacturer,
are some important operational points.  So, I have included some
diagrams, as well.  They will hopefully assist in the visualization of
the process.

	It is a unique thermal process where the PVC tubings are set parallel
to each other in what we refer to as holders, and Dr. Aubuchon referred
to as chucks.

	The welder wafer here, as he mentioned, is a copper wafer and it is
positioned in a perpendicular plane to the tubing.  The wafer is heated
to a temperature of approximately 500 degrees Fahrenheit.

	[Slide.]

	When this temperature is achieved, the wafer is then raised and crosses
the plane of the tubing.  As it does this, the tubing is melted.  The
wafer remains stationary with the melted ends of the tubing adhering to
the wafer.

	[Slide.]

	Then, also, as Dr. Aubuchon mentioned, the holders undergo some
movement, so that the left hand tubing is moved to the rear and put into
alignment with the right hand tubing, so these are the two pieces of
tubing that we are connecting.

	[Slide.]

	When they are directly opposite each other, the wafer is then lowered. 
As the wafer recedes, the melted tubing is fused together and form a
weld that has maintained the internal tubing sterility.

	Once this cools, because it was heated to 500 degrees Fahrenheit, after
the cooling process, the welded tubing can be removed from the device
and handled.

	[Slide.]

	This is a very critical step in the process because it is as this point
that the operator must conduct an inspection of the weld integrity.

	Each weld must be visually inspected.  This can be accomplished by once
the tubing is removed from the device, rotation of the welded tubing in
a 360 degree fashion and visually inspecting the weld.

	If the weld is acceptable, it will appear as in Diagram A, where you
can see that the two pieces of tubing have fused together and are in
alignment.  If it is unacceptable, it will be visually recognized as
what appears in View B, where you can see that there is a gap in the
alignment of the two tubing pieces.

	Also, during this visual inspection, if any leaky welds are detected,
they should be treated as air contaminated and handled accordingly, so
that their out-date should change or that there should be a
discontinuation of processing of those units.

	[Slide.]

	As we have discussed previously, the weld integrity inspection is so
critical that it is noted in both the FDA Guidance for Industry, and
here are the specific wordings from that document, as well as the AABB's
21st edition of their standards.

	Again, two standards addressing the need for completeness of the weld
and appropriate actions to take if the weld is not intact.

	[Slide.]

	Once the operator verifies the weld integrity, the weld is opened,  as
Dr Aubuchon said, by simply rolling the tubing between your thumb and
forefingers and the weld will open.

	[Slide.]

	This is a picture of an open and acceptable weld. Right here is the
welded site.  These were two separate pieces of tubing prior to the
weld.  You can see at the connection site that they are perfectly
aligned and there are no leakages.

	[Slide.]

	I would like to discuss from the manufacturer's viewpoint and from an
operator's viewpoint some general considerations when making a weld.

	It is an automated process once the tubing is placed in and the new
wafer is advanced, and we are going to talk a little bit later about the
reuse of wafers and the single use, and hopefully will answer the
question that the committee posed.

	This entire process occurs in approximately 30 seconds once the tubing
is placed and the Start button is initiated.

	[Slide.]

	A new wafer is used for each weld connection and this has always been
the policy in the United States.  Again, a picture of the cartridge of
wafers and just for scale representation, an individual wafer in a hand.

	[Slide.]

	The sterility of the component and the system is maintained by key
features of the welder.

	Number 1 is that the heated wafer kills any bacteria or spores
encountered on the outside of the tubing, and I will show you some
studies to substantiate this claim.

	Number 2 is that the melted tubing adheres to the wafer and forms a
seal which prevents any atmospheric contamination from entering the
system.

	[Slide.]

	Sterile tubing welders are semi-automated devices with built-in checks
to monitor the proper welder function and user operation.  These include
clamp interlocks, wafer checks, audible alarms, and indicator lamps.

	When there is a problem detected by the device, an audible alarm occurs
and indicates the process failure to the operator.

	[Slide.]

	Preparing a good weld is operator-dependent in a number of areas.  The
devices are intended for use by trained individuals in settings, such as
blood banks, hospitals, and laboratories, but the devices are easy to
use and the operating instructions are very straightforward and simple.

	[Slide.]

	One of the operator-dependent instruction in areas is, as we described
with the operation, is that the tubing clamps begin in an aligned
position, but at the end of the welding cycle, there is a different
positioning of the tubing clamps, so in order to initiate the welding
process, the operator must take an action to return that to an aligned
position.

	The operator must also verify proper placement of the tubing.  It is
indicated on each of the devices on the deck of the devices where the
dry tubing goes, which would be such as to the leukoreduction filter, or
the wet product, such as would be coming from the blood component.

	This is dictated by the manufacturer through testing that we have done
to demonstrate that this proper positioning verifies sufficient weld
strength to guarantee the integrity of the weld.  Only PVC tubing should
be utilized.

	[Slide.]

	Other considerations the operator must take into consideration is that
the tubing length must extend beyond the tubing holders.  The tubing
must be properly seated.  You can imagine if you are trying to align two
things, if you don't have them on the same plane, it is never going to
happen, and that the clamp covers play an integral role in keeping that
alignment, and that they must be properly locked into place before
beginning the welding process.

	Again, there are audible alarms that will alert the operator if this
has not occurred.

	[Slide.]

	Hopefully, getting to the question here.  Failure to advance a new
wafer prevents the weld cycle.  Again, an audible alarm and/or a visual
indicator will alert the operator to this.

	The wafers are single use only and reuse of wafers is in direct
opposition to the operating instructions, and voids any and all device
warranties.

	[Slide.]

	In 1993, there was an abstract presented at the AABB by Hawker and
others from the UK, where they conducted a study with repeatedly used
wafers.  What they were able to demonstrate was that the wafers became
contaminated with the solidified PVC from the tubing.  The welds
demonstrated low tensile strength, and we will talk about what that
means later on.

	The contamination resulted in misalignment of the welded tubing and
ultimately, they had weld porosity, and that is to say that they had
leaky welds.

	[Slide.]

	Other things that the operator needs to consider when they are making
the weld to ensure the integrity is to not touch the clamps, not open
the clamps, and to not pull on the tubing.  These are all important in
maintaining the integrity of the weld.

	With each of the devices, we do have bag supports that are provided to
help elevate and support the bags, so that everything can remain on the
same plane and move freely.

	[Slide.]

	In speaking about the devices, they are manufactured, as I said
earlier, by Terumo Medical Corporation.  We are an ISO 9001 facility
located in Elkton, Maryland.

	Each device that is manufactured there is subjected to what we consider
as rigorous release criteria, and this includes 10 dry to dry welds
being made on every device, and 20 wet to dry welds.

	[Slide.]

	All of these welds, 30 welds in total, must meet the following
criteria.  Tensile strength or the force that can be exerted on the weld
before it will break must be demonstrated to be equal to or greater than
15.3 pounds for wet to dry welds and greater than or equal to 15.9
pounds for dry to dry welds.

	The minimum tensile strength of any of those 30 welds must be
demonstrated to be above or equal to 11.2 pounds.

	All of those welds must be aligned and easy to open, and they are
subjected to an air pressure leak test with pounds per square inch of
compressed air.

	These five criteria combined assure the weld strength and the integrity
of every weld made on that device before it is released.

	[Slide.]

	Turning to some data that supports the sterile connection abilities of
the devices, I would like to look at our 510(k) submissions.  This is
the reference here to the submissions.

	The performance testing that was supplied in these submissions included
sterility testing and weld strength testing, as well as four other tests
that demonstrated acceptable test parameters, and that is on file with
the device applications, but won't be discussed in this presentation as
it is not relevant to the discussion.

	[Slide.]

	For the TSCD, the study design for the sterility testing is what I
believe Dr. Aubuchon modeled his test off of, so we are going to discuss
a couple of tests, and they are all very similar in design.

	In this case, the exterior of the tubing was coated with Bacillus
subtilis.  This was chosen because these spores demonstrate a high
resistance to dry heat. There were 1,215 total welds prepared; 405 were
test welds with the coated tubing.

	But then there were two additional welds, that is, 810 welds that were
made to each of these test welds, and that was to dock on the bag with
the growth media on one side of the weld and a transfer bag on the other
side of the weld, and then the fluid traversed the weld site.

	There were three devices tested.  There were multiple manufacturers
tubing encompassed, both dry to dry and wet to dry welds were utilized,
and when wet tubing was used, it was filled with 5 percent human serum
albumin.

	[Slide.]

	The conclusion of this study demonstrated that there was no growth in
any of the 405 cultures after 14 days and that the interior tubing
sterility of both the dry to dry and the wet to dry combinations was not
compromised by the welding process.

	There was positive growth exhibited in the control tubing which
verified the viability of the organism chosen for the test system.

	[Slide.]

	The SCD312 underwent similar sterility studies. In this case, the
exterior of the tubing was coated with either Bacillus circulans spores
or Staph epidermidis, so again a heat resistant spore and a commonly
found skin contaminant that might be found in a real-life application of
this device.

	Six hundred welds were made with the B. circulans tubing, 50 with the
Staph epi, and 6 devices were utilized.

	The conclusion in this test was also that the interior tubing sterility
of the welds was not compromised by the welding process.

	[Slide.]

	Also published in an article by Nicholas in the American Biotechnology
Laboratory in July and August of 1987, is a study entitled, "A Sterile
Connection Device for Cell Culture and Fermentation Systems."

	In this study, Nicholas had two aspects to the study.  She looked at
sterility studies, as well as airborne contamination studies.

	The lengths of tubing were sterilized by ethylene oxide gas prior to
any welding or manipulation.  Then, the tubing exteriors were coated
again with the B. circulans.  A microbiological growth medium was
present in the tubing, this was the medium, and there was also a
bacterial growth indicator.  758 welds were made.

	[Slide.]

	At the conclusion of the sterility studies, it was ascertained that
there was no microbiological growth after 96 hours in the test system,
while the control tubing exhibited growth at 48 hours.

	They made the conclusion that under the correct operating procedures,
all welds were shown to be sterile.

	[Slide.]

	As I mentioned, they also looked at the airborne contamination studies.
 So, they designed this study similarly to the previous protocol
described except that they did not coat the tubing with spores at this
point.  They placed the device into a sport-laden atmosphere with an
average spore density of 260 spores per liter of air, and they performed
114 welds.

	At the conclusion, they found that all of these 114 welds were sterile
and the sterility of the system was maintained.

	[Slide.]

	So, the Nicholas study does offer us some practical evidence that the
impact a sterile tubing welder can have in a fermentation of cell
culture lab.  When the study was published in 1987, the sterile tubing
welder had been implemented in their laboratory and at that time they
had performed 400 welds and 5,000 connections were made with no loss due
to system contamination.

	Prior to the implementation of the device, they were losing
approximately 10 percent of the runs due to contamination.

	[Slide.]

	They also did greater than 1,500 connections to bioreactor systems, and
they did not have to abort any runs because of the contamination when
using the device.  Again, prior to implementation of the device, they
were losing approximately 10 percent due to contamination.

	[Slide.]

	I think I can go through this without anything. I think Dr. Aubuchon
has provided us a very good overview of his study.  The only point that
I would like to make is, as I said, the tubing is still coated with
organisms and as he had alluded to, in a very, very high concentration,
it is a very, very vigorous test for the device to be challenged with.

	[Slide.]

	Again, as he had described, he did three phases of the study, passing
the liquid across the weld, incubated at room temperature, and showed
that all cultures were sterile at the end of the time.

	[Slide.]

	I would like to emphasize, as Dr. Aubuchon did, as well, that they had
two incomplete and leaky welds, but they were doing some innovative
procedures there where they were leaving the tubing wet and making welds
to provide an additional stress and challenge to the system.

	So, it would be interesting to know if that wasn't done, if any of
those two incomplete or leaky welds would have occurred.

	His conclusions, as he already described, but again to emphasize, the
fact that he is advocating visual inspection is in agreement with the
manufacturer's instructions, the FDA, and AABB guidance documents.

	[Slide.]

	At the AABB in 2001, we presented a study that we had done at Terumo
Medical Corporation, myself and some associates had looked at the weld
quality with various tubing combinations from multiple manufacturers.

	We utilized two TSCD's and two SCD312's.  We were looking at wet to dry
welds, and we made a total of 320 welds.

	[Slide.]

	What we saw is summarized here.  This is for the TSCD.  Again, these
were the tubing combinations that we were looking at, the wet tubing
noted first, followed by the dry tubing, the average tensile strength
release criteria, which I mentioned earlier, and the minimum tensile
strength, and you can see that all the values are well above both of
those minimal requirements, and we had no air pressure leaks.

	[Slide.]

	The results for the SCD312 are very comparable.  Again, no air pressure
leaks, no leaky welds.

	[Slide.]

	So, the overall results of our study show that all the welds were
aligned, there were no leaks.  The weld strength exceeded the minimum
strength requirements, and we achieved acceptable weld integrity with
each of our 320 welds.

	[Slide.]

	Recently, Terumo Medical Corporation has undertaken another internal
study where we are looking at an expanded number of tubing types and
manufacturers in conjunction with sterile tubing device.

	Again, we are looking at two TSCD's, two SCD312's. Here, we looked at
dry to dry and wet to dry welds, and a total of 2,400 welds were made.

	[Slide.]

	Here are the tubing combinations as expressed only as dry to dry or wet
to dry, but you can see again the average tensile strength is well above
the minimums, and the minimum tensile strength for the dry to dry welds,
the minimum that we encountered was a weld strength of 15.3 pounds, and
for the wet to dry, a minimum of 11.7.  This is out of 1,200 welds. 
Again, no air pressure leaks in 2,400 welds.

	DR. NELSON:  I wonder if you could summarize because we are way behind
and we need to have time for the questions for the committee.  If we
don't do this, we are not going to have time.

	MS. MANLOVE:  Absolutely.  Okay.

	[Slide.]

	This is just a graphical representation again to see that everything is
well above the minimum, which is demonstrated by the red line.

	[Slide.]

	So, we feel that these two studies show that the sterile tubing welders
consistently prepared welds that exceeded the minimum tensile strength
regardless of the manufacturers type and combination of tubing used.

	[Slide.]

	I wanted to briefly show you some data from our QA Department.  They
are responsible for tracking and trending our Quality Assurance
Department.

	Since January of 2000, these reports have described unacceptable welds,
which are misaligned, leaky, or hard to open welds occurring at a rate
of 3 for every 200,000 welds or 0.0015 percent.  This number is derived
based upon these reports as the numerator and the number of wafer sales
in that same time period as the denominator.

	The low incidence of unacceptable welds, we believe further supports
the performance and reliability of the device.

	[Slide.]

	Dr. Aubuchon has discussed the Merten's paper.  I think the only thing
that we feel incumbent upon us to comment is that there were two blatant
author recognized areas where they were not in compliance with the
manufacturer's instructions.

	They were reusing wafers and, as I mentioned earlier in the discussion,
the Hawker group was able to accurately identify what that did to welds.
 Despite the visualization of the one leaky weld that they saw, they
continued utilizing that unit in the pooling and storage process.  Our
guidance documents in the U.S. would not have permitted that.

	DR. NELSON:  I wonder if you could just conclude or summarize.

	MS. MANLOVE:  Summary.

	DR. NELSON:  There, you are.  Okay.

	[Slide.]

	So, in summary, I would like to remind the committee that the Terumo
sterile tubing welders have been in use for over 15 years.  In that time
period, there has been no reports of transfusion reactions or
contaminated units.

	The rigorous sterility testing studies that I have put forth here and
that Dr. Aubuchon has put forth approximate 3,000 welds that were made
that demonstrate consistency and reliability of the devices and indicate
that there were no incidents of contamination in any of the successful
welds.

	We consider that these studies are scientifically sound and controlled
studies, and that they, coupled with the years of use and the quality
assurance data that I have presented, offer confirmation of the ability
of the sterile tubing welder to safely prepare the sterile welds for
products stored at room temperature.

	[Slide.]

	Furthermore, we feel that the use of the sterile tubing welders to pool
and store platelets for greater than four hours, when combined with a
bacterial detection system, is appropriate.

	We believe that the available data on the sterility of the sterile
connection device procedure supports the use of this procedures to
collect the samples under debate for bacterial detection from in-date
platelet products.

	[Slide.]

	We are confident that the sterile tubing welders maintain the integrity
of closed systems when used according to the manufacturer's instructions
and in accordance with FDA and AABB guidance and standards.

	Thank you.

	DR. NELSON:  Thank you.

	Questions?

	Okay.  The next speaker is Dr. Steven Wagner from American Red Cross.

G.  Design of Clinical Trials for Clearance of

         Devices Intended for Screening of Platelet

Products Prior to Transfusion

Steven Wagner, Ph.D.

Background

	DR. WAGNER:  Hi.  My name is Steve Wagner.  I am with the American Red
Cross.  My stomach is grumbling, so I am going to try to go as quickly
as I can.

	[Slide.]

	I am going to talk today about blood culture methods for screening
platelet components.  I am going to provide a background for Jaro
Vostal's talk on potential designs for clinical trials for release of
platelets as a function of culturing.

	[Slide.]

	Just a very quick comment.  Sepsis is probably the first recognized
infectious disease risk of transfusion that has clearly been indicated
in the times around World War II. The frequency of
transfusion-associated bacterial sepsis was greatly reduced with the
advent of closed systems for collection and storage for blood and with
the dramatic improvements in safety realized by viral testing, bacterial
sepsis remains as the most frequent infectious disease adverse event in
transfusion medicine.

	[Slide.]

	We have seen these numbers before.  I don't really need to go over
them.  I do want to make a point in terms of fatalities that are
reported to the FDA, that between 1990 and 1998, 16.7 percent--sometimes
that is misquoted as 10 percent--of the reported fatalities to FDA were
due to sepsis.

	[Slide.]

	From the BACON study, we know that in platelet components, sepsis or
fatalities from sepsis is measured in about 1 in 450,000 units.  In red
cells, it is much less infrequent because of their cold storage, about 1
in 7,700,000 units, and because of that, most people are focusing on
trying to intervene with culturing platelets.

	[Slide.]

	In plasma, to my knowledge, no observed fatalities from sepsis have yet
been observed.

	[Slide.]

	There are two systems that have been cleared by the FDA for screening
of platelets for bacterial contamination.  One uses a color change or a
rate of a color change in a pH-sensitive disk or sensor area, and it is
presumably by bacterial-generated carbon dioxide evolution, and the
other, from another manufacturer, involves the detection of a reduction
of blood gas oxygen caused by bio-oxygen consumption.

	[Slide.]

	This is just a comparison of the two systems that I alluded to.  One
system is the bacT/ALERT.  The cleared component is for apheresis
platelets.  It involves 4 ml that is cultured in aerobic bottle and 4 ml
that is cultured in anaerobic bottle.

	The sampling time that is permitted is greater than 24 hours.  This
allows the bacteria to grow to a level where, when you take a sample
that there is a more likely event that bacteria will be present.

	The incubation time after inoculation into the culture bottles should
be greater than 24 hours.  The number of evaluations of the culture is
continuous.

	Another system that has been cleared, that is marketed by Pall, has
been cleared for both leukoreduced random donor platelets, as well as
apheresis platelets.  The volume required in the pouch where the oxygen
is sensed is 2 ml, but in actuality, it uses 6 to 7 ml of a platelet
concentrate simply because filling the tubing, going through a filter
requires some volume of platelets.

	The sampling time that is recommended is 48 hours, but it is permitted
for sampling to occur after 24 hours. The incubation time is recommended
to be 30 hours in this system, but again is permitted to be after 24
hours, and the number of evaluations for a platelet component is once.

	[Slide.]

	Culture systems are quite sensitive and by definition they are able to
detect one viable organism that is capable of growth if that sample is
inoculated into a culture medium.  That requirement is dependent on, of
course, first, the initial bacterial load in the component.

	Most people who work in this area believe that the initial bacterial
load is quite low.  From essentially one organism in an entire platelet
component to 10 organisms per ml. The rate of growth in the platelet
component also determines whether you get a sample in your syringe for
inoculation into a blood culture instrument and also the volume of the
sample.

	[Slide.]

	For fast-growing organisms, it is pretty clear that if you wait greater
than 24 hours to take your sample and you wait longer than 24 hours to
incubate your sample, that you are going to have a very, very high level
of detection.

	In the two studies cited here, one through my lab, which was, by the
way, sponsored by Organon Technika, which is the manufacturer of one of
the devices, and also through another lab, which is an industrial lab in
Gambro BCT, it showed 100 percent detection of fast-growing organisms.

	[Slide.]

	But slow-growing organisms and low bacterial loads represent the most
stringent conditions for evaluating culture conditions.  In these sorts
of systems, Staph epidermidis is probably the most frequently implicated
slog-growing organisms identified in clinical cases of
transfusion-associated sepsis.

	[Slide.]

	From our laboratory, we were able to show if you sample immediately
after culture, using a very low inoculum,  a tenth of an organism per
ml, you don't detect anything, and after around 24 hours, you can detect
around 67 or so percent of the cultures as culture positive.

	If you are less stringent and inoculate with 1 organism per ml or 10
organisms per ml, the 67 percent detection actually goes up to 100
percent detection.  If you wait longer than 24 hours and sample at 48
hours, you essentially detect everything.

	[Slide.]

	So, how large should the sample volumes be?  The answer depends on when
you sample.  If you sample at day zero or an early time, before 24
hours, it turns out that the larger the sample volume, the better the
frequency of detection.

	If you wait for one or two days before you take a sample, and during
that time, of course, the bacteria will be proliferating in the
platelets components, we were able to find that both a half a ml sample,
as well as a 2 ml sample yielded identical frequencies of bacterial
detection.

	[Slide.]

	I also wanted to address a question of whether anaerobic culture really
is needed.  The partial pressure of oxygen and platelet components is
between 40 and 100 ml of mercury.

	There have been two cases where strict anaerobes have been associated
with clinical cases of sepsis, and the two cases both involved
Clostridium perfringens.  In one case there was a fatality, and in
another case there was morbidity.  One case was in a red cell unit,
another case from a pooled platelet unit.

	The microbiological textbooks indicate that Clostridia cover an entire
range and the need for anaerobicist, and many are not fastidious, so
many of them you would pick up in an aerobic culture bottle anyway.

	This compares to many scores more bacteria that have been implicated in
transfusion-associated bacterial sepsis that have been able to grow up
in non-anaerobic conditions.

	[Slide.]

	How long to incubate?  In these very stringent conditions, we found
that the incubation time was inversely related to the time when you
initially sampled, so if you waited a day for sampling, it could take
you a little bit over a day for detection with this sample volume and
with 2 ml, that didn't change.  If you wait 48 hours, it takes about a
half a day.

	[Slide.]

	There has been some talk about 100 percent QC of platelet components. 
This would be done on day 1 or 2 sampling, and after sampling, sometime
soon after sampling, the platelets would be released.

	There would be adequate platelet availability through the week as long
as the shipping was less than one day and in most cases, we can get our
platelets to hospitals within a day, but it does require a failsafe,
real-time communication between the blood collection center and the
hospital.

	This has been done before as has been indicated or mentioned for NAT
testing.  What are the ramifications when a contaminated product is
infused, that is later to be determined to be culture-positive?

	I really can't answer that question, but I imagine that would be of
great importance to both the blood providers, as well as the recipients.

	[Slide.]

	The implications of using culture in terms of platelet release are a
bit different.  My analysis is that hospitals probably wouldn't receive
platelets until day 3 because it takes some shipping time, there is some
time until sampling, and there is some incubation time.

	This has been studied by Chang Phang at the American Red Cross. 
Assuming that all centers do not collect on weekends, what that would
mean is that there would be no platelets available on Thursdays.  With
no collection on a Friday, for example, there was a long weekend, that
would mean that there would be no platelets available on Wednesdays.

	On a long weekend with a Monday holiday, that would mean that there
would be no platelets available on Fridays.  So, I guess what I am
trying to address is there are some availability issues involved with a
culture release model where the hospitals don't get the platelets until
day 3.

	There conditions would require either uniform weekend collections,
which is possible, but quite a change for the blood providers, or an
extension of platelet storage time.

	[Slide.]

	In terms of the extension of platelet storage time, Jim Aubuchon hasn't
presented his data, but there is an abstract out indicating that
platelet properties and survival look initially good after seven days of
storage, the data are encouraging.

	An extension of storage might offset the cost of testing by reducing
the percentage of outdated platelet components, so that is good for the
blood providers, but microbiological data needs to be collected to
support extension of the platelet storage time with the introduction of
a particular bacterial test.

	[Slide.]

	So, after introducing bacterial culture, the development of a seven-day
platelet component would be facilitated by demonstrating that the
frequency of repeat culture-positive units is similar after five and
seven days, and I think that would probably form the basis for
determining whether a seven-day product is safe or not.

	[Slide.]

	In conclusion, bacterial culture is a sensitive method for detecting
bacteria in blood components, aerobic cultures should detect a great
majority of clinically important organisms.

	The choice of sampling and incubation times are an important
determinant of detection frequency with sampling and incubation times
chosen for acceptable detection frequencies.  A 100 percent QC culture
release notification model should be compatible with adequate platelet
availability.

	A quarantine release model for bacterial culture would require weekend
platelet collection or an extension of the platelet storage time. 
Studies suggest that seven-day-old platelets maintain their in vitro and
in vivo properties and data need to be collected on the microbiological
risk of storing platelets for five compared to seven days following the
introduction of bacterial culture.

	Thank you very much.

	DR. NELSON:  Thanks very much.

	Questions?

	It seems like, as opposed to two days after collection, one day after
collection, the problem is I guess you would have to have longer
incubation times to make up for the earlier collection, so the time at
which the platelets were released would not be shortened by earlier
culturing.

	DR. WAGNER:  That's right.  I think the reason for that is bacteria
grow in the culture about as well as they do in the bottle, so you still
need the same amount of time whether you slice it one way or the other.

	DR. KLEIN:  But in point of fact, the agents that are the ones we were
really most worried about are the ones that grow faster, so clearly,
what you did was the way to do the experiment with those that grow most
slowly in the lowest concentrations.

	Maybe we could calculate in terms of reactions and deaths what the
interdiction would be if we sampled, for example, at 24 hours and then
released at 24 hours after culture.

	DR. WAGNER:  I agree.  I think that is worthy of study.  I think that
the bad actors are the gram-negatives that are fast growers.

	DR. ALLEN:  Two questions.  Let me ask the first to you and perhaps Dr.
Klein could comment also.

	Given the information available now, would you recommend the culture
release notification model or a quarantine release model, or do you
think we need more study?

	DR. WAGNER:  I am with the Red Cross, so I am biased in this a little
bit.  I believe that logistically, right now, what we can handle is a
culture release model and I think that we would need a longer platelet
storage time greater than five days to be able to handle a quarantine
release model.

	DR. KLEIN:  I would just comment that I think you can make the culture
release model work.  It is not going to be perfect, but it is going to
be much better than what we have now without worrying about having a
lack of availability of platelets.

	DR. ALLEN:  The second question.  You commented on the paucity of any
data suggesting that anaerobic bacteria,  by and large, are a
significant problem in terms of platelet contamination.

	Can you do an aerobic culture only, would you recommend that, or do you
think you still need to follow the recommended model of the aerobic and
anaerobic bottles?

	DR. WAGNER:  My answer is I believe anaerobic culture would detect the
great majority, a vast majority of clinically relevant cases of sepsis.

	However, in the product insert for the bacT/ALERT, it said it was
recommended that both aerobic and anaerobic cultures be performed.  I
wanted to bring this up because personally, I disagree with that.

	DR. SIMON:  I know the FDA has not put before us the question of a
culture release model for discussion, but I think you raise the issue
tangentially, but certainly the liability issue is going to be on the
minds of any blood center that adopts such a bottle or considers
adoption of such a model whereby they would release the unit once they
have taken the culture and not wait at least for a 24-hour result.

	The Blood Centers of America are very risk averse for obvious reasons. 
I think that could result in a significant impact on availability of
platelets for patients in need.  So, I just bring that up as something
that is in the background.

	DR. WAGNER:  There are alternative ways of looking at it that are, as I
think of it, shades of gray where you took a sample at 24 hours, kept
it, and did not send it out of your facility for 10 or 12 hours, but you
are not really doing release from a quarantine and then let the units
out.

	That would interdict most of the fast growing organisms and then the
slow growers, you would have to call on later.

	So, I think that it is difficult to look at something as dynamic as
culturing in kind of a digital system.  It is more of an analogue type
of system.

	DR. NELSON:  In order to have time for lunch, I wanted to have Dr.
Vostal talk about a proposed study design for evaluation.

H.  Proposed Study Design

Jaro Vostal, M.D., Ph.D.

	DR. VOSTAL:  Thank you very much.

	I will try to sprint to the finish and get us to lunch before the
lunchroom closes.

	What I would like to start off with by saying that this is our current
thinking about clinical trial design and we are really open to
suggestions and discussion about how this should be designed.

	Dr. Wagner has done a very nice introduction for me, so I will actually
be able to skip some slides in the beginning.

	[Slide.]

	The issue of 100 percent QC of platelet products was raised.  We think
that we still need a clinical trial of automatic bacterial culture
devices or ABC devices even though there will be 100 percent QC of
platelet products because the QC monitoring will not assure that
products are culture negative at the time of transfusion either day 5 or
day 7 because the devices have not been validated for this issue.

	So far no clinical data is available on whether a negative culture
early in the storage period is predictive of a negative culture at day 5
or at day 7.

	[Slide.]

	So, the intended use of these devices is to screen bacterial
contaminated platelet products prior to transfusion.  The evaluation
process that we will be looking at will be laboratory testing, as Dr.
Wagner covered, and we think we require a clinical trial.

	[Slide.]

	Now, if you do go through a clinical trial or if the device goes
through a clinical trial, what kind of label can you put on your product
if you have been screened by a device such as that.

	We think the appropriate label would be bacterial culture negative for
up to five days of storage for five-day-old platelets or a bacterial
culture negative for up to seven days of storage, and the asterisk here
is this requires that the storage must be under conditions validated to
adequately store platelets up to seven days.  Actually, that is a
separate issue from the contamination rate.

	[Slide.]

	In the laboratory testing of these devices, as has been already
described, you spike in bacteria at a certain concentration and you can
then follow the growth of bacteria in the platelet product over the
storage period.

	With your device, you can sample at different time points and culture
to get a result either 24 or 48 hours later, and with this design, you
can also determine the sensitivity at the point of collection and CFUs
per ml at the time of sampling.  So, this would be a design of the
laboratory type studies.

	[Slide.]

	Actually, this slide just briefly talks about the different organisms
that we recommend that are tested during the laboratory studies, and
these are described by Mark Brecher's paper in Transfusion in 2001.

	[Slide.]

	The information you get from laboratory studies is the approximate
level of sensitivity and this is a moving target.  It is based on when
you sample and it is based on the device.

	We think for day 1 sampling, sensitivity should be on the order of 10
to 100 CFU per ml.  The other things you get out of the laboratory study
is the optimal sampling time, the length that you keep it in culture,
and the optimal sampling volume.

	[Slide.]

	So, moving on to the clinical trial, we think that the trial should
demonstrate that a second culture taken at the end of the storage period
confirms the results obtained from a culture taken early in the storage
period.

	So, a comparison study where you have a culture early on in the
culture, later in storage, to see if the results agree.  This is just a
graphic demonstration of what a study could look like.

	This would be an ideal study where you actually take your sample early
on and then you wait until the full length of the storage, right now it
would be day 5, and then take your second sample and see if you get
agreement.

	The reason it is ideal is because this would be a high-risk day for
platelet contamination.  The problem with the study is you actually have
to wait to out-date and you lose the ability to transfuse these platelet
products, which is probably not good for the clinical community.

	[Slide.]

	This just runs through the ideal study.  Basically, you collect your
first sample time point at a time point that is identified by laboratory
studies and which has the optimal chance of assuring that the product is
culture negative at day 5.

	The second culture is collected at out-date.  The primary endpoint of
these studies would be agreement between the first and second culture,
and we can discuss the level of confidence that we want for this type of
agreement.

	As I mentioned, the design is not practical because you lose someone's
transfusion products, however, this design could be modified to look at
platelets that are going to be outdated anyway.

	Let's say if you screen all of your platelets with the first culture,
and then only culture the platelets that are going to be outdated at day
5, and even you could hold those up to day 7, so you could do a study
like that if you only looked at the outdated platelet products.

	[Slide.]

	Another way to do this would be to actually transfuse the products
during the study, and you can transfuse them up to day 5, and collect
your second culture at the point of transfusion.

	Because the risk would be with longer storage, it would be better to
have sort of a waiting of the data towards the later end of the storage,
so we have suggested here that day 5 makes at least 25 percent of the
total samples, but at least in this type of design, you would be able to
transfuse your products, which would make it lot more cost effective
than the other study.

	[Slide.]

	We are calling this a realistic study.  You have the confirmatory
sample at a time point day 2 to day 5, and day 5 samples should
represent a high percentage of the collected data.  Again, you are
looking for agreement between the first and the second culture.

	Now, as has been pointed out, if you could extend the storage out to 7
days, you would actually have an added benefit from these transfusion
products, so you could actually offset the cost of doing these studies.

	So, we are considering that such a study would be possible, and we
think it should also be designed in a similar way that you have a first
sample early on and then the second sample would be at the point of
transfusion, just like it was in the day 5 study, and then you would do
this at day 6 and day 7, as well.

	Since these products are the ones that have the highest risks, since
that was the reason, bacterial contamination was the reason they were
taken off the market,  we would like to see a high percentage of the
second culture be done on day 6 and day 7 platelets.

	[Slide.]

	Now, how to get around the question of transfusing these products at
day 6 and day 7, since they are at high risk for bacterial
contamination, we thought that maybe it could be screened by yet a third
culture of a third bacterial detection method, for example, after day 5,
could transfuse products if a bacterial detection screen is done before
transfusion, just to make sure that you don't have a highly contaminated
unit that you are going to transfuse because you second culture done on
this product would come back 24 hours later.

	So, it is screened by bacterial culture, then sampled at day 5.  If you
decide to use a culture as the third detection mechanism, if you sample
at day 5, then, you can transfuse that if it's negative by 24 hours.

	If you use a non-culture detection method, such as Gram stain or
dipstick or something else that is less sensitive, but may be
appropriate for units that would be highly contaminated with the
bacteria, you could transfuse as soon as these tests come back negative.

	Again, the confirmatory second culture needs to be taken at the point
of transfusion for comparison with the first culture.

	[Slide.]

	This is again a graphic demonstration, so you would be taking your
first sample, your second sample at the time of transfusion, and you
would guarantee the safety or you would attempt to guarantee the safety
of these day 6 and day 7 units by a bacterial screen either by culture
method or by some other bacterial detection method.

	[Slide.]

	So, what would be the size of this clinical study? Well, it is actually
very difficult to estimate this because it depends on the expected
contamination rate of the platelet products.

	We have heard several numbers mentioned today.  It could be 1 in 1,000,
1 in 2,000, 1 in 3,000, and also the size of the study depends on the
level of certainty that the first culture would be predictive of a
culture-negative platelet product at the end of the storage.  You can
choose  your level of comfort at 99, 95 percent or less.

	[Slide.]

	So, I have had some help with my biostatistician colleagues.  They
actually pointed out that this should be more in agreement between
sensitivities, similar sensitivities at the 99 or 95 or lower level.

	If you choose this type of agreement, you would need to screen or at
least collect 300 contaminated units, and your expected contamination
rate is 1 per 1,000, you have to screen about 300,000 units.

	If you decrease your agreement level, this number goes down,  and we
feel actually that probably 95 percent agreement would be appropriate,
so it may be somewhere in the order of screening 60,000 units, but again
this depends on what the expected contamination rate is.

	So, this is sort of a rough draft of a clinical study proposal.  We
welcome any comments or discussion that you would have.

	Thank you.

	DR. NELSON:  Thank you, Dr. Vostal.

	Questions, comments?  It sounds like a big study.

	Yes, Mary.

	DR. CHAMBERLAND:  Given the difficulty that you alluded to, the ideal
example of actually being able to culture units that truly were on the
shelf for five days or seven days, would another possible variation be
the day that they are actually going to be transfused?

	Let's say it's day 3 that they are selected to be transfused, and you
take the sample, could you just incubate, you know, maintain that sample
out for a total of five or seven days and then culture?  You know,
maintain the sample at similar conditions to which the platelets are
normally maintained.

	Obviously, there is big volume differences and whether that would
preclude that as being a valid approach,  but I was just curious about
that.

	DR. VOSTAL:  That is an interesting suggestion.  I guess the problem
with that would be that the growth of the bacteria may be different in
this new environment that you put them in, you know, the smaller volume,
less gas permeability, then what actually would be going on in the bag
itself.  So, it might not represent the growth curve that you would see
in the actual product.

	DR. FITZPATRICK:  I think one of the goals needs to be the earliest
point of detection that correlates to the five and seven day level of
contamination, so if you sample only at day 1 and then at the point of
transfusion, you miss a block of time that may be important.

	So, I think you would want to sample at a 24-hour interval up until the
point of transfusion, so that you can determine if day 1 didn't work,
and didn't correlate, now you have got to repeat the study and try it at
day 2 or try it at day 3.

	So, if you don't see a correlation between day 1 and day 5 or day 7,
you haven't collected the data you need to determine when that
correlation occurs.  I just would support what Mary said.

	I think if you worked with the manufacturers, while you couldn't find a
perfect way to collect an aliquot at the time of transfusion, you could
find a way to prepare a 10 ml pouch or a 15 ml pouch that approximates
the conditions, that you could store until day 5 or day 7, that would
help give you that information without wasting the products.

	DR. NELSON:  You might also have to vary the incubation times based
upon when the culture was taken, so it is a bit of a complex experiment,
but I can see that it would be useful.

	How solid do you think the 1 to 1,000 estimate is, because if that is
way off, then, all of a sudden you are talking about an astronomically
larger--I mean if it is much lower than that, to get the results might
be even more of a problem.

	Is that pretty solid, the 1 in 1,000, do you think?

	DR. VOSTAL:  Actually, I don't that is very solid. I mean nobody really
knows what the true contamination rate is.  It will be pretty much a
guess.

	DR. SIMON:  See if my interpretation is correct,  but from what Dr.
Williams said, I gather that if the industry wishes to exceed a quality
control standard and actually test all units, do a culture on all units
and use it as a release criteria, either culture release or wait for the
results and release, then, the practical impact of your clinical trial
would just to allow the company to make a claim.

	We could already have actual culture of all units before the clinical
trial was done.

	DR. VOSTAL:  Right.  I guess it depends on when you take your culture. 
If you take your culture early on, we are not really sure right now
whether that culture will be predictive of what will happen at the end
of storage.  I mean that is why we need the study.

	If you take your QC like at day 3, and you are willing to wait for the
results, so you transfuse at day 4, that probably would work, but I
think you would lose three or four days of transfusion.

	DR. SIMON:  You are saying that you are dubious about the whole concept
of the 24-hour culture as eliminating almost all of this problem?

	DR. VOSTAL:  Oh, you mean taking sampling at 24 hours.  Well, I think
as Steve pointed out, the longer you wait, the higher your sensitivity
gets because the bacteria grow to a higher level.

	So, I think it's a tradeoff.  If you sample early on, you might miss
some, and the question is how many do you miss and will that be a risk.

	DR. NELSON:  From a practical standpoint, five days is the limit now,
but how many platelets are actually released, you know, one, two, three,
four days earlier than that?

	Somebody has those data, I guess, but given "never on Thursday"
scenario that Dr. Wagner presented, it seems a little bit complicated.

	Do you have a comment?

	DR. KUNERT:  I just had a quick question.  On your sample sizes, was
that assuming the two cultures or was that assuming three cultures
lengthening out to day 7?

	DR. VOSTAL:  It would be looking at the two cultures.  The third
culture is actually just to provide safety if you plan to transfuse the
day 6 and day 7 product.

	DR. KUNERT:  What data are there to suggest that you wouldn't have
cultured it out at, say--I am not sure what the assumption is on the
second culture--but if it is day 3 or 4, that you would then culture it
on day 5?

	Staph epi would be the biggest example, but I don't know even know with
Staph epi that--I mean you should be able to culture it at day 3 or 4,
so is it the concern mainly fastidious organisms or is there any
particular scenario you had in mind for that?

	DR. VOSTAL:  I guess if you are saying that if we culture at day 3,
that should be sufficient to cover day 4,  5, and 6, out day 7, right?

	DR. KUNERT:  As Dr. Wagner pointed out, it depends on the time of
sampling to the time of culture, but there could be some point where
there should be a level of confidence where you might not need to
culture at day 5 depending on those parameters.

	DR. VOSTAL:  You have to optimize it, I guess, because if you culture
at day 3, are you going to not transfuse day 1, 2, and 3 platelets, or
are you going to transfuse them or hold them.

	I think you have to play around with the logistics of the study and
logistics in the blood bank to try to optimize it.

	DR. NELSON:  Yes.

	DR. SNYDER:  Ed Snyder, New Haven.

	Do you have a similar approach that might be useful for random donor
platelets?  I mean that could be used, but you would have to sample each
bag.  We use a four-unit pool,  so that would be a fair amount of
sampling.

	I assume you wouldn't let us pool and store the pool before release and
testing at various times.

	DR. VOSTAL:  Well, I think we are actually having discussions about
pooling upfront and storing pools, however, we need more data on that in
terms of at least platelet efficacy and also in terms of whether these
devices can--you know, if the growth of bacteria in the pools is
different and whether the devices can pick up that contamination.

	So, we would expect a separate study done on the pools themselves.

	DR. STYLES:  I was just going to suggest that if you are going to
undertake such a large study, that you want to incorporate DNA-based
screening techniques within that study instead of having to go back and
repeat it with the advent of PCR and all.  I mean you are going to avoid
the whole need to wait after culture if those techniques come to
fruition.  Just a thought.

	DR. VOSTAL:  I mean these studies are designed to look at culture
devices and, you know, the sensitivity and the time you have to wait to
get a readout, but if there is a screening method that would be
immediate, I think that would be a lot better.

	DR. NELSON:  Here is the dilemma.  We have got four people that wanted
to make statements in the open public hearing and then we have to
consider the questions, and the lunchroom closes in about 15 minutes.

	I propose that, of the three alternatives, lunch is a higher priority
at this time.  Why don't we break now. Let's come back at 2:30.

	DR. UNDERWOOD:  Those people that are speaking in the open public
hearing after lunch, you know my rule is five to seven minutes.  It is
now five and half each, so please be prepared.

	Thank you.

	[Whereupon, at 1:35 p.m., the proceedings were recessed, to be resumed
at 2:30 p.m.]

AFTERNOON PROCEEDINGS

[2:30 p.m.]

I.  Open Public Hearing

	DR. UNDERWOOD:  This is the open public hearing for the bacterial
contamination.  As I admonished those speakers prior to lunch, if you
can make your presentation as brief as possible.

	We have four scheduled speakers for the open public hearing on
bacterial contamination:  T.J. Smith from Medi-Flex Hospitals, Dr. Roger
Dodd, Kay Gregory, Dr. Bianco.  Are those people here in the room?

	Those that will not need to use the slide projector, if you can proceed
and perhaps use the mike in the center aisle, please.

Kay R. Gregory

	MS. GREGORY:  My name is Kay Gregory.  At this time I am representing
the American Association of Blood Banks.

	The American Association of Blood Banks (AABB) is the professional
society for over 8,000 individuals involved in blood banking and
transfusion medicine and represents approximately 2,000 institutional
members, including blood collection centers, hospital-based blood banks,
and transfusion services as they collect, process, distribute, and
transfuse blood and blood components and hematopoietic stem cells.

	Our members are responsible for virtually all of the blood collected
and more than 80 percent of the blood transfused in this country.  For
over 50 years, the AABB's high priority has been to maintain and enhance
the safety and availability of the nation's blood supply.

	The AABB believes that bacterial contamination of platelets is the most
significant current infectious threat from blood transfusion and
appreciates the opportunity to comment on this issue.  For decades,
bacterial contamination has been recognized as a significant risk
associated with room temperature storage of platelets.  The AABB
believes the time has now come to take action on this issue.

	As other infectious risks of transfusion have been reduced, the
magnitude and relative importance of bacterial contamination of
platelets has become more pronounced.  Various innovative strategies
have been and are being developed to address this risk.

	Although no single method or strategy provides a perfect solution, the
AABB believes that multiple approaches may be appropriate for
consideration.  Methods to prevent and detect bacterial contamination in
both apheresis and pooled platelets made from whole blood have been
implemented in other countries.

	These methods have undergone clinical evaluation in this country,
demonstrating the ability to detect some bacterially-contaminated units.
 The AABB notes that the FDA has recently approved two culture-based
bacterial detection systems for quality control testing of
leukocyte-reduced platelets.

	At this critical juncture, the AABB sees a valuable opportunity for
cooperation between the transfusion medicine community and FDA.

	The AABB reviews its voluntary Standards for Blood Banks and
Transfusion Services on a planned basis.  The next edition of these
Standards, the 22nd edition, has just been published for public comment
and proposes two significant changes with regard to decreasing the risk
of bacterial infection for recipients of platelet transfusions.

	The first focuses on prevention of bacterial contamination of the
donated unit, and involves changes to the skin preparation method. 
Based on the data reviewed, AABB has recommended that alcohol/tincture
of iodine be the method of choice, with chlorhexidine being acceptable
for individuals who are allergic to iodine.  On the basis of the data
reviewed, the Standards Committee has concluded that green soap is not
acceptable for skin preparation.

	The second change the AABB has proposed is a draft standard requiring
that facilities have methods to detect bacterial contamination in all
platelet components.  In light of the fact that no single system or
method is effective in eliminating the risk of bacterial contamination
in all components, the AABB has declined to be specific as to the method
of bacterial detection required in this proposed standard.

	There are a number of logistical and scientific issues to be resolved
prior to implementation of any detection system, but the AABB believes
it is critical to begin to address these issues now.  The AABB
recognizes that some facilities may opt to use a method that gives
immediate results, while others may be able to adopt culture
technologies.

	It is also relevant to note that this proposed standard would require
screening of all platelet components. If the goal is to reduce
infections in recipients, it is essential that all platelet components
be evaluated.  A statistical sampling approach runs the risk of not
effectively decreasing the rate of bacterial infection.

	It is much more feasible and practical from both a logistical and a
product loss standpoint to perform bacterial detection, especially using
culture methods, on apheresis platelets.  However, the entire need for
platelet transfusion is not currently, nor will it be in the foreseeable
future, met by single donor apheresis platelets.

	Whole blood derived platelets are necessary to ensure an adequate
supply of platelets.  The potential application of culture methods to
detect bacterial contamination in apheresis platelets cannot be allowed
to render platelets from whole blood an undesirable component.

	To this end, the AABB recognizes that detection techniques such as
Gram's or Wright's stain, or dipstick monitoring may initially need to
be used for whole blood derived platelets.

	The AABB believes that the FDA can facilitate bacterial detection of
whole blood derived platelets by reexamining its current thinking under
which platelets pooled in either the blood collection facility or the
transfusing facility, regardless of the use of sterile methods, cannot
be used beyond four hours after pooling.

	The FDA's current thinking makes the culture of pooled platelets
impossible.  In the interim, alternative, albeit less ideal, methods,
including microscopy with acridine orange, Wright's or Gram's stains, or
dipstick monitoring of glucose and/or pH with appropriate thresholds are
available for use at the time pooled platelets are released.

	The FDA appears to have indicated that it would require in vivo studies
of platelet effectiveness before considering extending the storage of
platelets pooled using sterile methods to five days, as is currently
allowed for non-pooled product.

	However, such in-vivo studies are difficult to perform, expensive,
require the enrollment of large numbers of patients from multiple
institutions, and are difficult to analyze due to multiple, unavoidable
confounding factors.

	In light of existing in vivo data from Europe concerning the five-day
storage of pooled platelets derived by the buffy coat method and in
vitro data showing the similarity between platelet-rich plasma derived
platelets and buffy coat platelets, the AABB urges the FDA to examine
ways in which it could expedite approval of the extended storage of a
pooled platelet product.

	The AABB urges the FDA to act quickly to consider what data will be
required to extend platelet storage to seven days, provided that an
acceptable bacterial detection system is used.

	In light of the challenges and tremendous opportunity for improving the
safety of the blood supply through the implementation of the bacterial
contamination methods described above, the AABB requests the following
assistance from FDA:

	1.  Regulatory support towards accomplishing AABB's current goal of
requiring bacterial detection and interdiction of contaminated products.

	2.  Regulatory support in developing consensus on arm preparation
solutions and techniques, with a specific emphasis on prohibiting the
use of green soap.

	3.  Discussion of data required to increase the storage time for random
pooled platelets with a particular focus on whether in vitro data on
platelet bacterial growth rates is acceptable.

	4.  Discussion of the data needed to extend the out-date of platelets
to seven days.

	As has been the case relating to the development of new tests for
emerging infectious diseases, the blood banking and transfusion medicine
community and the FDA must understand the need to implement less than
perfect solutions, while we work to improve the available methodology
and technology, recognizing that such incremental steps will improve the
safety of the blood supply.

	Thank you.

	DR. NELSON:  Thanks very much.

	Roger.

Roger Y. Dodd, Ph.D.

	DR. DODD:  Thank you very much, Ken.

	My name is Roger Dodd.  I am the Executive Director, Biomedical Safety
at the American Red Cross.  At the moment I am representing the American
Red Cross, which collects about half of the blood components used for
transfusion in the United States.

	One of our strategic priorities is:  "To provide high quality, safe
products."  The American Red Cross thanks the Food and Drug
Administration and the Blood Products Advisory Committee for this
opportunity to address a topic of great importance to platelet
recipients in the United States.  We applaud the FDA for its attention
to the issue of bacterial contamination of platelet components.

	The Red Cross agrees with the AABB statement relating to the serious
nature of bacterial contamination and recognizes that measures should be
taken to reduce or eliminate the occurrence of transfusion-related
sepsis.

	We recognize that an immediate, single solution is not currently
available and acknowledge that attention to aseptic practice and to
appropriate skin preparation continue to be a critical foundation for
maintenance of bacterial safety.

	We further agree that it is highly desirable to implement means to
detect bacterially contaminated platelet units and recognize that some
approach to diversion of the initial collection volume may complement
such detection.

	We challenge researchers and manufacturers to develop rapid, highly
sensitive tests that may be used to assure their platelets are
bacterially safe; ideally, such methods could be used prior to release
of products.  In the meantime, we recognize that FDA's approval of two
culture-based methods for platelet quality control is a step in the
right direction.

	The Red Cross is in the process of determining the feasibility of
implementing procedures to assure quality control for bacterial
contamination of all apheresis platelets and will discuss with the FDA
available options to assist hospital customers in reducing the risk of
transfusing any components that fail to meet bacterial QC requirements.

	In common with the AABB, the Red Cross is concerned about the ability
to complete such QC on random donor platelets without compromising their
availability and efficacy.

	We hope that the FDA will be willing to consider the concerns expressed
by the AABB and thus to assist the Red Cross in fulfilling its mission.

	Thank you for your attention.

	DR. NELSON:  Thank you.

	Celso.

Celso Bianco, M.D.

	DR. BIANCO:  I am Celso Bianco.  I am speaking for America's Blood
Centers.

	America's Blood Centers (ABC) is a national network of
locally-controlled, not-for-profit community blood centers that provide
nearly half of the U.S. blood supply from volunteer donors.

	Collectively, America's Blood Centers' total blood collections exceeded
7 million donations in 2001.  ABC members operate in 45 states and in
Quebec, Canada, and serve more than half of the 6,000 hospitals in the
U.S.

	Members of America's Blood Centers thank the FDA for the opportunity to
participate in this public discussion about the reduction of the
incidence of bacterial contamination of blood components.  Bacterial
contamination is the second cause of transfusion-related fatalities
reported to FDA, representing 10 percent of the cases with an average of
five reports a year.

	Only hemolytic reactions due to errors cause more fatalities--an
average of 18 a year and half of the reported transfusion-associated
fatalities (Jong-Hoon Lee, M.D., CBER, FDA September 1999).

	ABC members also agree that measures to reduce the incidence of
bacterial contamination of blood components should be implemented. 
However, they believe that a number of unresolved issues must be
considered by this committee and by FDA before the agency issues any
specific requirements.

	Bacterial contamination of blood components is a far more complex
problem than viral contamination.  Substantial reduction of transmission
of HBV, HCV, and HIV by transfusion has been achieved by screening
assays that are specific for each virus,  as well s by donor history
questions and donor deferrals.

	Viruses do not replicate during component storage; what is in the donor
is in the blood sample collected for testing and is in the blood bag.

	Bacteria, on the other hand, are everywhere.  There are thousands of
species that may contaminate blood products, and they replicate during
storage.  They may be present in minuscule amounts in the donor's
circulation, they may survive skin disinfection, and there are no
specific tests.

	The sensitivity of disinfection and detection systems varies according
to the type of bacterium.  Thus, while everyone agrees that something
should be done, there is no clear agreement about what should be done.

	The American Association of Blood Banks' Standards Committee is
proposing new standards for skin disinfection in the next edition of
AABB's Standards.  The Standards Committee has also proposed the
implementation of bacterial detection systems (without specifying how
this should be done).

	We all agree that disinfection of the venipuncture site should be
performed using the most effective method possible.  Recent studies
suggest that tincture of iodine would be better than current methods.

	A second approach to reducing the incidence of bacterial contamination
adopted in some European countries, e.g., The Netherlands, is attaching
a diversion pouch to the collection bag.  The first several ml of
collected blood are diverted to the pouch and used for testing.

	This prevents skin contaminants and the skin plug often generated by
penetration of the needle from entering the collection bag.  These
diversion pouches are available in some apheresis sets, but are not yet
approved for whole blood collection systems.

	Unfortunately, skin disinfection and diversion pouches only reduce skin
and environmental contaminants. Several bacteria of importance are in
the donor's circulation and are not affected by these measures.
Detection systems appear, at first sight, to be the solution.

	In theory, bacterial culture and detection of bacterial growth could
resolve the problem of bacterial contamination of blood components, and
FDA has approved two such systems in recent months.  However, the
approval is specific for quality control, not for release of blood
components as free of bacterial contamination.

	Concerned about bacterial contamination, European blood agencies have
decided to adapt clinical laboratory culture systems to their
operations.  In The Netherlands, platelets from whole blood are prepared
by the buffy coat method, pooled, and a sample from the pool is placed
in a culture system with automated detection of bacterial growth.

	After 24 hours, if the culture is negative, the platelets are released
to hospitals.  If growth is observed in subsequent days, the hospital
and physicians are notified.  This method for the preparation of pools
of platelets from whole blood is not approved in the U.S.

	Hema-Quebec, our Canadian member, has implemented bacterial detection
systems for platelets collected by apheresis and is studying the
adoption of the buffy coat method for preparation of platelets from
whole blood.

	Among the 75 ABC member centers, 8 have decided to implement bacterial
detection systems in the near future.  They plan to use he ones
currently approved for QC of apheresis platelets, in a way similar to
that used by the Dutch.  These systems are complex and expensive;
moreover,  they cannot be applied in a practical manner to platelets
derived from whole blood.

	Current FDA regulations prevent us from pre-pooling platelets from
whole blood.  When pooled, they must be transfused within four hours,
even if the pooling is performed in a closed system, using sterile
connecting devices.

	For this reason, the implementation of bacterial detection systems
threatens the survival of platelets from whole blood.  Members of this
committee should be aware that there aren't enough platelets collected
by apheresis to supply the needs of the U.S. healthcare system.

	Last year, ABC members distributed about 550,000 platelets by apheresis
and 1.5 million units of platelets derived from whole blood, and it
would take several years to reach sufficiency if we were to convert
entirely to platelets by apheresis.

	In addition, many hospitals are resistant to the conversion, because of
the substantial cost differential between the two components.

	Some less cumbersome and less expensive approaches have been proposed
for the screening of random donor platelets for bacterial contamination.
 One is the use of a reagent dipstick for pH and glucose; a pH lower
than 7 and/or a glucose level of less than 250 mg/dl would be considered
indicators of bacterial contamination.

	The sensitivity and specificity of dipsticks is not yet fully assessed.
 In the past, some centers have screened platelets with a Gram stain
immediately before transfusion.  However, we know that these two methods
are much less sensitive than systems based on bacterial culture.

	In our opinion, a number of practical issues need to be dealt with
before restrictive standards or regulatory mandates are issued for
interventions designed to reduce the incidence of bacterial
contamination.

	The mode of application of the systems approved for quality control is
still unclear for us.  What would be the corrective actions triggered by
the finding of an occasional component with bacterial growth?  We can
think of personnel retraining and very little else.  How should we
interpret these findings from the QC point of view?

	Where should bacterial detection be performed?  At the blood center
where the components are prepared, or at the hospital, closer to the
transfusion event?  The requirements for the two approaches are quite
different.

	Detection at the collection facility requires high sensitivity, and
results obtained at the time of release of the platelets (consider the
five-day out-date) may not be predictive of the bacterial load at
expiration.  On the other hand, there are no reliable systems for
testing close to the transfusion event, when platelet concentrates are
pooled.

	One of the approved systems (Pall) focuses mainly on aerobes; the
manufacturer of the other is recommending cultures for aerobes and
anaerobes (BioMerieux).  Are cultures for anaerobes warranted,
considering that platelets are stored in gas permeable bags in an
oxygen-containing atmosphere?

	Recognizing that anaerobes are rare causes of clinical bacterial
contamination, and sometimes not detected in vitro until beyond the
expiration of the product, we do not think that use of anaerobic media
is a key initial part of this initiative.

	What inoculation systems should we use to prevent false positive
results?  Do we need to use laminar flow hoods?  How do we deal with
false positives?

	The Pall BDS is an endpoint system.  Specimens are inoculated 24 hours
after collection and the cultures read at least 24 hours later.  In
contrast, the BioMerieux bacT/ALERT is a continuous system, raising the
question of when cultures should be considered negative.

	Then the BioMerieux system is used and the cultures continue to be
followed after release of the platelets to a hospital, if subsequently
positive, what should physicians be told (since in many cases the
platelets will have been transfused)?

	To what level should centers or contract microbiology services identify
positives?  Is there a need for performance of antimicrobial
susceptibility assays? Probably yes.

	Could cultures be inoculated at the collection facility and read at the
hospital that received the component?  If so, how would specimens be
identified?  What software modifications are needed to assure
correlation between components and culture results?  In this case,  how
should reports of positive results be handled?

	How do we validate these systems?  What are the positive controls?  How
can we measure the efficacy of the detection procedures in light of the
low frequency of events?

	Finally, should a recommendation be made for implementing these very
expensive new procedures, there should be a consistent message to
hospitals and insurers explaining that their benefit far exceeds their
cost.

	Considering these and many other issues that need to be reviewed, we
respectfully request that this committee and FDA consider the following:

	Support collection facilities that implement methods to reduce the risk
of bacterial contamination by skin contaminants in blood and blood
components.

	Facilitate the licensure of bacterial detection systems for component
release, allowing claims such as "negative for bacteria at time of
release."

	Support collection facilities that decide to implement 100 percent
Quality Control for their apheresis platelets.  Their experience will be
invaluable for progress in this area.  Regulatory actions could have
serious inhibitory effects and delay the implementation of procedures
that will certainly increase the safety of transfusion.

	Encourage the development of alternative technologies for bacterial
detection that are less laborious, less expensive, and can be applied at
the hospital level, closer to the transfusion event.  This is essential
for the survival of whole blood derived platelets and for the
fulfillment of patient needs.

	Speed the regulatory process for the extension of the expiration date
of platelets to seven days if negative for contamination.

	Allow pre-pooling of platelets from whole blood using approved sterile
connecting devices, based on the long and successful European experience
with buffy coat platelets.

	We strongly believe that these actions will encourage the
implementation of means to reduce bacterial contamination of platelets
and hence increase the availability of safer platelets for transfusion. 
When we reach this stage of development, we will welcome FDA regulation.

	Thank you very much for the opportunity to present our point of view.

	DR. NELSON:  Thanks, Celso.

	T.J. Smith?

	MS. CROSBY:  T.J. Smith has asked me to give the presentation.

Cynthia Crosby

	MS. CROSBY:  I am Cynthia Crosby.

	As we go forward, I am going to skip through these slides really fast,
but I want to challenge the FDA Advisory Committee to understand the
modes and mechanisms of antimicrobial solutions in choosing what I am
hearing a plea from your Red Cross, the ABC to adequately assess skin
preparation prior to the venipuncture.

	[Slide.]

	Understanding antiseptic agents is very easy and readily available in
the texts that are out there.  I am with Medi-Flex.  We have been in the
business for 17 years of providing aseptic tools to deliver antiseptic
products to the donor site.

	Our bread and butter is in the donor prep market outside the United
States by them using alcohol followed by tincture of iodine.  Our bread
and butter in the United States is blood culture kits that provide
alcohol followed by tincture of iodine.

	[Slide.]

	Why is tincture of iodine superior to your current AABB recommendation
of povidone-iodine?  Povidone-iodine is complex with an inert polymer
that must dry for the polymer, must dry for the iodine to decomplex, to
have a kill mode.

	[Slide.]

	With tincture of iodine, you have readily available iodine.  It is
there in the position to have an immediate effect prior to your
venipuncture.

	[Slide.]

	We have had great demand from the blood donation facilities for a
chlorhexidine type based product whether it be tincture or aqueous.  We
are hearing that every day if this product has been approved by the FDA
to be used for blood donation.

	Right now we have several facilities that are running validations.  The
National Blood Authority in the UK is in the mode right at this time of
implementing a chlorhexidine/alcohol/tincture for blood donation.

	Carl McDonald presented at the International Blood Banking Conference
in Canada that stated that there was an equivalence to the tincture of
iodine two-step procedure, alcohol followed by tincture of iodine.

	In this, there was 2.76 log reduction with tincture of iodine versus a
2.6 log reduction with the chlorhexidine/tincture.  Over 99 percent of
the bacteria were killed at the time of the venipuncture.

	[Slide.]

	What makes chlorhexidine ideal is its affinity to bind to the skin
surface.  Ideal antiseptics should be broad spectrum, they have to be
rapid.

	[Slide.]

	They have to be rapid because of one key component, and that is
convenience and compliance to protocol.  As our staffing nurses are
reduced, our average age nurse is now 47 years old in our facilities,
and as phlebotomy teams are reduced, compliance and convenience to
protocol is necessary.

	Industry has met that demand by combining products that will have a
quick dry time, a quick kill.

	[Slide.]

	As of July 2000, there was an NDA approval for a new
chlorhexidine/tincture product, 2 percent chlorhexidine and 70 percent
isopropyl.  It has been demanded from the health care worker or
clinician for the care of catheter insertion and catheter lines.

	In 10 years, we have heard the data over and over again in the 50
percent reduction to catheter-related bloodstream infections.  We see it
now in the studies that are followed in the data that I gave you as a
package that shows the 50 percent reduction by going to use a
chlorhexidine/tincture prior to a vascular procedure.

	[Slide.]

	Povidone-iodine is your current prep.  It's a two-step procedure. 
Tincture of iodine and CHD are far superior in the clinical and in the
in vitro data to povidone-iodine. Tincture of iodine is a two-step
procedure.  It is proposed for the standards of blood banking and
transfusion.

	Two percent chlorhexidine based products that are currently available,
which only happens to be one, is a one-step procedure.  It is equivalent
to the two-step tincture of iodine.  It is proposed as an alternate
prep.

	One thing I can state is that we see huge compliance to protocol in the
hospital community.  New CDC guidelines for the prevention speaks I
think, and I am going to go out on a limb here, to your donation
facilities, that is, educated and dedicated resources are necessary to
reduce your contaminations.

	Maximal barriers and the understanding of maximal barriers are
necessary in reducing contamination, and 2 percent chlorhexidine-based
antiseptics are preferred, they are the preferred antiseptic of choice
by the CDC, and they have superior efficacy compared to povidone-iodine.

	[Slide.]

	Preventative measures are the highest quality in medical treatment and
the most cost efficient.  They are simple.  We just have to make the
decision to use those preventative measures.

	Thank you.

	DR. NELSON:  Thank you very much.

	Pall Corporation also had a statement to make, but they have decided to
submit it for the record rather than to read it or present it at the
meeting at this time.

	We are no longer in the open public hearing.

	Maybe we can go to the questions.  John.

Questions for the Committee

	DR. LEE:  I guess I will just read the question one more time regarding
donor arm prep.

	Do available scientific data support preferential use of an
isopropanol/tincture of iodine procedure for preparation of the donor's
phlebotomy site?

	I guess I would just comment that some of the comments we heard during
open public hearing session just now didn't really refer back to some of
the material that I presented in the morning.  The fact that
isopropanol/tincture of iodine is proven to be superior in the donor
setting seems a bit of an overclaim in view of what I presented.

	DR. SIMON:  Is it appropriate to start the discussion?  I guess it is
an interesting topic because I have a long-standing interest in it, and
actually, data that were not presented here, but there was a study done,
the United Blood Services in 1993-94.  It was only presented in abstract
form and it is referred to in the Goldman paper, and it showed that
tincture of iodine was superior in eradicating organisms from the skin.

	I actually wrote the AABB at that time and proposed that we make the
change then, and that obviously did not occur.  So, I think that there
is data and I think there is data in the blood culture literature, as
well, to indicate that looking strictly at data, it would appear that
tincture of iodine is superior to the povidone and that, in addition,
chlorhexidine is also superior.

	Actually, we use chlorhexidine in our laboratory for our blood
cultures, but when I questioned our microbiologist on why we made that
choice instead of tincture of iodine, it was because we have a lot of
these cultures drawn by non-dedicated personnel, not personnel that we
control, intensive care, emergency room nurses, and so one-step
procedure, which the chlorhexidine is, is superior or we feel we get
superior compliance and went with that instead.

	I think there is data to support the change as AABB plans.  I must say
that I was impressed with your presentation and I think from the common
sense point of view, one could well raise the question, does it really
make a difference in the donor setting where you have so many other
things going on, and you have sort of a multifactorial situation.

	But I would guess if we are this concerned about this subject to be
willing to move to some of these other steps that we are going to
discuss, I would think, as a first step, that we would want to have the
most effective removal of organisms from the arm to begin with when the
unit is drawn.

	I think that the tincture of iodine, as AABB has proposed, with the
chlorhexidine as the backup for iodine-sensitive individual, that would
be the step to go, so I would support this.

	I would put into that statement, however, that I don't think the plasma
industry should be required to go this direction because none of their
products are stored are either refrigeration or room temperature.  They
are frozen,  and they have not had a problem with bacterial infection in
that industry.

	So, I would think that they should be okay to go with either one, but I
would support the AABB position at this time.

	DR. ALLEN:  Thank you, Toby.  I think that is a good introduction to
the discussion.  I have been mulling over,  given the way in which the
question is worded, whether I would vote yes or abstain.

	My feeling is that there is some evidence that tends to lean towards
tincture of iodine in isopropanol, or the combination, the two step, as
clearly the preferred way to go although I would have real trouble given
the broader view of 2 percent chlorhexidine in isopropanol as an equal
alternative.

	I guess my question is what is the implication of voting yes for this
in terms of action that the FDA might take given that the AABB already
has revised proposed standards out that use this with the chlorhexidine
as an alternative.

	So, if we vote yes for this, what are the implications in terms of how
the FDA is going to use this information?

	DR. LEE:  That's a million dollar question.  Dr. Epstein, would you
like to field that question?

	DR. EPSTEIN:  The question before the agency is whether we should issue
a regulatory guidance recommending preferred use of isopropanol/tincture
of iodine.  It has occurred from time to time that we disagree with an
industry voluntary standard, and then we may take a regulatory position
to override it.

	So, in this case, our options would be to remain neutral and let the
preference be voluntary or to endorse it through a regulatory
recommendation.

	DR. ALLEN:  I would certainly feel comfortable and I would vote yes if
it is to be a voluntary recommended standard.  I am much less
comfortable I think to take it to the regulatory level.

	DR. EPSTEIN:  Well--okay.

	[Laughter.]

	DR. NELSON:  Actually, this committee doesn't exactly make regulations.
 We are supposed to help the FDA evaluate data and evaluate what is
there and what is missing.  So, I don't think you need to consider
yourself a judge or a congressman at this point, I guess.

	DR. KLEIN:  I appreciated Dr. Lee's very thorough review, but I must
say I sort of discounted the studies that simply showed that you can't
culture something or you culture something less frequently from the
site, because that really isn't a good demonstration of what might be
the risk in the bag.

	Actually, I think you could be fairly badly misled by those studies. 
What you really want to know is what the risk is of contaminating the
component.  We don't have those studies unfortunately, and I think the
best that we have are the blood culture studies which by and large show
that all of these preps are just about the same, I guess, with the
exception of soap, which I don't think anyone supports.

	The problem with those studies, of course, is that the background level
is so high that it could mask slight differences between different arm
preps.

	So, I don't have any problem with saying that this is a good
preparation, I guess as the AABB is suggesting in its standard.  I would
hate to see anything regulated based on that, however.  I don't think
the data support it.

	DR. NELSON:  In these studies, apparently what happened, the way they
did it is they cultured before the preparation and then they applied the
material and then recultured.  It was stated as a percent reduction of
flora.

	DR. LEE:  Well, many of the blood culture studies were comparisons.

	DR. NELSON:  Not the blood cultures, the McDonald.

	DR. LEE:  The blood collection studies, yes, it was cultured before and
after.  Again, when represented as percent reduction, the values don't
strike out at you as much, the difference between 99 and 98, I don't
know what that really means.

	DR. NELSON:  Mary.

	DR. CHAMBERLAND:  I guess I am puzzling over this, too, because I guess
what are the criteria, what is the level or rigor that is going to
demonstrate that one of these preparations is "preferential."

	I agree with Dr. Klein's comments that log reduction of bacteria skin
contaminants is not necessarily at all the whole story.  I just wanted
to make sure because it was a quite a load of literature that the
committee was given to review in advance, in a relatively short period
of time, so I think it is really hard for us to basically digest all of
this.

	I do want to commend you because I think you did really put together a
really nice, critical review.  As I understand it, there are no data for
this proposed prep that look at contamination rates in bags of
platelets, that is correct, it is just all skin studies.

	DR. LEE:  That is my understanding.  The thing that comes closest is
the third study that I described, within which cultures of actual
platelet units were done, but that study did not include the tincture of
iodine/isopropanol method.

	DR. CHAMBERLAND:  So, as a surrogate, then, you brought up in some of
your slides some reviews of literature that is available on looking at
blood culture data.

	DR. LEE:  Right.

	DR. CHAMBERLAND:  And the various preps.  Again, I want to make sure I
have got the bottom line here straight, but in looking at the blood
culture data, were there any clear winners here?  This
isopropanol/tincture of iodine did not emerge in the blood culture
studies as showing a preferential--

	DR. LEE:  No, it did not.  There are some conflicting results.  Some
showed marginally better performance, but there were many others,
equally as many, that showed no difference.

	DR. CHAMBERLAND:  I think it comes down to--again, this is what I am
struggling with--if FDA is signaling that they are potentially
considering guidance, then, you usually have to put forward what your
criteria are that you are going to evaluate.

	I guess a fair question is if the only data for this new prep are skin
culture data, is that adequate, is that the whole story, would you
require more rigorous data.

	In the end, at the end of the day, how much of a difference is it going
to make.

	DR. LEE:  I see Dr. Epstein at the mike.

	DR. EPSTEIN:  Maybe I can shed some light.  I think Dr. Lee said this
earlier.  What has happened here is we have had a series of workshops
over a period of years.  At every single workshop or symposium, someone
shows the studies on the isopropanol alcohol/tincture of iodine prep and
says, look, this reduces bacterial contamination of the skin.

	We all agree that we don't have a good endpoint study in platelets, and
the FDA has not come forward in the face of that with a recommendation
for preferred use of that procedure because we keep taking a look at the
larger literature and scratching our heads and saying we are not ready
to do this.

	On the other hand, the AABB has now, listening to the same workshops,
the symposiums, reading the same literature, has decided to make this a
voluntary standard.

	Now, there are two positions that the FDA could take.  One is we agree,
so we will make it a regulatory standard, or we are not so sure, but we
are not opposed to the industry voluntary standard.

	Again, and I guess this comes back to you, Jim, we are not asking you
to decide the regulatory position.  We are just asking you whether you
are, in effect, in agreement with the FDA that the science is too soft
for us to take a regulatory position, because after all, the pressure on
us has been to do just that.

	DR. SIMON:  I just will say a few more words in support of my position,
I guess, for a yes vote on the question.  I think we have to remember
that what we used to cleanse the skin was based on this kind of
analysis.  That is, we wouldn't have any prep at all as far as I know if
we had as an endpoint, the desired one, of the infection of the unit.

	So, we have sort of empirical choice of agents that remove bacteria
from the skin, because that is what we want to do with this step of the
procedure, and the data would indicate that tincture may be superior.

	I was a little surprised at the skepticism about the blood culture
studies because the reason we did the study at UBS back in 1994 was the
literature on the blood culture studies, which indicated that povidone
was inferior.

	I might be somewhat influenced by this 1999 study by my former mentor
at medical school from Barnes-Jewish, which showed tincture of iodine to
be superior, and I don't believe that we would use povidone in our
laboratory for blood cultures at this time.  We would use either
chlorhexidine or tincture of iodine.

	So, I think that there is enough data there given that what our
objective is with this step of the procedure is to remove bacteria from
the arm.  We may have to do other procedures to make sure that the
platelets are absolutely sterile, but that would seem to be the most
reasonable first step.

	DR. FITZPATRICK:  I would agree with Toby, but prior to making
regulation, because really the only data we have for regulation is from
the Vox Sang article, and since the tincture of iodine wasn't in that,
there isn't a comparison to use, so you would need a comparison.

	But as far as just surface contamination, I think there is ample
evidence, and we did used to use tincture of iodine, and the biggest
complaint was that techs and nurses were getting cut in their fingers
when they broke the ampule initiative the tincture of iodine swab that
was used to prepare the arm, and now there is a method where you don't
have to break a glass ampule to do that.

	Since it just says preferential, it is pretty soft as it is.

	DR. LEW:  I think maybe to be fair, to be more specific, though, 
because everyone is struggling with the fact, I think, that there is no
good studies looking at many different types of preparations and saying
that this is The superior with a capital T.

	But if you can just say preferential, the tincture compared to what is
used now as a recommended standard, I don't think anyone would have a
major problem with that.  It is just trying to imply this is the best
overall.  The data is not there, and it is not appropriate to I think
regulate based on no data.

	DR. NELSON:  Are we ready to vote on this?  It is now an open public
hearing, but if you have got something critical that we haven't
considered.

	DR. WAGNER:  Just a cautionary note, and that is of placing too much
emphasis on whole blood cultures done at hospitals on patients who may
already, some fraction of which would be bacteremic, to compare two
different skin preparation methods.

	The bacteremia in the population may overwhelm the differences.

	DR. NELSON:  The difference of differentiating a contaminant from a
real infection.  We have recorded that.

	DR. ALLEN:  Would the FDA consider an amendment, if we add at the end
of this, the sentence as written, "for preparation of the donor's
phlebotomy site compared with the current procedure using
povidone-iodine"?

	That compares this one versus that one without making a statement about
any other preparation.

	DR. LEE:  We could do that, interpret the question in that light.

	DR. NELSON:  Right, since that is what is used, I guess.  Okay.

	DR. SMALLWOOD:  Would you repeat that?

	DR. ALLEN:  Do available scientific data support preferential use of an
isopropanol/tincture of iodine procedure for preparation of the donor's
phlebotomy site compared with the current standard procedure based on
povidone-iodine preparation?

	DR. SMALLWOOD:  Voting will be by roll call as required.

	The question as amended reads:  Do available scientific data support
preferential use of an isopropanol/tincture of iodine procedure for
preparation of the donor's phlebotomy site compared with the current
standard procedure based on povidone-iodine preparation?

	Allen.

	DR. ALLEN:  I vote yes and I think we need to look at other
preparations, and we need to study the donor acceptability of the
tincture of iodine.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  I am going to vote no.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  No.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  Yes.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  No.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Stuver.

	DR. STUVER:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  No.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  No.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon, you would--

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  The results of the voting.  There are 6 yes votes, 7 no
votes, no abstentions, the acting non-voting industry representative
agrees with the yes vote.

	DR. NELSON:  Now that we have solved that, do we have a second
question?

	DR. WILLIAMS:  Just a very brief clarification.  Like this question,
several of the other questions deal with issues in which there is a
developing industry standard, as well as a potential developing
regulatory policy.

	Several of these initiatives are draft standards. They need to go
through a membership comment period and final acceptance, so just to
clarify, these are not existing industry standards at this time.

	Committee Question 2.  Do available data on the sterility of the
sterile connecting device procedure support the use of this procedure to
collect samples for bacterial detection from in-date platelet products?

	DR. NELSON:  Discussion?  Yes.

	DR. SIMON:  I think the data were fairly overwhelming to answer this
yes, however, just with the caveat that we are introducing another
complexity into the system, which at a breakdown at some point, could
lead to more bacterial infections than what we have now, but I think
certainly the data we are presented support it and obviously, the FDA
has approved this for products that remain in-date and can be infused.

	So, I would think that we would vote yes on this one.

	DR. NELSON:  When it breaks down, that will lead to another question
later on.

	Are we ready to vote?

	DR. SMALLWOOD:  Question No. 2.  Do available data on the sterility of
the sterile connecting device procedure support the use of this
procedure to collect samples for bacterial detection from in-date
platelet products?

	Allen?

	DR. ALLEN:  Yes.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  Yes.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  Yes.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  Yes.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  Yes.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  Yes.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Stuver.

	DR. STUVER:  Yes.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon?

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  We have unanimous yes on the vote for the second
question.

	DR. WILLIAMS:  Does the committee concur with FDA's proposed
statistical approach to providing quality control for platelet
contamination?

	DR. NELSON:  Discussion?  Toby.

	DR. SIMON:  Well,  I have a real problem with this, but in a way I
think Dr. Epstein clarified it as a method for the agency to be certain
that the regulated entities are performing their functions as they
should.  I guess it is something to go along with.

	I just don't think that this kind of approach will ultimately improve
safety, and it is going to cost a lot and small organizations will have
to wind up culturing all their units.  So, I think it probably logically
leads you to what the AABB is proposing, which is a release criteria,
but I don't see anything wrong with what the FDA has proposed other than
I think it is going to be a lot of data collection without a lot of
value from it.

	DR. FITZPATRICK:  A comment and a question.  I think Toby's comment
about the small facilities having to culture every unit is very valid. 
In the past, FDA has attempted to find a means of providing those
smaller facilities an alternate method, and I think it is essential that
there be an alternate method for them especially if they demonstrate
they are in control over a period of time.

	The other question I have is when a facility finds itself above the 0.2
percent level, are they then to restrict the release of products until
they sample enough products to come below the 0.2 percent level, and
what is the impact on supply of that.

	DR. WILLIAMS:  That is a good question.  If a process is deemed out of
control, I think under normal circumstances, one would basically curtail
production, reassess, and revalidate all the processes and continue,
revalidate and resume production as soon as possible.

	But I think to continue manufacture of product for a procedure that is
significantly not meeting a current standard, it would probably be
arguable that production should continue.

	There is a supply issue and I think probably in that situation,
approaching FDA for a potential solution is probably the best impact. 
Jay may wish to comment, as well.

	DR. EPSTEIN:  Well, typically, industry establishes alert levels and
action levels, and I think the discussion hasn't gone that far.  It may
depend what measure you get.  You know, if the estimated rate is 0.4
percent, that is not the same thing as if it was 0.25.

	I don't think we have all the answers there, but I think that Alan's
answer is the correct one, a system that is clearly out of control
shouldn't continue, and it does throw into question the quality of the
products released.

	But that said, more work is needed, more thought is needed to figure
out what the appropriate actions are at any given level.

	DR. SCHMIDT:  I was struck by Dr. Bianco's comment about what do you do
with the information, and the thing I thought back to is a recent event
where a patient who had multiple transfusions, suffered from I think it
was a Klebsiella, and the result of this, I just read the newspaper
account, was shutting down the blood for the whole state more or less,
and some strange things were said in the press.

	I guess what I am saying is when you get to this type of iffy things
and you are looking at the quality of the entire place based on
something that may or may not relate, this can be a pretty dangerous
situation.

	DR. FALLAT:  I was impressed with the fact that there is such a
variation in the literature and you really can't give a standard right
now, so I am not sure how we can set up an industrywide standard when we
don't know what the standard is.

	I was also impressed with the comments made by the CDC representative
that said that there may be several questions that need to be answered,
and I think it would be much more reasonable that this be done as a
pilot or a study to see what your response is with a certain number of
centers that might be willing to cooperate in such a study rather than
making it an industrywide quality control approach at this time.

	DR. ALLEN:  Thank you for saying that.  I concur.  I like this
approach.  It is certainly one that is analogous to what is used in
industry for quality control.  I think it ought to be pursued.  As Dr.
Epstein said, there is still a lot to be worked out on it, and I guess I
am struggling how do I vote to encourage the further development of this
and perhaps a pilot testing of it without indicating that I think it is
anywhere close to being ready to be put into a regulatory mode.

	DR. KLEIN:  I don't think that this in any way as described is going to
improve public health.  First of all, we don't know what the right
number is, and it is more likely that it is less common, which would put
the facilities at risk of being out of compliance for no reason of their
performance.

	Second, as I stated earlier, a high percentage of the endpoints that we
are looking at are not going to be addressed by what we do when we
prepare platelets.  They are from organisms that are circulating in the
donor.

	Point three, I think, is that what we are really looking for is a
release criterion, and you obviously can't do that because the testings
are not licensed for that, but I think that this is not a good
compromise.

	I think it will be laborious and potentially affect supply, and not
improve public health in terms of bacterial contamination of platelets.

	DR. NELSON:  Is the Red Cross going to uniform culturing of platelet
donors?  If so, that could end up being a pretty good database to decide
what the level should be.

	DR. DODD:  Ken, I take it I can comment on this?

	DR. NELSON:  Yes.

	DR. DODD:  The Red Cross is considering whether or how it could engage
in 100 percent quality control of apheresis platelets.  That doesn't
mean that we are actually going to do it although I think there are
other blood agencies that are moving down that track, and I think it
would give an opportunity for collecting appropriate data.

	Ultimately, if and when we do that, we would like to do it in some way
that prevented the usage of a bacterially contaminated product.

	But I did want to take the opportunity of asking either the agency or
the committee if it considered the fact that as of today, the two
methods that have been approved for quality control, they are approved
only for leukoreduced platelets.  I wonder what impact that might have
on answering this question.

	DR. VOSTAL:  I can actually address the question about leukoreduced
platelets.  The reason those devices are approved for that product is
that is because that is the only type of data we saw when we were
clearing the device.

	DR. FITZPATRICK:  I have Dr. Allen's same dilemma as to how do we
answer this question.  As I understand Dr. Epstein's desire for a way to
measure compliance and encourage compliance, but we need a stepwise
approach to that.

	One of the key elements that I think came out today was the pooling
random donor platelets and being able to maintain them as a pooled
product, and then test them.  I think that would add to what we can do
to ensure and reduce bacterial contamination for the patient, if we have
a way of quarantining and release or testing prior to release.

	As Dr. Epstein said, there are a number of unanswered dilemmas here
with this, and the approach taken for the leukoreduction guidance, I
think applies here, but because of the low incidence of positive units,
the N becomes so large that it is very difficult to apply that
statistical model to this.

	I think future exploration and an alternative method to find a way to
establish compliance and standards, and then monitor on a periodic basis
might help, but I find it really difficult to say I concur with this as
the appropriate approach at this point.

	DR. CHAMBERLAND:  I just find myself in the same dilemma.  I would hope
that the committee really wants to signal their strong endorsement as
has been echoed by industry that the time is now to really take
bacterial contamination seriously and to take steps in that direction to
prevent and then appropriate monitoring.

	I don't feel at this point comfortable enough that this proposed
approach, exactly what it will accomplish, and will it incur any real
risk related to supply, et cetera, so I am concerned that if I vote no,
that that might be viewed as not feeling that this is an important
problem.

	We are being asked to really vote on a really detailed, specific plan,
and I think it takes a bit of time to sort of model that and work that
out, and I just wonder if maybe you have all done that and done some
real live kind of testing of this from a modeling perspective to see
what might happen, but I just a little bit uncomfortable that a yes vote
and develop a guidance and it's out there without thoroughly
understanding the implication.

	So, I feel somewhat caught here a bit.

	DR. SIMON:  I guess a question and a comment.  Alan, could I ask, you
are proposing this or FDA is proposing this for both apheresis and
random donor, that is correct, isn't it?

	DR. WILLIAMS:  That's correct.

	DR. SIMON:  One other possibility, and I don't know how FDA feels about
this, would be to table this question for now and at the next meeting,
discuss this issue of quality control and the release issue together,
and see which is the best direction to go to reduce risk, because I
understand that AABB had some of these same kinds of discussions when
they came up with their proposal.

	Since industry is moving towards a sort of different approach for
release, and since it raises all kinds of implications, and I think
there, really where the supply issues get raised, both in terms of units
being held and also in terms of what happens to random donor platelets
in this setting, and are blood centers going to consider it too
hazardous to provide these because they are not amenable to the same
approach, so it has become a very complicated issue.

	Although I don't see anything wrong with what FDA has proposed, it
appears that it may need a more comprehensive discussion and taking into
account both this approach and the release approach.

	DR. DiMICHELE:  I just wanted to echo the comments of several of the
committee members.  It seems to me, although I certainly defer to those
members of the committee who are blood bankers and have had tremendous
experience with this, but it seems to me that the issue of how we
document sterility of platelets is still unclear.

	So, therefore, it becomes very difficult to embark on this type of a
study if we are really not clear on how we are really to measure
platelet sterility in the first place.

	I would agree with those committee members who suggest that maybe the
next step is really to develop a pilot study to really help us to
understand how to measure platelet sterility before we embark on a
widescale regulatory quality control measure that actually attempts to
do this.

	I don't know again the best design for this, but I know that there are
many units of platelets, for instance, that would get discarded anyway
because of hepatitis B core positivity, et cetera, units that would not
be used, and the question is, is whether a study can actually be
designed using units that cannot be salvaged, to start looking at this
in kind of a pilot way.

	DR. NELSON:  There have been a few pilot studies on cultures, and they
are not large enough.  The issue is that to really get this number and
to figure out what is acceptable would take a substantial size study I
think.

	DR. DiMICHELE:  Well, that is if it were a natural history study, but
one of the things that you can do with platelets that wouldn't be used
anyway would be spiking experiments and manipulative studies that might
be able to give you data in a very different way.

	Again, I certainly can't comment on being an expert on the design of
such a study, but it might be considered.

	DR. NELSON:  Are we ready to vote on this one?

	DR. SMALLWOOD:  Question No. 3.  Does the committee concur with FDA's
proposed statistical approach to providing quality control for platelet
contamination?

	Allen.

	DR. ALLEN:  I abstain with strong support for the FDA's continuing its
work in this area.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  I also will abstain.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  No.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  No.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  No.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  No.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  No.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  No.

	DR. SMALLWOOD:  Stuver.

	DR. STUVER:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  No.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  No.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  No.

	DR. SMALLWOOD:  Dr. Simon?

	DR. SIMON:  I would abstain.

	DR. SMALLWOOD:  Results of voting on Question 3. There were 11 no votes
and 2 abstentions, and the industry representative took an abstention
position.

	DR. VOSTAL:  We will move on to Question 4.

	The question reads:  Does the committee concur that data derived from
FDA's proposed clinical trial design would be appropriate to support
clearance of devices for pre-release screening of platelet products for
transfusion?

	DR. NELSON:  Discussion?

	DR. FITZPATRICK:  Of the many designs that we were offered, which one
would you like us to endorse?

	DR. VOSTAL:  I think the basic concept of whether we should require a
clinical trial to evaluate these devices and whether the clinical trial
should be of design where you have two cultures and you are looking for
agreement between a culture early on in the storage period versus at the
end of the storage period.

	DR. FITZPATRICK:  So, would you rephrase your question then?

	DR. VOSTAL:  I guess we could, if you don't like that one.

	DR. KLEIN:  If we answer yes to this, is it pretty nebulous.  If you
want to know if we think there should be a clinical trial, designed
appropriately, I think that is a legitimate question, but it is hard for
us to pick the appropriate design at this point from what you are
offering.

	DR. VOSTAL:  So, we could change it just to say whether the committee
would endorse that a clinical trial is necessary, and the second part of
that question would be whether it should be of a design with the two
cultures.

	DR. KLEIN:  I think that that would answer the question, and I think if
you are going to do it, you need to do it with two cultures.  Having
said that and as a federal employee who looks at our budgets, I think
this is going to be a very large study.

	I think it is going to take a very long period of time, and I think it
is going to be a very expensive study. I look at the culture technology
really as a transient technology.  We all want something that we can use
at the endpoint of issue, and my guess is that in three to five years we
are going to have something that we will be able to do.  Maybe by the
time the study gets done, and we have the results of the study, we won't
use this technology.

	So, having said that I think you need a study in order to license this
for that use, and that this study would be appropriate, I wouldn't
participate in the study and I wouldn't spend any money on it.

	DR. ALLEN:  Thank you, Harvey, that's a hard one to follow.

	I certainly am strongly supportive of the general concept.  I think
very definitely, clinical trials of in-use situation, using real live
materials to the extent that that is possible is very helpful to support
the evaluation approval of devices for this, as well a variety of other
purposes.

	I think the study designs that were presented were probably the purest
in terms of sampling at the beginning and at the end, and maybe at a
couple of mid-points, and I in general would be supportive.  I am not
quite as negative as Harvey on this.  I agree that they will be
large-scale studies and possibly fairly expensive and difficult to
accomplish.

	I think they could be important even in an area where technology will
be evolving rapidly.  Nonetheless, I think that the FDA, in terms of
putting out trial designs, needs to be flexible, because I don't think
that in every instance that you have got to have exactly the same design
and all of the bells and whistles.

	I am supportive of the general concept.  We did hear that there were
many different clinical trial designs that were proposed, and I would
urge flexibility on this, but I think the concept is one that I
certainly am supportive of.

	DR. FALLAT:  Am I correct in understanding that one of the outcomes of
this would be to get more data on the five- to seven-day platelet
release and capability of release, and if so, I would be very much in
favor of that sort of study design.

	DR. VOSTAL:  I think that it depends on the intent of the study,
whether you want to have your device to be used for extension of the
storage period, so that it will be designed to look at day 7 platelets.

	DR. SIMON:  I would agree and being supportive of clinical trials and
design, but would also agree that this is going to be, the kinds of
numbers we are talking about, it is going to be very costly.

	Unlike Dr. Klein, I would do it if somebody gave me the money, but I
think it would be a major task to do. What I am concerned about, because
I think it would be advantageous to go to seven-day platelets having
done the studies in the '80s to support that before the bacterial
contamination issue became in the foreground.

	I would certainly like to see something done that would support
seven-day platelets.  From what I am gathering from your comments,
somebody would need to do this type of study to get permission to do
that extension or to get clearance from the FDA to sell seven-day
platelet or to label their platelet product as being good for seven
days.

	I would certainly like us to find a less severe way to get to that
point, because what I am afraid is going to happen with AABB instituting
a voluntary standard for release of all products, and if the
manufacturers see that their devices are being used on a national basis,
they may not have an incentive to move to do this type of study.

	So, I have the same concerns the other members of the committee have,
but I certainly am supportive of doing clinical trials.  I would hope
that we could work through a model that wouldn't be quite as expensive
to do.

	DR. DOPPELT:  In regards to the cost, how would this be funded?  Would
the banks be required to sort of absorb the cost?

	DR. VOSTAL:  I think we would be looking to the manufacturers to
sponsor their studies because it will be for their device.

	DR. CHAMBERLAND:  Jim, did you propose a wording change, or it was just
kind of--and I tend to agree with you that the idea of a study is
something that I am in complete agreement with--but I would say that I
am not sure we have the design that has been articulated and that there
might be a need for some flexibility?

	DR. ALLEN:  You could perhaps add one word, the FDA's proposed clinical
trial design concept.

	DR. VOSTAL:  Sure, that would make it plural.

	DR. NELSON:  Are we ready to vote on this?

	DR. SMALLWOOD:  I need to have the correct wording.

	DR. VOSTAL:  Clinical trial design concepts would be appropriate.

	DR. DOPPELT:  Is part of the proposed study that would give data to
extend from five to seven, that is or is not included in the current
proposal, because that was one of the four choices, so I just want to be
sure what we are voting on.

	DR. VOSTAL:  There are several issues there.  One would be having a
device for release of platelets up to five days.  Another one would be
to have a device for release of platelets up to seven days.  Or you
could have a study that covers both, pushing storage out of seven days.

	So, I think it depends on what the manufacturers want to have on their
label and what kind of study they are willing to sponsor.

	DR. SIMON:  I take it we can take it for granted that somebody who came
to you and asked to have extension of platelets to seven days, which I
think people like Dr. Aubuchon are doing based on the bacterial
detection device, you would not permit that with a culture at 24 days,
and you would have to have follow-up studies at seven days.

	DR. VOSTAL:  Right.  I mean if you wanted to be able to say that your
device is capable of detecting or making sure that you have a
culture-negative product at seven days, and you are going to be sampling
at 24 hours, we would like to see data that supports that.

	DR. FITZPATRICK:  So, the key to the design of the study and the
concept would be that you want a study designed to correlate the time of
sampling and the result of that sampling to the bacterial condition of
the product at the time of its out-date.

	DR. VOSTAL:  Right.

	DR. FITZPATRICK:  I think I could support that pretty well.

	DR. SMALLWOOD:  Question No. 4 as amended.  Does the committee concur
that data derived from FDA's proposed clinical trial design concepts
would be appropriate to support clearance of devices for pre-release
screening of platelet products for transfusion?

	Roll call.  Allen.

	DR. ALLEN:  Yes.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  Yes.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  Yes.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  Yes.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  Yes.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  My answer to the question is yes.  My caveats are in the
record.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Stuver.

	DR. STUVER:  Yes.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  The results of voting for Question No. 4, as amended,
is a unanimous yes.

	DR. NELSON:  The next subject is Human Parvovirus B19 NAT Testing for
Whole Blood and Source Plasma.

	Dr. Yu will give an introduction and background.

	DR. SMALLWOOD:  We are now approximately an hour and 23 minutes behind.
 So, maybe if we move quickly, we may be able to meet our goal of 6:30
in completing this.  If not, we will have a continued session regarding
this discussion at a later meeting that we will announce.

II.  Human Parvovirus B19 NAT Testing for

Whole Blood and Source Plasma

A.  Introduction and Background

Mei-ying W. Yu, Ph.D.

	DR. YU:  Hi.  So, now the topics of the discussion is Parvovirus B19
NAT for Whole Blood and Source Plasma.

	My name is Mei-ying Yu.

	[Slide.]

	I will provide the introduction and background and then Dr. Kevin
Brown, who is a B19 expert from NIH, will present the overview of
parvovirus B19 infection.  Then, there will be industry data
presentations.

	They will update data presented at the December 2001 FDA NAT Workshop,
and they will provide data for NAT sensitivity, testing algorithm, time
to resolve to single donations or donors, prevalence and levels of both
B19 DNA and anti-B19 antibodies and profile in serial bleeds, if any.

	The presentations will be made by, first, Dr. Susan Stramer of American
Red Cross, and then will be NGI, Dr. Andrew Conrad.  However, Dr. Andrew
Conrad is sick and he cannot make it here, so I think the committee has
all his slides, so there will be no one to present his.

	Then, there will be consolidated data presentations organized by the
PPTA.  There are three speakers:  Dr. Barbee Whitaker, Dr. Steve
Petteway, and Dr. Ed Gomperts.

	Then, I will come back to address the FDA Perspective and Questions for
the Committee.

	[Slide.]

	Now, I have a lot of background information I need to cover.  We have
talked about B19 NAT issues in quite a few meetings.  They include the
BPAC held in September 1999, FDA NAT Workshops, and NHLBI Parvovirus B19
Workshop both held in December 1999, FDA NAT Workshops held in December
2001 and then the BPAC in this year, March, and another one is the ad
hoc PHS Panel Committee Meeting held in July 2002.

	Parvovirus B19 has been extensively discussed in September of 1999
BPAC.  BPAC agreed then that pending a policy on screening whole blood
donations, FDA need not require studies to validate the clinical
effectiveness of NAT for B19 DNA under IND for plasma for further
manufacturing.

	So, parvovirus B19 NAT was considered as an in-process test, so it is
unlike HIV, HCV, and HBV NAT, as a donor screen test, however, BPAC did
not recommend resolving to the single donation or donor.

	For S/D treated pooled plasma, the reactive 20 unit subpools were
discarded when tests completed, labile components had expired.

	[Slide.]

	At that time, BPAC did recommend to quarantine and destroy in-date
units when possible.  So, FDA requires that the tests be reviewed under
biologic license application--that is called BLA mechanism--for the
manufactured product and that the tests be validated as analytical
procedures with respect to sensitivity, specificity, and
reproducibility.

	[Slide.]

	At both the FDA NAT workshops and NHLBI Parvovirus B19 workshop held in
December of 1999, a strategy for standardizing B19 NAT was outlined. 
FDA also proposed a B19 DNA limit that is less than 104 geq/per ml for
manufacturing pools.

	This limit was mainly derived from the B19 transmission associated S/D
treated pool plasma in a Phase 4 study in healthy donors.  In those
product lots that have less than 104 genome equivalents, no B19
transmission in recipients.  These are in zero-negative recipients.

	The residual virus will be complex or neutralizing by anti-B19, always
present in large pooled products, and also the manufacturer procedure,
that is our viral clearance procedure to remove the residual B19
infectivity.

	I want to mention one more point.

	[Slide.]

	We subsequently revised the limit.  The limit was then set as 104 IU/ml
because in year 2000, the WHO standard for B19 NAT and CBER working
standards for B19 DNA are all available, so we have since then revised
the limit to less than 104 IU/ml.

	That limit seems to be technically achievable by most manufacturers.

	[Slide.]

	Fractionators are performing high titer B19 minipool NAT screening by
in-house methods to lower the viral load in manufacturing pools.  In the
December 2001 FDA NAT Workshop,  we learned that the sensitivity of NAT
assay used to exclude donations ranged from 105 to 107 geq/ml.  That is
per original donation.

	The reactive minipools are resolved to single donations.  Testing
results are used to reject reactive donations.  Now, today, you will
hear more the update by the industry speakers.

	[Slide.]

	Establishments collecting whole blood units that are used to prepare
the recovered plasma and transfusible blood components would like to
implement high titer B19 NAT screening similar to that used by source
plasma fractionators.

	So, Dr. Susan Stramer did present some retrospective data and some
study data, and she described the Phase 1 approach that is not resolved
to single donations, labile components would have expired, and in Phase
2 approach, that would resolve to the single donations by a
free-standing test kit.

	Now, Dr. Sue Stramer is going to update the data later.

	[Slide.]

	In March 2002, BPAC, FDA's current thinking on B19 NAT was presented. 
The recommendations that FDA is considering are the following.  For
plasma, when identified, high-titer B19 reactive units should not be
used for further manufacturing into injectable products.  This is to
ensure that the FDA proposed limits less than 104 IU of B19 DNA/ml for
manufacturing pools can be met.

	Now, for whole blood, we say when feasible, B19 reactive minipools
should be resolved to identify the individual reactive donors prior to
release of the component for transfusion, and units from reactive donors
should not be used for transfusion.

	For whole blood, when testing is done subsequent to product release,
in-date components from potentially reactive donors should be retrieved
and discarded.  Even when performing an in-process test, testing and
identification of the individual reactive donors constitute medical
diagnostic testing, therefore, such testing would require the use of an
FDA-approved investigational mechanism.

	[Slide.]

	Informed consent should be obtained from blood and plasma donors
subject to such high titer NAT testing.  Reactive donors should be
identified and be informed of their reactive status and be provided with
medical counseling.

	Lastly, because of the transient nature of the infection and rapid
involvement of the immune response, such donors would be suitable to
donate when they test nonreactive.

	[Slide.]

	So, in March 2002, BPAC, the discussion largely focused on the apparent
lack of the medical benefits that might justify donor notification, so
consequently, FDA convened an ad hoc PHS panel in July of this year.

	The panel members include Drs. Harvey Klein and Kevin Brown of NIH,
Larry Anderson, Mary Chamberland, and Bruce Evatt from the CDC, and CBER
representatives.

	[Slide.]

	The conclusion by the PHS panels are regarding the donors, there is no
medical benefit in identifying high titer B19 NAT-positive donors
informing them of their reactive status and providing them medical
counseling.

	Regarding close contacts of the high titer B19 NAT-positive donors,
there are potential medical benefits to donors contact especially those
at risk, for example, persons with certain anemias, pregnant women, and
immune-suppressed or compromised individuals.

	Now, Dr. Brown will elaborate on these medical benefits later in his
talk.

	[Slide.]

	So, FDA is taking a stepwise approach in resolving B19 NAT issues
concerning whole blood and source plasma.  At this meeting, FDA is
seeking advice on the issues that are listed here.

	The need to reduce the risk to transfusion recipients by withholding
high titer positive units of whole blood and its components from use. 
The need to temporarily defer the high titer donors and whether
potential benefits to close contacts of B19 infected donors warrant
notifying high titer donors, and if so, what would be the time frame for
notification.

	Accordingly, we ask four questions.  Before you listen to Dr. Brown and
industry presentations, I would like you to bear in mind these questions
that we are going to ask.

	[Slide.]

	First, if donations of whole blood are tested for the presence of human
parvovirus B19, are risks to transfusion recipients sufficient to
warrant withholding high titer positive units.  These are equal or
greater than 106 geq/ml from use for transfusion.

	[Slide.]

	The second question is, is temporary deferral of positive donors
warranted in the setting of:  (a) whole blood donation?  (b) Apheresis
donation?

	[Slide.]

	The third question.  Do potential medical benefits to contacts of
parvovirus B19 infected donors warrant identification and notification
of positive donors?

	[Slide.]

	Finally, if yes to Question 3, should donor notification be limited to
settings where testing and notification can be completed within several
weeks of donation?

	Thank you.

	DR. NELSON:  We will come back to these questions.

	Dr. Brown.

B.  Overview of Parvovirus B19 Infection

Kevin Brown, M.D.

	DR. BROWN:  I am going to give an overview of parvovirus B19 infection,
mainly concentrating on the areas that I think you need to be able to
sort of answer the questions that Mei-ying posed to you.

	[Slide.]

	So, what is parvovirus B19?  It is a small, 22 to 24 nanometer diameter
icosahedral virus.  These are viruses by immune EM.  They are
non-enveloped, so solvent detergents don't work too well.

	They are relatively heat resistant because of the small genome, which
is only 5,500 nucleotides of single-stranded linear DNA.

	It has a high conserved genome and up until a few years ago, it was
said the variation was less than 2 percent of the DNA level.  There have
variants that have been described in the last two years, but I am not
going to say very much more--I am not going to say anything more about
it because they have not been isolated from blood except at extremely
rarely, when we found them in liver samples and other people have found
them in skin.

	[Slide.]

	So, the parvoviruses are divided into three genera.  They are the true
parvoviruses of which canine parvovirus or porcine parvovirus are
members.  There are the dependoviruses, also known as the
adenoassociated viruses, but B19 comes into the third genera, which are
erythroviruses, so-called because they are highly erythrotrophic and
they are only known to replicate in erythroid progenitor cells.

	So, the cells that the virus replicates in are these cells here.  The
precursors are the red cells, that is, the BFU-E and CFU-E.

	[Slide.]

	This is the virus itself.  The virus encodes for one non-structural
protein and two structural proteins.  The two structural proteins are
encode VP1 and VP2.  VP2 is the major structural protein.  It is 58
kilodaltons.  There is about 60 copies in the virus.

	Ninety-five percent of the variant, as I say, is VP2, and you can
express this in baculovirus and it self-assembles to form capsids. 
These VP2 capsids hemagglutinate and it was using this property that was
able to go on and show that the receptor for parvovirus B19 is
globosidal, known as Blood Group P antigen.

	The VP1, which is a minor component, has an additional 223 amino acids
at the 5-prime end of the VP2.  If this is expressed, it does not
self-assemble, but it is thought to be the main site of neutralizing
epitopes.

	This is again to show you what the virus looks like, but also to make
the point that it is even quite different, not at the DNA level, but
even at the structural level, looks quite different from the true
parvoviruses.

	[Slide.]

	As I said, I was able to show that parvovirus B19 uses globoside, that
is a glucose single lipid, demonstrated here as its receptor for viral
entry into cells.  I was also able to show that individuals that do not
have P antigen on their red cells or on their cell surface cannot be
infected by B19 either in vivo or in vitro.

	[Slide.]

	The discovery at the receptor for B19 does go on to explain a few of
the things that we didn't know about B19, a lot of the pathogenesis of
which I am not going to say more about it, but globoside is found on
erythroblasts, as you would expect.  It is found on megakaryocytes,
megakaryoblasts, vascular cells, on the cells in the placenta, does
cause transplacental infection, which we will come to, and it is found
in the thecal hematopoietic cells and myocardial cells.

	[Slide.]

	Turning to the epidemiology of the virus, the virus is a very common
infection.  Everywhere that people have looked, they have been able to
find it apart from some isolated communities in South America and
Africa.

	Virtually, all countries where people have looked, 50 percent of adults
have B19 antibody with seroconversion happening in childhood and also as
young adults.  The calculations are that there is an annual
seroconversion rate in women in the USA of about 1 percent.  That is 1.5
percent per year.

	[Slide.]

	It shows classical epidemic behavior, with temperate countries,
increased peaks in the spring.  It also shows variability between the
years, so some years there is a much higher peak of virus than others.

	Looking at natural infection, the incubation period has been calculated
to be between 4 and 14 days depending on the presentation of the
infection, with a maximum up to 20 days.

	[Slide.]

	The major route of the natural infection is by the respiratory route
and it is actually fairly infectious.  In studies that have been done in
susceptible individuals, there is a 50 percent attack rate in household
contacts and 25 percent attack rate in schools or nurseries.

	There also is a high level of viremia and blood products can have virus
in them, and you have already heard about how pooled products can be a
source of B19 infection, but there are cases of single component, which
I am going to come to in a few minutes.

	[Slide.]

	The kinetics of B19 infection were really well established by some
volunteer studies which were done in the UK in the 1980s.  These were
all adult volunteers who had different concentrations of virus dripped
into their nose. They were then followed to see what would actually
happen to them.

	There were a couple of individuals who had pre-existing antibody to
B19.  They remained well, there was no viremia, and as I say, they have
no symptoms.  There were a couple individuals who had low levels, which
were called equivocal IgG levels.  They had a fever and chills, headache
and myalgia, only they didn't develop the second stage infection.  They
also boosted their antibody response.

	However, in those that were antibody-negative, and received more than
108 genome copies into their nose, they had this typical biphasic
illness, so that there was this level which was associated with viremia,
with fever, chills, headache, myalgia.  At that time, there was a drop
in their reticulocyte count, and then a second phase infection which
resembled fifth disease as the antibody came up and the virus cleared.

	However, it is important to note from this that these assays were done
really before PCR was well established and the viremia was measured by a
dot blot, and the dot blot has a sensitivity of between 105 and 106
genome copies/ml.  So, although the viremia could only be measured for
about four or five days, the fact that it goes back to normal or back to
the zero line doesn't mean to say there is no virus actually present,
and that will become relevant in a few minutes.

	Also, there were no infectious assays performed. It is actually very
difficult to grow this virus, and there were no neutralization assays
performed, so we don't know whether these antibodies actually are
neutralizing or not.

	[Slide.]

	The virus can cause a wide range of different disease depending on the
host characteristics, and I am just going to go through and illustrate
each of these.

	[Slide.]

	The majority of infections caused by B19 are asymptomatic and it has
been estimated between 25 percent and 50 percent of infections are
asymptomatic even in an outbreak situation.

	The commonest presentation of B19 infection is erythema infectiousum
also known as fifth disease, slapped cheek disease because of this
characteristic slapped cheek appearance that children get, and also
academy rash, because of the outbreaks that are often seen in schools.

	The prodrome, which is at the time of the viremia, is usually missed or
there may be just mild symptoms, and the diagnosis is usually made at
the appearance of the rash. In children, it is particularly the slapped
cheek on the cheeks.  In adults, this stage is usually missed, but then
there is a second stage where you get this reticular pattern on the
limbs.  It comes and goes and pruritis is very common.  There is no
specific treatment.  It is usually a fairly innocuous infection and
often the parents are more worried about the rash than the children are
about the symptoms.

	[Slide.]

	However, in adults, and especially in women for reasons that we don't
understand, it is often associated with arthropathy or even a frank
arthritis.  This is a peripheral distribution especially in the small
joints.

	Again, it can persist for months.  It often lasts between two weeks to
two months, but can, in fact, last for six months or even up to several
years.  The problem is that it may resemble acute rheumatoid arthritis,
especially as the rheumatoid factor can be positive and also
autoantibodies are often present.

	So, diagnosis, this is B19 as opposed to acute rheumatoid, does have
implications especially for the management of these patients.  These
patients respond to nonsteroidals.

	[Slide.]

	The first disease that was associated with parvovirus B19 was transient
aplastic crisis, and this was seen in patients with increased red
turnover.  Basically, there is a transient arrest of erythroid
production and in those who have a high erythroid turnover or dependent
on that regular erythroid turnover, they develop this aplastic crisis,
which was originally described in sickle cell disease, but it can be
described, it has been seen in many cases of hemolytic anemia, in fact,
any cases where there is increased erythropoiesis including even in
acute hemorrhage.

	Often pronormoblasts, such as are illustrated here, can be seen in the
bone marrow, sometimes in the peripheral circulation.  It is a
self-limiting infection.  As soon as the virus clears, which is about
four to five days, the antibodies come up.

	The reticulocytopenia results.  There is a single episode in a
lifetime, and treatment is supportive to get them through the aplastic
crisis.

	[Slide.]

	However, in some individuals who can't mount an immune response or
can't neutralize the virus, they will go on to develop a persistent B19
infection.  This is being described in a variety of different cases of
congenital immunodeficiency.  I am going to illustrate a case of
Nezelof's in a minute.

	In cases with iatrogenic immunodeficiency, such as chemotherapy or
immunosuppressed patients, it is also quite commonly seen in patients
with AIDS.

	[Slide.]

	So, this is a chart with Nezelof's syndrome who had dot blot positive
instead of the usual three to four days, actually lasted for many
months.  This was associated with an anemia and a reticulocytopenia.

	This was despite the presence of low levels of antibody.  There was IgM
and IgG present, but the virus was there at high titers.

	[Slide.]

	These cases are often very well treated and respond well to commercial
immunoglobulin which has high titers of neutralizing antibody.  This is
a case of a patient with HIV infection who again had high levels of B19
DNA present for many months and was treated here with IVIG. There was a
brisk reticulocyte response peaking and then coming down to normal and a
response of the hemoglobin, and the virus actually disappears.

	It is interesting to note at the time that the antibody responds, many
of these patients have the symptoms of fifth disease.  They have the
arthropathy and they also have the rash.

	[Slide.]

	The next case I want to talk about is where you have a combined, where
you have an immunocompromised or immunosuppressed individual with high
erythropoiesis, by which I mean the fetus.

	In fetal hydrops, there are many, many case reports now in the
literature that are very similar.  B19 can be found in all the tissues. 
In some of these cases, this is associated with myocarditis, globoside
is actually found in the fetal heart.  Some of these spontaneously
resolve.

	[Slide.]

	This is a combined study again in the UK where they looked to see what
the risk of B19 was or the outcome was of confirmed B19 infection in
pregnant women.  So, this is a stage that the mothers have their B19
infection. You can see there is an increased fetal risk due to B19
infection, particularly in weeks 9 to 20, so the first half of
pregnancy.  There is an excess of fetal hydrops at this stage.

	[Slide.]

	The conclusion of this double study was that both asymptomatic and
symptomatic infection is associated with hydrops or fetal loss.  There
is a mean interval of about six weeks between the maternal infection and
the fetal symptoms, but 1 in 10 infections, confirmed B19 infections
before week 20 will result in fetal loss due to B19.

	The treatment for hydrops, if it is early, and many of these cases have
been treated with intrauterine blood transfusion with positive results.

	There have been no studies that have confirmed congenital abnormalities
due to B19.  There is a few case reports.  We have described three cases
of congenital anemia following a maternal B19 infection.  Different from
what happens with the persistent infection, these congenital anemias did
not respond to immunoglobulin treatment.

	[Slide.]

	Despite those well attested associations with B19, there are a number
of different disease associations that have been described in the
literature, often based on PCR results, and some of these are
controversial as to whether this is a chance association.

	Certainly, it is being proposed that B19 causes PCH.  It definitely
causes some cases of hemophagocytic syndrome, ITP, vasculitis, Kawasaki
is more questionable, some cases of hepatitis have been ascribed to B19.
 Recently, particularly in Japan, there is a group that is claiming that
B19 may cause rheumatoid arthritis.

	[Slide.]

	This is returning to the picture that I showed earlier based on the
data from the volunteers, which has held up to be fairly good, but as I
say, the first studies were done in the absence of PCR, but when PCR is
now actually added, instead of actually just being positive for three to
four days, the PCR actually remains positive for months, if not years.

	Again, these studies were done without quantitation, so the question is
how high is that virus, is it infectious, and what is it doing.  I just
want to make the point that the viremia that people talk about in the
old literature is associated with this dot blot sensitivity of about 106
genome copies/ml.

	[Slide.]

	The fact that that virus persists at low level detectable by PCR, it is
not really surprising then that when people have gone back and looked in
blood donors, you can actually find virus.

	So, this is not supposed to be exhaustive, but just to show some of the
studies where people have looked at healthy blood donors by a variety of
different methods, that do have different sensitivities, so it is not
surprising that you get different prevalence rates with the highest
sensitivity those based on PCR have given you the higher prevalence of
the virus.

	I am going to go and say more about this study by Yoto, but I want to
make the point that these samples, 1,000 samples were taken at the time
of an outbreak in the community, which is why they are probably higher
than some of the other numbers that people have.  I could allude to the
study by Jordan.

	[Slide.]

	The fact that the virus is present, can go up to 1012, 1014 genome
copies/ml, it can be a problem in pooled products, and all these
products, there have been cases in the literature where B19 has been
said to be transmitted from them.

	But single components, it is much rarer.  There are some case reports,
but there are not an awful lot of them.

	[Slide.]

	I want to go through some of them, and again this is not exhaustive,
but it is supposed to be I think for the solitary.  This was actually a
study by Yoto, which is the group that did the prevalence of B19 by PCR
at the time of an outbreak.

	That was actually triggered by this case that they found.  This was
14-year-old boy who had ALL, who was being well maintained on
chemotherapy.  He received two units of red cells.  They have no
additional information on the red cells, so they went back and looked at
1,000 blood units to see what the prevalence of B19 was, but they didn't
test these two units.

	The child actually developed a profound anemia.  He actually went on
and had a peripheral blood stem cell transplant, was treated with IVIG
just as CMV prophylaxis, not actually because they had recognized that
he had the B19 infection at that stage, but he had a viremia that lasted
for a month.

	Certainly, I can't imagine that they would have done a stem cell
transplant if they had known that he had an acute B19 infection going on
at the time, so this was an infection that was missed.

	[Slide.]

	This is a second case.  This is a 22-year-old woman who had thalassemia
major diagnosed at the age of 15 months, so had been receiving two units
of red cells monthly, so estimated actually received probably about 500
units of blood at this point.

	She again developed a transient aplastic crisis and severe heart
failure.  It turned out she had received two units of red cells nine
days previously, and one of the donors that they were able to show was
IgM positive, DNA positive.  Unfortunately, there is no information
given as to what they mean by "positive."

	This individual had an invasive esophageal echocardiography, was
misdiagnosed as having subacute bacterial endocarditis, treated for a
month with antibiotics before they actually realized what the diagnosis
was.

	So, again, a patient that was mistreated because the diagnosis wasn't
thought about.

	[Slide.]

	This is a third case which was again picked up retrospectively.  This
was one of the cases that was found by Jordan in that study of looking
at the prevalence of B19 in their blood donors.

	What they did was they identified positive blood donors and then went
back to see what happened to the recipients of that blood.  One of the
individuals had a severe anemia, and this was the individual.

	This was 49-year-old man who had a liver transplant for chronic HCV. 
He received one unit of red cells two days post transplant.  The donor
was IgM negative, IgG negative, but DNA positive.  Again, unfortunately,
they don't give a titer, but it was strongly positive on their
ELISA-based assay.

	Four months later, he was thought to have recurrent hepatitis, he was
found to be anemic, found to have reticulocytopenia, was treated with
red cells.  B19 wasn't even considered.

	Eight months later when Jordan contacts them to find out what happened
to the recipient of this blood, they find out that he had a B19
infection at this stage.  He has, in fact, seroconverted.  He is IgM
positive, IgG positive, B19 and DNA negative.

	[Slide.]

	These were the 10 recipients that they were able to follow up on.  It
is interesting to note that this was the only patient that received B19
DNA positive.  Now, as I say, we don't know what they mean by that
titer, because they don't actually give a titer.  Had B19-like symptoms,
so was followed up.

	Interestingly, this patient was IgM, the donor was IgM negative and IgG
negative, so presumably this was an acute B19 infection that the donor
had, but was asymptomatic at the time.

	[Slide.]

	The last case I want to talk about is a case where there was
transmission by platelets.  This was a 36-year-old who had CML treated
with chemotherapy, received a bone marrow transplant, had been tested
before and was shown to be B19 IgG prior to the transplant.

	Over the next six months, he had increasing anemia and eight months
later, B19 was diagnosed.  The patient was treated with IVIG.  They were
able to go back, and he had had blood from 90 donors, and they were able
to actually test all of those.

	Two of them had B19 DNA.  One of them, Donor A, it was greater than 106
genome copies/ml.  I can say that because it was dot blot positive, and
I used to work in this lab and I know the sensitivity of the test that
they use.

	They also sequenced the donor's virus, and he had the same B19 sequence
as they found in the patient.  There was a second donor that was
positive.  This had been given prior to the transplant.  This had less
than 106 genome copies/ml.  It was only positive by PCR, not by dot
blot.

	But again this diagnosis was only made eight months after the event.

	[Slide.]

	So, what about levels of B19 DNA?  As I say, there are not that many
studies where people have published where they have looked at actual
levels, but this is a group from Italy where they have taken blood.

	I think it sort of makes the point they divided them into those that
were DNA positive, IgM positive, and IgG negative at the early stage,
and you can see they have relatively high viremia levels, those who were
IgM positive and IgG positive, so within the first couple of months,
following an acute B19 infection, lowered, but there are still some that
are above this cutoff of 106, and then some that were IgG negative and
IgG positive, which suggests they had a B19 infection more than two
months earlier, and yet one of these individuals had greater than a 106
genome copies/ml.

	So, the question that I think that we really don't know the answer is
although we know that B19 can persist for months even at relatively high
titers, if this level of B19 is infectious if it is given as a unit of
blood.

	[Slide.]

	The problem is that we really don't have the answers to some of these
questions.  The virus is very difficult to grow in the culture.  It can
only grow in human bone marrow explants or thecal livers as a source of
hematopoietic cells.

	There are a few cell lines that have been described, but they really
aren't that highly permissive. You can detect virus either by looking
for inhibition of colonies, but it is very insensitive looking by
immunofluorescence for capsid proteins or NS proteins.

	Some people have described in real-time PCR looking for DNA increase. 
We prefer to use RT-PCR looking for viral transcripts to distinguish
replication RNA from viral DNA.

	[Slide.]

	I put this slide on really just to illustrate that even if you have an
infection assay, this is using different concentrations of virus at
three different cell lines that are said to be explicit for B19.

	In the most highly sensitive, you can detect down to 104, 103 genome
copies per ml,  but in another cell line, you can't really detect
anything even at 106, so even the cell lines themselves have problems in
their sensitivity as to what you call infectious, and there is obviously
limits to how much virus you can actually put in there.

	[Slide.]

	So, coming back to the FDA questions, who is at risk of parvovirus B19?
 Well, you could argue anybody who is seronegative is at risk of B19
infection, but fifth disease is a relatively innocuous disease and apart
from the problems with arthritis or arthropathy, the main problems in
these patient that have increased erythropoiesis, patients who are
immunosuppressed or immunocompromised, and the pregnant woman and the
fetus.

	By definition, these aren't going to be your blood donors, but they may
well be your contacts of your blood donors.

	[Slide.]

	Is there any way that we can prevent B19?  Well, there is a vaccine
that is in Phase 1 trials, but it is going to be several years at the
least before we actually have a vaccine.

	What about passive immunotherapy?  The only thing that we have is IVIG,
we don't have any antivirals that we know work for B19.  Again, you have
got the problem with the time interval.  If you are trying to actually
stop the viremia, you have got four days to actually do anything about
it.  So, time is limited.

	[Slide.]

	This is my last slide.  This again shows the time course of B19.  What
I tried to do here was to put when you expect to see the symptoms at the
different risk presentations.

	So, the transient aplastic crisis is really at day sort of 4 to 7
following your infectious exposure.  You have not got a lot of time to
do anything there.  The persistent anemia starts probably just about the
same time, but lasts actually months, so even with the late
notification, you might actually be able to go in and make a difference.

	The fifth disease is usually at about sort of 21 days, but the
arthropathy can last for several months.

	So, what are the benefits of notification?  Well, you might be able to
intervene.  I did mention pregnancy, but certainly in the pregnant
woman, you might be able to monitor and see what is happening.

	I think it is important for accurate diagnosis and may be important for
treatment especially chronic infection and also for the monitoring of
pregnant women who might be at risk of hydrops.

	Thank you.

	DR. NELSON:  Thank you, Dr. Brown.

	Questions?  Mary.

	DR. CHAMBERLAND:  Thank you for that nice presentation.  I had a
question.  It seemed, and I am sure this will come up in the discussion
again, that the benefits of notification are going to be not so much
with respect to the donor, but the close contact.

	That is clearly going to probably be linked to the period of time over
which notification can take place after the unit has been tested.

	In your review article, which was distributed to the committee, under
Transmission, it is stated that the case-to-case interval is 6 to 11
days irrespective of the type of B19 related disease.

	Now, if I understand that correctly, it would suggest that let's say in
a typical household setting, a contact setting, that there is really a
fairly limited period of time in which you would have to, if you will,
inform a donor and potentially interdict subsequent transmission.

	Sort of tying that in with the other comment in your slide, you say the
slide that looked at the titers of B19 DNA relative to IgM and IgG, you
ask the question is this level of B19 infectious if given as a unit of
blood.

	Actually, I think the sort of complementary question is, is this level
of B19 infectious through the traditional respiratory route.

	I am just wondering if you could comment on that because I was quite
struck by this statement in your article about this fairly tight time
period.

	DR. BROWN:  It seems to have held us in sort of fairly good stead, this
sort of illustration, because this time of the viremia does seem to
correlate well to the time of infectiousness in close contacts.  So,
this seems to be, as I say, a very good estimate.

	So, as soon as the IgM and especially the IgG starts to come up, then,
you don't actually see exposures continuing to happen.

	Maybe that sentence is a bit ambiguous.  There is a slight difference
in the time between exposure and the disease itself, because the
transient aplastic crisis will appear earlier, which is the first phase.
 If you are looking for fifth disease, it's the second stage, and you
see the viremic stage is missed, so that usually sort of comes up a
little bit later.  So, that is why you have quite a wide range, but the
infectious time is actually very tight exactly.

	DR. SIMON:  If I put that in practical terms, so if we detected
someone, and I gather that fairly insensitive techniques are being used
because we are looking for high titer, so we would be likely to detect
someone at about day 7 time frame.  So, the infectious period, you are
going back to day zero when you say 6 to 11 days?

	DR. BROWN:  Yes, these numbers are based on having high titer virus
dripped into your nose, which is not really an every-day occurrence and
probably doesn't mimic what is happening either if you get a unit of
infected blood or if you are living with somebody who is actually
copping the virus out.

	DR. SIMON:  So, unless we could get to that donor within four or five
days, it probably would not have much utility.

	DR. BROWN:  Exactly.  That is what I was sort of trying to get to here.
 If you actually want to actually go in and actually intervene before
this viremic stage, you have got a very, vary narrow time cap, and I am
not sure, quite frankly, that it is realistic.

	However, if you are trying to actually go in and do something about
patients who have persistent anemia, now, you have got a much longer
time because they are being viremic, but they will continue to be having
symptoms, and they would actually do well with actually being treated
even if it quite a bit later on in their disease.

	DR. SIMON:  Treated with what?

	DR. BROWN:  The only treatment that we have is IVIG, and it actually
works very well, as in the HIV case.

	DR. SIMON:  I am told by our expert that it is considered experimental
therapy.  I know it's off label and that not everyone accepts that it is
beneficial particularly if you don't know the titer of parvovirus
antibody in the preparation.

	DR. BROWN:  It is the only treatment that is available, there is
nothing else.  So, I think when people have actually looked for
neutralizing antibodies in most commercial sources of IVIG, they are
actually at high titers.  I don't know if anybody else wants to comment
on that.

	DR. WU:  I think we ought to let other industry presenters to present
and maybe they have some data, and if not, we do have some data that we
can present later on to shed that light.

	DR. DiMICHELE:  I was really surprised to see the number of people who
get infected in middle age and beyond. Is there any difference in the
morbidity from this infection in the middle age and older age population
compared to the younger population?

	DR. BROWN:  None that I am aware of other than the fact that women, and
it doesn't seem to be particularly old or young, but certainly from 20
up would seem to be more likely to have the arthropathy and the
arthritis, but otherwise, there doesn't appear to be any difference if
you get your B19 when you are 80 as opposed to when you are 20.

	If you get it when you are 8, you will probably have less symptoms and
you will have the slapped cheek.  That is all I can say.

	DR. NELSON:  Thank you, Dr. Brown.

	Dr. Stramer.

C.  Industry Data Presentations

American Red Cross

Susan Stramer, Ph.D.

	DR. STRAMER:  Thank you very much.

	This is a compilation of several presentations. Firstly, most of the
data that I will present were presented at the AABB last year, then
presented as Dr. Yu referenced, at the December 2001 FDA NAT Workshop,
and I have added some new data and comments at the request of FDA.

	I also want to mention that I am not only discussing B19, but I will be
making some remarks about hepatitis A virus, because we really can't
separate the two as we are trying to screen our plasma derivatives for
nonenveloped viruses.  Parvo is one, but so is HAV.

	So, my presentation is contaminated with some HAV material, and I will
try to limit those remarks.

	[Slide.]

	As we have heard, manufacturers of plasma derivatives have implemented
NAT for nonenveloped viruses and such testing will likely be implemented
for recovered plasma.

	Most parvovirus B19 NAT programs target the elimination of equal to or
greater than 1 million copies/ml as already referenced by Dr. Yu.

	Studies of HAV and B19 frequencies in recovered plasma are limited. 
Dodd and coworkers at the 1997 AABB reported some results from screening
pools of 512 at NGI, and those results, we had zero positives for HAV
out of 20,000 donations screened, but a frequency of 1 in 1,400 for B19.
 This was using a very sensitive test and as Dr. Brown referenced, the
prevalence is very dependent on the sensitivity of the test that is
used.

	[Slide.]

	Dr. Yu also mentioned solvent detergent treated plasma and through the
three-year experience that has been reported at Vitex for NAT screening
of S/D plasma and final product, which is 2,500 donations, and that is
at NGI for HAV, or by their in-house testing for B19.

	Now, they may screen pools of 100 that are comprised of five subpools
of 20, and if a pool of 100 is found positive, they resolve to the
subpool of 20, and if that subpool of 20 was found positive, they
discarded all 20 individual units.

	But from their testing, their frequency for HAV was about 1.5 million,
and about 1 in 800 for B19 DNA.  When we started to think about
developing programs for B19, we tried to recover some of these positive
units prior to discard, such that we could research the frequency of
B19, the titers of B19, and try to get in-house some positive materials.

	So, what we were provided was greater than 1,000 units, and we have
tested these representing 20 positive subpools of 20, but to our
surprise, of those subpools that we tested, the individual units
comprising those subpools, only 23 of greater than 1,000 units tested
were B19 positive at NGI using a standard test.  Those were from 16
subpools.

	So, of the 72 subpools tested, only 16 were positive.  That indicated
that we had a 77 percent false positive minipool test result using the
sensitivity of the Vitex procedure.

	[Slide.]

	When I presented these data at AABB and at the FDA NAT Workshop, we
only had a couple of positives identified to that point, and this again
supports the data presented by the earlier speakers.

	Here, you can see, even though these are separate units, if you align
these by titer, and then you look at the IgM and the IgG concentrations
or presence of those two antibodies, you can see that only in this very
high titer positive unit there is no IgM or an IgG.  Those that had
lower titers had the presence of IgM.  As titers decreased, IgM
disappeared and then all were positive for IgG.

	[Slide.]

	For the 23 that we have identified at the conclusion of the study, we
haven't completed the IgM and IgG testing, but I just present to you the
titers of the positives of these 23 that we found.

	You can see only five here of three years of Vitex screening, only five
represented very high titer units.  The rest were lower titers and they
do what you would expect here, maybe had a mix of IgM, IgG, and perhaps
some of these represented further contamination as 5 of 16 of these
pools that contained these positives, contained multiple low level
positives suggestive of contamination.

	[Slide.]

	So, from the Vitex studies, we know that HAV was infrequent, B19 NAT
false positivity may be an occurrence that we have to deal with,
especially when you are dealing with very high titer units.

	Low level B19 DNA positive, IgG positive samples do occur, and
individuals with early acute B19 infection have high viral titers and
are likely to be IgG negative.

	[Slide.]

	So, in order to prepare for some type of screening program, which I
should have said at the onset we are not yet doing anything for parvo. 
We have done a couple of pilot studies, and I am going to present
results from two of those.

	We did the first study with NGI, and it was an unlinked study to
determine the frequency of HAV and B19 in recovered plasma.  The testing
was done from our surplus NAT negative tubes for HIV and HCV our NAT
negative PPTs that were sent to NGI.  Once at NGI, they were pooled and
tested.

	The testing used at NGI, and NGI won't be here to present, they use a
four-test reaction for both HAV and B19. That is two different primer
pairs that they run each in duplicate.  So, each of the four tests, all
four tests have to be negative for them to report out a negative result.
 If any of the primers or replicates to primers is reactive, it is
considered a reactive.

	So, we took a half million donations that we sent to NGI.  They were
pooled into 100 pools of 512, which is the standard matrix that NGI uses
for pooling.  We tested HAV without dilution and if there was a positive
result, we would resolve to the individual donation and quantify.

	For B19, NGI's standard algorithm is to take the pools of 512, perform
a 1 to 1,000 dilution, and then test. So if we had a positive pool, we
then would resolve to the individual donation, quantify, and look for
antibody.

	One point that we added to the study is if a pool was negative at 1 to
1,000 dilution, we wanted to see what would be in those pools that they
were tested undiluted.  So, 1 to 1,000 dilution negative was further
tested NEET.

	NEET, that is the 512 pool was tested without dilution,  and if that
was positive, the same thing.  Resolve to individual donation,
quantified, and IgM and IgG testing performed.

	I do want to comment because this is in your materials from NGI, that
the sensitivity of the testing that we used at NGI was 20 copies/ml,
actually 22.4 to be exact, and that is what is in NGI's presentation.

	If you then multiply that by a pool of 512, multiply that by 1 to 1,000
dilution, you get a sensitivity for the donation of 1.2 times 107, so
this is where we get the greater than or approximately equal to 106, but
this is the sensitivity of the NGI test.

	[Slide.]

	So, the results.  For HAV, these were easy.  They were all negative. 
For B19 in performance of the 1 to 1,000 dilution, we had 3 positive
pools including 4 positive donations.  Two of the positive donations
occurred in the same pool.  So, the 4 positives gives you a frequency in
the study of 1 in 12,800.

	These are the 4 positives, the IgM present, results of IgM testing. 
All were IgG negative as you would expect from relatively high titers
especially in these two. No quantifiable IgG.

	One of the high titer ones was IgM positive, and one of the lower
titers was also IgM positive.  Interestingly enough, these two were
detected even though they were below the limit of sensitivity of the
tests we were using, perhaps being in the same pool, there was some
additive effect.

	[Slide.]

	Now, when we took the remaining 97 pools that were negative at 1 to
1,000 dilution and ran them without the 1 to 1,000 dilution, we had an
additional 34 positive pools including 95 positive donations, which gave
us really an unacceptable yield of 1 in 528.

	Again, if you look at the titer of the samples found, IgM, IgG
presence, and then this is the number of samples within these various
titers, you see there are some high titer samples, actually 1, that
probably should have been detected at the 1 to 1,000 dilution but
wasn't, and that one was IgM positive, and plus/minus for IgG.

	The lower titers had IgM as you expect, but also had IgG.  The lower
titer samples had a mix of IgM and again most of them have IgG.

	[Slide.]

	We did a similar study with GenProbe because since we are talking about
a Phase 1 and Phase 2 approach, the NGI would represent the Phase 1
approach where samples would be sent out for testing, but we recognize
in the future that in-house testing may occur, so we wanted to look at
the GenProbe test.

	So, what we did is we took NAT negative pools, pools of 16, about 2,500
of them, and representing April collections, and I should have said in
the NGI study that I showed previously, those also represented
springtime collections, and that is important because it is the highest
time of reported parvovirus prevalence.

	This represented about 40,000 donations.  They were tested at GenProbe
using a combination test for B19 and HAV, that had about 600 copies/ml
test sensitivity.  This was about 100 percent detection level, which in
our pools of 16, was about 9,600.

	Of those 2,547 pools tested, 24, just under 1 percent of these pools of
16 were reactive for B19.  There were no reactives for HAV once again. 
If we assume that there was one B19 positive donation per reactive pool
of 16, that would give us a prevalence of the sensitivity that we were
using this test at 1 in 1,700 versus what I showed you for the NGI
study, which used a less sensitive procedure, of 1 in 12,800.  So, in
this case, it was 7-fold higher.

	If you look at product loss, because again at a pool of 16, if we have
a positive, we have to discard all products, due to discard of all
members of a reactive pool when dealing with 16, would be about 1 in
100, which is unacceptable.

	Based on the distribution of quantitative results for those 24 positive
pools, the addition of the ones with 1,000 pre-dilution, which is what
NGI does, would result in a prevalence that was comparable to NGI, of 1
in 13,000.

	[Slide.]

	If you look at the titers of the 24 positives that we got in the study,
the vast majority would be expected to be below the limit of detection
of the GenProbe test in the pool setting, so those are unexplainable
findings or due to contamination, which is likely the outcome.

	Here are some moderate level positives that were detected.  We don't
have antibody signal results on these. Again, we have 3, these are the 3
high titer positive samples.

	[Slide.]

	Currently, we are also doing--this is moving on to something else--we
are doing manufacturing pool testing.  That is where many recovered
plasma units are pooled, approximately 3,250 liters, and these
manufacturing pools are tested for virus prior to final manufacturing.

	I don't want to dwell on this because this is HAV, but again it shows
you the type of result output that you get from NGI.  This is audio
radiogram, and I said that NGI performs two primer pairs, they do them
in duplicate, so here you see a set of four results.

	Lanes, all one are their positive controls, 17 through 19 are all
positives, so this one they didn't have a positive.  I mean there are
certain criteria, and these all met the validity criteria.

	But here you have an unknown and you can see here that it may not be
positive on all of the four rafts, but in this case, it certainly was
positive on three, constituting a positive result.  So, that shows you
the type of data output one would get.

	[Slide.]

	So, the conclusions from the presented studies is that blood collectors
considering implementation of B19 screening will have to evaluate NAT
methods that are relatively insensitive to prevent issues from
contamination and detection of low level NAT positives.

	The frequencies that I showed were 1 in 12,800 using the insensitive
method at NGI.  If you consider only the two very high titer positives
we had, the frequency is 1 in 25,600, and those were IgG negative.

	The frequency, if we increase the sensitivity, was 1 in 528 with
moderate titer samples that were plus/minus for IgG, but positive for
IgM.  Now, if you compare the GenProbe methods, we could get comparable
results depending on whether we do a dilution, which would yield a 1 to
13,500 result, or 1 in 1,700 frequency if we used their sensitive
method.

	[Slide.]

	High-titer screening methods may not capture all infections B19
positive units, however, the infectivity of antibody reactive, low-titer
positives is unknown, as has already been referenced.

	This study defines expected yields of B19 if sensitive and insensitive
NAT methods are used.

	This study also demonstrates the infrequent occurrence of HAV in
recovered plasma, which is about 1.5 million to a million.

	[Slide.]

	So, where does that leave us?  It leaves us with a discussion of Phase
1 and Phase 2, and I will try to answer some of the FDA's questions
here.

	One mechanism to do Phase 1 testing that the Red Cross will likely
implement is the method with NGI where we would outsource the testing to
NGI, and the process time for the testing would exceed the dating of
labile components.

	So, by the time we got test results, the only thing we would have are
frozen components.

	[Slide.]

	Now, how do we limit this, so we are only dealing with recovered
plasma, and not the issues surrounding FFP?

	Following the completion of our current HIV and HCV NAT testing, we
would take our NAT tubes and identify those that correspond to recovered
plasma.  Those recovered plasma tubes would be pooled into pools of 16. 
They would be sent to NGI for further pooling, into pools of 512.

	NGI would test for HAV and parvovirus, following a 1 to 1,000 dilution
for the parvovirus.  If negative, the product with the plasma would be
fractionated.  If positive, we would resolve to the pool of 16, and all
in-date frozen products would be discarded, and the good news here is
that we wouldn't have any FFP because they were never tested to begin
with.

	So, the question is then how do we address Phase 2, which would be
testing in-house using a commercial kit. This would represent real-time
testing in pools.  Most likely in our scenario at the Red Cross, we
would maintain our current pools of 16, and as I mentioned with the
GenProbe procedure, we could do a pre-dilution step as part of the
assay, whether that is a 1 to 1,000 dilution to reach about a 107
copy/ml per donation level or 1 to 100 dilution, so we get to 106.  That
remains to be determined.

	Reactives would be resolved to the individual donation within real
time.  So, what does that mean?  For product release, in reactive pool
resolution, for the latter involving usually 3 rounds of testing, we
have anywhere from about a 10 to a 48 hour per donation turnaround time,
and that is really based on our current NAT testing now.

	If pools are negative, our turnaround time is about 10 hours.  If a
pool is reactive, requiring resolution testing, and then if it is a
multiplexed test, discriminatory testing, final results may not be
available for 24 to 48 hours.  So, this really represents the range of
when results are available.  In this model, no product release would
occur unless the units not only test HIV/HCV negative, and in the future
West Nile, but also HAV and B19 negative.

	[Slide.]

	The B19 sensitivity level would initially be set for the removal of
high-titer units that is greater than 106 copies/ml.  We really couldn't
make claims for labile products because we are not removing all parvo,
we are just removing high-titer units.

	Really, again, as reference, we would really need to determine the
needs for recipients of labile products, what the level of sensitivity
should be, who should receive these products, et cetera.

	Donor notification, management of products from NAT-reactive donors'
previous donations and recipient tracing, which we hope won't have to
occur, would have to be determined.  But regarding donor notification,
as has been addressed in the questions from Dr. Brown's presentation, is
our time required for donor notification, varies by the marker, but it
is generally two to three weeks and for some markers where we outsource
supplemental testing, it may take the full 56 days or up to 8 weeks.

	Our timeline for any type of Phase 2 implementation is dependent on the
regulatory policies that FDA mandates, availability of test kits, and I
didn't even list here all of the implementation issues, such as those
outlined for West Nile virus.

	Thank you.

	DR. NELSON:  Thank you, Dr. Stramer.

	DR. SCHMIDT:  What is NGI?

	DR. STRAMER:  Oh, I am sorry, National Genetics Institute.  They are a
clinical reference lab.

	DR. SCHMIDT:  Thank you.

	DR. STRAMER:  I was hoping to get a more challenging question than
that.

	DR. SCHMIDT:  I thought maybe it was a new disease.

	[Laughter.]

	DR. STRAMER:  It could be.

	DR. NELSON:  Actually, ironically, they were supposed to present next,
but I understand Dr. Conrad isn't here.

	DR. STRAMER:  I tried to address two points in his talk because he
presented differing sensitivity, but it is actually the same test at 20
copies/ml, pools of 512, and a 1 to 1,000 dilution.  According to my
$1.99 calculator, that comes out to 1.2 times 107, and I did it twice to
verify my initial results, so it is about 107 sensitivity for their
method.

	DR. FALLAT:  You have presented a lot of data and threw out an awful
lot of numbers.  Can you simplify it for me?  What do you think is the
best estimate from your large sample size of the general incidence of
this virus in donor pools?  I see numbers from 1 to 500 to 1 in 25,000,
and what is the best number?

	DR. STRAMER:  Using sensitivity that I think is reasonable to eliminate
the high titer units, I think we will wind up with a prevalence between
1 in 10,000 and 1 in 15,000.

	Now, if a cutoff of 106 or 107 is adequate, that is what the prevalence
would be.  If we dropped the cutoff, then, we are dealing with
considerably different numbers.

	DR. NELSON:  Next, from the plasma industry, Barbee Whitaker.

Fractionators/PPTA

Barbee Whitaker, Ph.D.

	DR. WHITAKER:  Good afternoon.  Thank you for the opportunity to
present the PPTA approach to reducing parvovirus B19 load in
fractionation pools.

	I would like to mention that there have been a few changes to the
slides that were distributed to the committee last week, and that you
should have the current version, the version I am presenting now in
front of you.

	I would also like to mention that we have three presentations as a part
of our industry presentation and I would like to respectfully request
that you hold questions until the end because it's a comprehensive
presentation.  Thank you.

	[Slide.]

	PPTA is the international trade association and standard-setting
organization for the world's major producers of plasma derived in
recombinant analogue therapies.  Our members provide 60 percent of the
world's needs for source plasma and protein therapies.  These include
clotting therapies, immune globulins, and alpha-1 antitrypsin among
other products.

	PPTA members are committed to assuring the safety and availability of
these medically needed life-sustaining therapies.

	[Slide.]

	Although transmission of parvovirus B19 is uncommon through plasma
therapies, PPTA recognized the particular vulnerability of specific
therapy recipients including pregnant women and immunocompromised
individuals. The industry opted to pursue a strategy of identification
and removal of high-titer units as described already.

	About the same time that we were looking at this, based on experiences
observed with the solvent detergent treated plasma for transfusion, FDA
encouraged the fractionation industry to limit viral loads in
manufacturing pools.

	About a year later, in 2000, the European Medicinal Evaluation Agency,
the EMEA, held a workshop to address viral safety of nonenveloped
viruses.  It was concluded that given the current extent of knowledge,
further introduction of regulatory requirements should be carefully
considered, and to date, there have been no further meetings on this
subject.

	Also, in 2000, the WHO released an International Laboratory standard
for parvovirus B19, allowing the standardization of various laboratory
tests particularly NAT.

	PPTA released its voluntary industry standard for the management of
parvovirus B19 in mid-2001.  This standard is one of five critical
standards in PPTA's Quality Standards of Leadership Excellence and
Assurance Program.

	The goal of our Standards Program is to enhance the margin of safety
and quality of each and every product that reaches our consumers.

	[Slide.]

	PPTA's parvovirus B19 standard requires in-process control testing of
incoming source plasma by NAT for parvovirus B19 DNA.  Plasma that would
result in a manufacturing pool exceeding 105 International Units/ml is
removed.

	Effective July 1st, 2002, manufacturing pools may not exceed 105 IU
parvovirus B19 DNA/ml.

	PPTA's standard is designed to enhance the safety of the finished
product and is based upon the recommendations of the September 1999
Blood Products Advisory Committee specifically and described by Mei-ying
a little bit earlier.

	The recommendation to treat parvovirus B19 is an in-process control
that no studies were required to validate clinical efficacy of B19 NAT
under IND for plasma for further manufacture.  The validation should
proceed as an analytical test only and that no clinical correlates were
necessary if no decisions regarding donor or recipient management were
taken.

	[Slide.]

	As Dr. Yu has described, FDA has requested additional data regarding
specific industry practices, and I would like to introduce Dr. Steve
Petteway of Bayer Biological Products, who will walk you through the
requested data.

	[Slide.]

	As agreed with FDA, industry data will be presented in an anonymized
fashion.  Companies represented are as follows:  Alpha Therapeutic
Corporation, Aventis Behring, Baxter BioScience, and Bayer Biological
Products.

	Following Dr. Petteway's presentation, Dr. Edward Gomperts of Baxter
BioScience will present the potential impact of donor notification.

	Thank you.

Stephen R. Petteway, Jr.

	 DR. PETTEWAY:  Thanks.

	The FDA has requested that we provide an update for our in-process
control testing of source plasma for parvovirus B19, but before I do
that, I want to address the FDA issues that are specific here from
testing algorithm through to profiles for B19 serial bleeds.  I will
address those.  I think that you have them in front of you.

	[Slide.]

	However, before I do that, I want to remind everyone that pathogen
safety is a comprehensive approach with effective redundant measures
that provide a high margin of safety.

	Beginning with the donor, with donor screening, management of
donations, and management through testing donations and inventory hold
and lookback, followed by manufacturing and specifically the management
of plasma or manufacturing pools, coupled with virus inactivation and
removal, and this coupling is very important in the whole safety
profile.

	Then, moving through the process ending with postmarketing surveillance
in support of our patients.

	[Slide.]

	Specifically for parvovirus B19 management, we focus on two of these
manufacturing safeguards, the plasma or manufacturing pool, and
in-process control of the plasma manufacturing pool again coupled with
virus inactivation and virus removal through the purification and
manufacturing processes.

	I think to understand the value of this in-process test method that we
have implemented, understanding the link between these two is critical.

	[Slide.]

	Prior to implementation of our testing paradigm, no plasm units were
tested for B19.  This resulted in manufacturing or production pools that
ranged from 101 to 109 International Units/ml, followed by process viral
reduction gave us a defined margin of safety, however, after
implementation, high-titer units are identified through minipool testing
and removed, now providing production or manufacturing pools with a
titer of 105 International Units/ml when coupled with the same process
viral reduction lead to an increased margin of safety.

	That is really the target of this testing is increasing the margin of
safety.

	[Slide.]

	To address the first issue that we were asked to address, the NAT
sensitivities for minipool testing and original units, what we are
presenting is targeted testing threshold for minipool testing as opposed
to analytical sensitivity.

	[Slide.]

	The reason for that is that the sensitivities required to achieve
reduced manufacturing pool loads are a function of minipool size and the
manufacturing pool size. They are a volume factor.  So, the differences
in the size of the minipools i.e., the volume or the manufacturing pool
greatly influence what this targeted cutoff is.

	Minipool and manufacturing pool sizes vary across the industry. 
Therefore, each manufacturer has set the testing threshold based on the
size of minipools and manufacturing pools to achieve the PPTA standard. 
That is why you see different threshold levels from company to company.

	[Slide.]

	The targeted threshold levels for original units we have calculated for
you here because we don't actually test the original units.  We
back-extrapolated from the minipools.  We remind you again that each
manufacturer has set the testing threshold based on the size of the
minipool and the manufacturing pool, and the goal is to achieve the PPTA
voluntary standard.

	You can see the differences, 5 times 105 to up to 107, and it's a
volume related issue.

	[Slide.]

	So, minipools that are reactive and based on the targeted threshold are
assessed and units are released or discarded based on individual company
processes for carrying out that activity.

	[Slide.]

	Looking at the prevalence and levels of B19 DNA in minipools, as you
can see, they range from 1 in 3 minipools down to 1 in 40 minipools.  Of
course, this is because the frequency in minipools is influenced by the
size of the minipool and it varies across the industry.

	[Slide.]

	The B19 DNA levels can range up to 1011 International Units/ml
depending, of course, on the titer and the donation.

	[Slide.]

	The next issue is the prevalence of reactive minipools, original units,
manufacturing pools, and the levels of B19 DNA in each, and we will
attempt to provide that for you.

	[Slide.]

	The frequency and levels of B19 DNA in original donations.  What we are
showing here is the frequency of discarded units, that is, in the
testing paradigm, because of the logistics of testing, all the units
that we discard are not necessarily greater than the threshold, so what
you are seeing is the units that we actually discard.

	It ranges from 1 in 2,000 to 1 in 5,000, and you can see that it
correlates with the threshold, and some companies have a lower threshold
than other companies have for identifying and dealing with units.

	[Slide.]

	For B19 DNA in manufacturing pools, I think this is a very important
slide and pretty graphically illustrates the whole point of this testing
and what we gain from it and the value of it.

	Prior to B19 in-process testing, this represents,  each line, each data
point represents the titer in a manufacturing pool and you can see that
many manufacturing pools have titers as high as 108 to 109 International
Units.

	Following the implementation of the B19 in-process testing, however,
there is a consistent reduction of the titer of B19 in the manufacturing
pools across the same time frame.  This data clearly demonstrates the
value of the in-process control testing for manufacturing pools of B19
NAT, and this again is our whole goal, is to reduce the load in the
manufacturing pools.

	[Slide.]

	Resolution Times.  Now, you can interpret resolution times in many
ways.  Resolution time may be from the time a sample is received in a
laboratory to the time the result is available.  So, we interpreted
resolution time as from the time collection occurs, the collection of he
unit, until a result is available.

	You need to understand that this doesn't include time for confirmation
testing and/or notification of donor, and the resolution times range
from a mean of about 25 to 60 days.

	As far as resolution times as it relates to the single donor, that is
not done in our process.

	[Slide.]

	Now, there are factors influencing resolution time, and I think this is
pretty critical in trying to understand how this works and what the
issues are.  These factors, of course, are shipping logistics, when
samples are moved, when units are moved, and how they are managed.

	Laboratory capacity and through-put and even seasonality of infection,
and I will show you in the next slide exactly what I mean by that.

	[Slide.]

	This is from one member company.  This is about two years, and this is
the trends in the incidence of parvo-positive units over time.  You can
see that it is cyclic, as you would expect, and at certain points in
time there is a very low incidence, i.e.,  the through-put through a lab
would be very rapid, at other points in time there can be 5 to 6 times
as many positives to deal with, so that is certainly going to affect the
through-put of a lab.

	So, I think we have to be very cautious in describing or relating
turnaround times, that they can be a little inconsistent and they can
vary on us depending on the conditions.

	[Slide.]

	The next issue was the prevalence and levels of anti-parvovirus B19
antibodies, if any.

	[Slide.]

	To summarize, anti-B19 antibody level is not affected by the
implemented in-process control measures, that is, removing the high
titer donations.  About 98 percent of manufacturing pools, whether
before testing or after testing, are above 10 International Units/ml,
and there are no manufacturing pools below 5.

	This demonstrates appropriate strategy for effective management of
parvovirus B19 loads in manufacturing pools while, importantly,
retaining necessary antibody levels.

	[Slide.]

	I will just make a couple more comments about that.  We were also asked
if it was possible to look at serial donations and look at the temporal
relationship of positives, negatives relative to serial donations, and
this happens to be a very high frequency donor, a profile from the high
frequency donor.

	There are about 12 of those, and this is one example, and this is meant
to be a prototype.  Each box represents a donation and the status of
that donation relative to our threshold, whether it is above or below
the threshold.

	What you can see is that at this particular point, we were very lucky
and we identified a donation actually at the point of initial infection,
and the increase in titer was very rapid up to a very high titer, and
then the next donations decreased very rapidly until it decreased below
the threshold, and then went back to nondetection or non-elevated based
on our tests.

	This does not mean there is no titer.  This means that it is
non-elevated relative to our testing paradigm.

	There is one important point to make here relative to antibody levels,
and that is that what we have superimposed is the expected IgG profile
based on publications based on the literature.

	We also have data that actually confirms this, so we have got data
where we actually looked at titer relative to donations, and we looked
at IgG.  What you can see, I think the first speaker already pretty much
said this, is that most of the high-titer donations that we remove are
not likely to have an impact on B19 antibody titers in either
manufacturing pools or immunoglobulin products.

	However, if we remove a great deal of the low-titer donations, then,
the possibility exists for a significant impact on the antibody titers
in both manufacturing pools and final product, and that is clearly one
of the main reasons why we have adopted this particular paradigm.

	[Slide.]

	So, in conclusion, then, PPTA member companies have implemented
appropriate processes which have been shown to be effective in managing
parvovirus B19 in manufacturing pools, thus achieving an increased
margin of safety for life-saving plasma protein therapies.

	Ed Gomperts is going to discuss public health impact of donor
notification and counseling.

Edward Gomperts, M.D.

	DR. GOMPERTS:  Mr. Chairman, colleagues, thank you for the opportunity
to talk to you this afternoon.

	I will focus briefly on parvovirus B19 infection,  summarizing briefly
what actually you have already heard, and then talk about resolution
times, as well as issues around donors and contacts.

	[Slide.]

	The infection itself and the virus is well documented, well reviewed in
standards, infectious disease textbooks, such as The Principles and
Practices of Infectious Diseases.  Essentially, it is an acute
self-limiting disease without chronic sequelae in normal individuals,
normally transmitted by the respiratory route.

	Most infections are asymptomatic.  Where symptomatic, the donor would
be deferred, symptomatic being fever, headache, malaise, myalgias, and
rash.

	Antibodies to parvovirus B19 confer life-long protective immunity. 
More significant sequelae are rare and usually occur in particularly
susceptible non-donor populations with pre-existing conditions.

	[Slide.]

	Parvovirus B19 infections typically resolve with the appearance of
neutralizing antibodies, in the case of IgM, approximately 10 days post
infection and 17 days, IgG, post infection, with a period of viremia
being about 14 days and in some cases this viremia may well persist for
a substantial period of time.

	The intense viremia, however, develops approximately one week after
infection, and this usually lasts about a week.

	[Slide.]

	Focusing on the donor notification and the counseling issue, as we have
heard from Steve Petteway, the average resolution time for NAT testing
ranges from 25 to 60 days.

	Additional time would be required to identify the unit, perform the
necessary confirmatory testing, and then to locate and communicate with
the donor.

	This is a fairly substantial period of time relative to the infection
and therefore an infected donor would already have cleared the virus and
developed sufficient antibodies to confer life-long immunity by the time
notification occurred.

	The infected donor also, on the basis of this fair amount of time, will
already have passed the infection to close contacts by the time of
notification.

	[Slide.]

	Focusing on the at-risk populations and close contacts.  From the point
of view of the donor population, these individuals are deferred.  There
are standard questions, for example, "Are you feeling well and healthy
today?" which ideally would exclude the individuals who have an acute
infection, that are feeling ill.

	Female donors, the question is asked, "In the past weeks, have you been
pregnant or are you pregnant now?"  Certainly, the questions would
exclude immunocompromised individuals.

	From the point of view of prevention of transmission of infection to
close contacts, as already mentioned, the turnaround time mean is about
25 to 60 days, and confirmation testing would be a minimum of additional
10 days and donor notification, anywhere from 3 days to months.

	[Slide.]

	In conclusion, this medical information related to an acute parvovirus
B19 infection would be nonactionable for both the donor and his or her
close contacts.  On focusing on the ethics, we may question the ethic of
notification of a donor regarding nonactionable medical information.

	Certainly, counseling a donor regarding nonactionable medical
information certainly presents difficulties.

	Finally, donor notification and counseling lacks public health benefit
as this is a non-chronic, acute, short duration viral infection which is
highly prevalent in the general population.

	[Slide.]

	To conclude and bring the presentations together, in-process control
measures are designed to enhance the safety margin of plasma therapies.

	Parvovirus B19 NAT test lacks value as a diagnostic or donor screening
method.

	Thank you.

	DR. NELSON:  The last three speakers are open for questions or
comments.

	DR. SIMON:  Just one quick one.  What is the confirmatory testing for
this virus?

	DR. GOMPERTS:  It would be a repeat B19 NAT test.

	DR. SIMON:  You would just repeat to make sure there wasn't an error?

	DR. GOMPERTS:  On the specific unit that is collected from that
specific donor.

	DR. DiMICHELE:  I can ask you, but the question would be for anybody. 
Has anybody ever heard of the CDC identifying a contact parvovirus
infection in a contact of a recipient of blood products at all?

	In other words, has the CDC ever identified infection in the contact of
a blood product recipient or a plasma product recipient to the best of
your knowledge?

	DR. GOMPERTS:  I don't know.

	DR. YU:  I believe for CDC, B19 infection is not a reportable disease,
right, Dr. Chamberland?

	DR. CHAMBERLAND:  We wouldn't have the data to answer that question. 
There may be isolated case reports that are in the literature that
someone can speak to that talked about secondary transmission in a
household where a transfusion recipient acquired it from transfusion.

	DR. DOPPELT:  I am a little confused.  How are you setting your cutoff
for what you consider as a high titer and a low titer?

	 DR. PETTEWAY:  The cutoff, as I said, is coupled to the manufacturing
process, and it is coupled to the target of achieving no greater than
105 International Units/ml in a manufacturing pool.  If that
manufacturing pool is 800 liters, then, the minipool screening and the
cutoff is going to be different than if the manufacturing pool is 5,000
liters.

	So, it is all linked together and it depends on the manufacturing
process for any given company.

	A cutoff at the minipool level or at the donor level, when that unit
now is diluted into the manufacturing pool,  the cutoff will be 105. 
The titer of that unit that is diluted will be dependent on the volume
of the manufacturing pool.  That is how companies are setting their
cutoff, and the key is the specification or the target at the
manufacturing pool.

	DR. DOPPELT:  How are you picking that particular target?  I mean is
this arbitrary?

	 DR. PETTEWAY:  No, actually, it is not.  Remember it's coupled to
viral inactivation and removal, and we picked that target because we
need to go below the target, so if we are at 105, in order to assure
that we don't go above 105, then, we need to be around 104, so if we are
around 104, so now we are excluding donations, some of which are
actually below 105, if we were to go much lower than that, we would
start excluding the low-titer donations that are high IgG, and we are
trying to avoid that.

	So, what we are trying to do is to eliminate the highest titer
donations without eliminating the donations that are high in IgG
antibody.

	DR. NELSON:  And by "viral inactivation," you mean antibody primarily,
right?

	 DR. PETTEWAY:  No.  I mean within the purification process for
manufacturing, we have the capacity to remove or eliminate virus to a
certain level, and the idea is to get the manufacturing pool, reduce the
load in the manufacturing pool, so the challenge on that capacity is
less and the margin of safety is greater.  They are coupled together,
and I think that is important to understand.

	DR. LEW:  I think I have a question that may be linked to what was
asked earlier.  My understanding is when CDC did their presentation,
they mentioned about genomic equivalents/ml based on a study that looked
at healthy donors, and you wanted to avoid anything that was greater
than 104 genomic equivalents/ml, because those were nontransmitting
lots.

	We didn't really get the details of that study where we get this cutoff
from, 106, and that is a problem that I am having, as well.  What are
the studies that show 106 is the greatest, which is a little different
issue looking at keeping antibodies in our plasma or IVIG.

	So, the first question is what is the correlation between International
Units/ml with genomic equivalents, is that 1 to 1?  The second is could
we get some details on how that particular cutoff was chosen, either the
104 or 106 based on how you manufacture?

	DR. NELSON:  Wasn't the cutoff chosen based upon a study that showed
transmission of some pools that had very high titers?

	 DR. PETTEWAY:  No, actually, that may not even be relevant to what we
are doing, so I will put a slide back up and try to explain.

	DR. LEW:  There is a different issue between trying to keep good
antibodies in these products versus this cutoff of not wanting to
transmit, and I would like better clarification on that.

	 DR. PETTEWAY:  Let me see if I can help you here. Remember that prior
to implementing testing, we had manufacturing pools or production pools
that were up to 109 International Units/ml.  During our purification
processes,  we are capable of removing virus, but removing virus to a
certain level.  That gives us a margin of safety, but it is based on the
starting load, how much can you remove.

	What we wanted to do was reduce this load, so by minipool testing and
removing the high-titer units, we are able to reduce the load to a
defined--we can call this a cutoff, but a defined specification for the
manufacturing pool of 105 International Units/ml.  That is the goal.

	Now, when that is coupled with the process viral reduction that is the
same here, but with a lower titer or a lower initial titer, then, we
have increased the margin of safety for the product.

	The transmissibility in solvent detergent plasma of 105 would not be a
criteria for choosing the load here. The criteria for choosing the load
here is a balance between removing the most virus possible while not
eliminating the most IgG possible.  That is why that was chosen, not
based on the 104 S/D plasma experience.

	DR. LEW:  Could you just say what is equivalents between genomic
equivalents versus International Units?

	 DR. PETTEWAY:  Well, the reason you have an International standard is
because people--John, go ahead.

	DR. SALDANA:  John Saldana from Canadian Blood Services.

	The correlation between International Units and genome equivalents is
about 1 to 0.6 or 0.8, and the reason we use International Units is to
get away from the discrepancy of people using different units.  I think
it was quite clear at the West Nile Virus meeting in November that
people were using copies/ml, genome equivalents/ml,  et cetera, and it
is very confusing, so the WHO has adopted the IU, which is an arbitrary
unit.

	I think that tends to standardize.  It is completely arbitrary, but we
try and make it as close to the genome equivalents as possible.

	DR. GOLDING:  Basil Golding, FDA.  This cutoff of 106 is obviously
arbitrary and it is true that most manufacturing processes will remove
virus, but the ability to remove virus is variable and this virus is
removed to a lesser extent than enveloped viruses, and most
manufacturing processes that I am familiar with will remove possibly 4
logs of virus,  so you are still going to have virus in the product.

	The point about the antibody, the caveat there is if you are making
immune globulin, sure, you will have antibody in the product and it is
going to neutralize low levels of virus, but if your fractionation
process separates your virus from your product like it could do for
certain calculation products, you could end up with 3 or 4 logs of
parvovirus in your product without any antibody in the product, so that
product could presumably transmit the virus, and if you look at the
hemophiliac population, the antibody titers compared to the rest of the
population indicate that that is exactly what happens.

	 DR. PETTEWAY:  Yes, that's true, and that is another reason why our
paradigm and the cutoffs that we chose wasn't based on the plasma S/D
experience.  I would also note that in many processes, we have validated
capability of removing parvovirus up to 108 or 109, and in others it is
less, so that is correct.

	DR. SCHMIDT:  I would like to see us remove from the discussion of what
we should do, the point about the contacts of the donors for three
reasons.  First, I think we are all interested in public health, but I
don't think we should add to the cost of blood and blood products some
public health activity, such as caring for or considering the contacts
of our donors.

	Secondly, I don't think it is within the authority of the FDA to
consider that.  We are supposed to be worrying about the product and we
are already moving back to the donor's health, and now we are talking
about the health of contacts of the donors.

	Thirdly, I think nowadays is this whole question of invasion of privacy
or not invasion of privacy depending which side of the fence you are on
these days, but we might not have any business letting people know about
those things when they didn't ask us for them.

	DR. FALLAT:  It seems to me, though, pertinent to that is there is a
big difference I guess between the plasma fractionators and their time
period and the blood center.  We have to keep that in mind when we look
at the time period where the donors or the recipients might be
contacted.

	DR. ALLEN:  I would be very cautious about certainly voting negatively
on this question, however, in the absence of some medical ethics
considerations, in the current environment, if an organization or an
agency has that kind of data, and it is taken down to the individual
donor level, I think there are many people who would feel that there is
an obligation to get that information back with an appropriate
explanation ever if there isn't any necessary medical or public health
significance to it.

	I think that goes back to the question of if you have got minipools, is
there an obligation to test back to the individual unit, and that
perhaps could be the level of discussion.  I think if you have taken it
back to that level of identification, there may be an obligation to
inform the person.

	DR. KLEIN:  I would agree with that, Jim.  I think that if you have
that information about a donor, and it was I who donated last week, and
you told me, I might not visit my pregnant daughter next week or I might
not go to the obstetrical unit or to the hospital where there are
immunodeficient subjects.

	I think that us not discussing this and deciding whether or not this is
an issue, we would be punting on that one.

	DR. SCHMIDT:  I guess this goes back to one of my old arguments with
Toby who presented the point of view that the plasma industry was
considering serologic testing for syphilis as a good thing because if
it's public health aspects, then, I see some relation here.

	DR. SIMON:  I never advocated that.  From a historical perspective,
that is the way it came about, you know, that the testing every several
months at a time when syphilis was more prevalent, I don't know if you
want us to start discussing this or not.

	DR. NELSON:  We will have time to discuss the questions raised by the
FDA.  Since we have come this far with all the presentations and people
have flights and are going to have to leave, could you present the
questions again?

	DR. WHITAKER:  Could I just answer the question that was raised a
minute ago?

	DR. NELSON:  Go ahead.

	DR. WHITAKER:  I would like to remind the committee that the test is a
threshold test.  You are not going to identify every one who has
parvovirus when you do the test, so you will be identifying individuals
with high titer viremia, but you may also be not identifying people,
individuals with high titer viremia just below the cutoff.

	The purpose of the test is the product and assuring a high margin of
safety and that not diagnosing a donor, and that there are some issues
here that really do need to be discussed and considered.

	It is not the same as an HIV test.

	DR. NELSON:  There are four people that wanted to make a statement in
the open public hearing.  If you could make a brief statement or even
provide it for the record or what I would like to do is maybe discuss
the questions, but the first is Kay Gregory.

	DR. CHAMBERLAND:  Ken, can I just ask a question because there is a
time issue here.  I think the committee really wants to give this a
thorough discussion and it is a difficult issue, and there are four
questions.  There is also open public hearing that has to take place.

	I think there is, by my watch, about 45 to 50 minutes left before the
scheduled adjournment and I think many of us, those of us from out of
town, scheduled flights to accommodate a 6:30 adjournment.

	If people realistically think that all of that can happen in 45 to 50
minutes, then, that is what we planned for, but if it is not
realistically able to happen, I think people are just feeling kind of at
a loss as to what exactly to do here.

	DR. NELSON:  I think if we have come this far and then we discuss,
let's say, the questions again at the next meeting, we would have to
sort of revisit all the issues.  I would like to try to do it in the
next 45 minutes if we could.

D.  Open Public Hearing

Kay Gregory, AABB, ABC, ARC

	MS. GREGORY:  You have the written statement and this time I am
actually representing the AABB, America's Blood Centers, and the
American Red Cross.  I am happy to just let you have it on the record,
but I think it may be giving a little bit short shrift to the whole
blood industry if you don't hear the statement.

	Primarily, what I really want you to be aware of is that given the
important and compelling competing safety priorities of implementing
West Nile Virus donor screening and performing bacterial detection in
platelets for the whole blood sector, and we are going to be doing this
in the next six to nine months, the additional capacity and work that
would be required to perform parvo B19 NAT as a donor screening test
simply cannot be absorbed.

	For example, performing it as a donor screening assay would require the
addition of another on-line assay requiring completion prior to all
product release, the pulling of samples and further testing to resolve
positive pools, the need for a confirmatory assay, and the alteration of
510(k) cleared computer systems to accommodate parvovirus B19 results as
a release criteria.

	Furthermore, current FDA policy would require that donor screening be
performed under an IND or an IDE, which would be an additional burden
for test kit manufacturers who are turning their efforts to West Nile
virus test development.

	We believe that the practical solution of performing parvovirus B19 as
an in-process control at this time is supported, and we are willing to
look at doing that, but we think going any further to call it a donor
screening and requiring notification, et cetera, is more than we can
absorb at this point in time.

	DR. NELSON:  Thank you.  That was a good summary.

	Let's move to the discussion and questions.

E.  FDA Perspectives and Questions

for the Committee

Mei-ying W. Yu, Ph.D.

	DR. YU:  I will try to be short.

	[Slide.]

	The first FDA perspective.  For whole blood donations, risks to
transfusion recipients are sufficient to warrant withholding high-titer
individual positive units that is greater or equal to 106 genome
equivalents/ml prior to release of blood components to use in
transfusion.

	This particular level was set, it is to minimize the risk of infection
in recipients and to prevent serious consequences of B19 infections in
high risk recipients and to avoid the removal of low titer units that
may not be infectious and which contain protective antibodies.

	[Slide.]

	Now, this is B19 profiles of B19 DNA and antibodies from the serial
bleeds from normal source plasma donors.  Dr. Gerald Zerlauth of Baxter
BioScience presented in December 2001 FDA Workshop, and we have been
collaborating with him very closely, so this is a very precious panel I
just want to point out, but unfortunately, NGI also have similar panels
from serial bleeds, I believe from 20 donors and show very similar
pattern.

	In red or reddish pink, that is the DNA level, and the B19 DNA level,
and in yellow is an IgM profile, and the blue is the IgG profile.  So,
as you can see from this source plasma donor, the B19 level goes up very
quickly to 1012 and then it drops quite sharply to 106 or a little bit
below when IgG became positive.  This is at day 14, anti-B19 became
positive, and the titer is around the 7 times 105 genome equivalents/ml.

	But the viremic period can be very, very long.  See, it tapered off
here, but then it remained very, very long time through 304 days, but
that is his last bleed, so it is 102 or 103 genome equivalents/ml level.

	The IgM was positive at day 10 and then the level is still very high, 2
times 1010 genome equivalents, so this 106 genome equivalents/ml above,
then most likely IgG will be negative.

	Now, we really don't know what is the infectivity, the minimum
infectious dose especially for those unpooled products that has no
antibody, but I want to tell you that the IgM in this particular donor
became negative at day 60 and this donor was positive at day 14, like I
said, and then later on it actually gradually increased and the level
reached to about 50 to 60 between 70 to 90 days IU/ml, 50 to 60.

	DR. NELSON:  This has been presented before to the committee.  I
wonder, could you move to the questions that you want us to consider?

	DR. YU:  Okay.  I will.  This is the level of IgG became around 1030 or
1025, that is what is found in IGIV level, in terms of 1 percent IgG
concentration.

	I wanted to answer Dr. Toby Simon's question.

	The key thing I wanted to show that profile is to say that you cannot
be too sensitive.  When you are too sensitive of the NAT, then, you are
getting those low level units that may not be infectious, and they
contain IgG.

	[Slide.]

	The second one is a temporary deferral may be warranted for high-titer
apheresis donors if positive donations can be resolved within several
weeks.  As you know, the donation intervals are 8 weeks for whole blood,
8 to 16 weeks for red blood cell apheresis, 48 hours for
plateletpheresis, and every 48 hours for plasmapheresis.

	Based on the industry presentation, we will see the resolution time. 
In Susan Stramer's presentation, she said that can be in 24 to 48 hours,
the mean time, so I put down two days here for the Phase 2 approach. 
Now, NGI unfortunately, Andy Conrad cannot come to present, but in one
of his slides, the average time to resolve to single donation is 4.6
days to be exact.

	Within such a short time, either two or five days, you can really
notify the donor, defer the donor if necessary, and then if there is a
medical benefit for the close benefit, they can be notified quickly.

	FDA actually got inquiry from plasma centers and they asked how long
they can defer these positive donors, but however based on the PPTA
presentations, you can see various companies.  The mean time, resolution
time is ranging from 25 to 60 days, so within the time period, you
really cannot do very much for donor deferral or medical benefit to the
close contacts, but the range is very, very high, you know, A and B
companies can be as short as 8 or 9 days.

	[Slide.]

	The third point is that FDA is seeking the BPAC's opinions on
conclusions made by the Ad Hoc PHS panels that there are sufficient
potential medical benefits to close contacts, but not to donors, to
warrant notification of parvovirus B19 donors.

	However, we believe such notification is likely to be useful only in
setting where testing and notification can be completed within, for
example, less than four weeks of donation.

	So, the questions for the committee.

	First, if donations of whole blood are tested for the presence of human
parvovirus B19, are risks to transfusion recipients sufficient to
warrant withholding high titer positive units that is equal or greater
than 106 genome equivalents/ml from use for transfusion?

	Is temporary deferral of positive donors warranted in the setting of: 
(a) whole blood donation?  (b) apheresis donation?

	The third question is:  Do potential medical benefits to contacts of
parvovirus B19 infected donors warrant identification and notification
of positive donors?

	Fourth.  If yes to Question 3, should donor notification be limited to
settings where testing and notification can be completed within several
weeks of donation?

	That's it.

Committee Discussion

	DR. NELSON:  Discussion?

	DR. SIMON:  Did you want to do these one at a time?  I think it is a
philosophic issue here.  We have started with a test which was an
in-process control, and it has somehow evolved into consideration as a
specific test for the removal of in-date units and for counseling of
donors.

	It only detects people who are at very high titers, so if you donate a
day or two, before you hit that titer, it won't detect you, if you
donated a day or two after, it won' detect you, so its public health
usefulness is very limited, it doesn't have the same kind of testing
characteristics for HIV or hepatitis B or hepatitis C.

	I think what has created a red herring here, I guess is the Red Cross's
intention, in its Phase 3, to do this in concert with the other tests
and to have a positive test result at the same time as they do for the
other viral markers, in which case they could remove the units, and I
guess it would make sense to do so, but that Phase 2 is a while away and
as they pointed out, even in their situation, there is still more time
needed before they would be able to contact the donor.  They have to
confirm the test result and then they have to put in the process all the
measures to contact the donor, and there is just a lot of other time
taken up.

	I think in terms of the industry, the plasma industry, you know, this
is coming way beyond any time for medical usefulness.  So, I just think
we have gotten off.  I guess, number one, I think if you had the data
within the same time frame as you do the others, I guess you would pull
in-date units, but I think it is unfortunate that we moved from looking
at this as an in-process test for the plasma fractionation product, to
begin to look at it for these other purposes, because obviously, a
somewhat lower titer in that setting could cause the problem.

	So, I think it is unfortunate.  I mean I guess that the common sense
answer to Question No. 1 is yes, but I certainly wouldn't defer the
donor who will get over the problem, and I think contacting, it is true,
of course, that you always have this ethical issue when you have
information, but it is not very useful information and the timing of it
is such that it is going to come at a time when it won' be useful to the
donor or the contacts, because simply the time it takes to do all of
this, whether you do it by certified letter or phone call that goes into
a voice mail, and by the time it gets back into the system, so that I
think is I guess representing industry, kind of my philosophic look at
it.

	I would hope we would not saddle the plasma industry with having to go
back to tell people 30, 40 days after they have donated about this.

	DR. SCHMIDT:  Considering the whole blood, we have heard a statement
from Ms. Gregory that they can't do it right now without impeding other
perhaps more important activities.

	Wouldn't it be sensible for the FDA to table this request for us to
consider this?  I know it means taking it up again next year, but that
might be a cheaper alternative to having everybody working on this
before then.

	DR. KLEIN:  We have spent a lot of time on the plasma industry, and as
best I can tell, the questions aren't addressing that, and it doesn't
make much sense, since they are not resolving to the single donor, and
the time frame would make that--

	DR. SIMON:  I think they are resolving to the single donor.

	DR. KLEIN:  If they are resolving to the single donor, then, the time
frame would make it impossible really to have any medical benefit either
to a donor or to the donor's immediate contacts or even distant
contacts.

	However, if, in fact, we are going to be resolving to the individual
donor within 48 hours, then, I think we need to address these issues
even though we may not be doing that for the next two or three years.

	I think we at least have to get on the track.  I think the FDA is
asking us for that advice.  I am not saying that we need to do it
tomorrow.  It seems to me that if you have a unit of blood that has a
high titer test positive confirmed for parvovirus, you simply don't want
to transfuse it.  I can't imagine that you would ignore that, so I think
the answer to No. 1, in my mind, is yes, and I presume we will get on to
No. 2 eventually.

	DR. NELSON:  Let's vote on No. 1.

	DR. BIANCO:  Let me just ask Dr. Klein an important question.  What if
in the whole blood sector, this testing is done after expiration of the
cell or components for the units that are going to recovered plasma,
which is what Sue Stramer presented?

	DR. KLEIN:  I think again the question we are being asked is about if
you have an in-date unit and you have a test result that indicates that
it may be infectious, and not only infectious, but potentially cause
morbidity and mortality.  I mean that is the question.

	If you want to pose the question differently, I may have a different
answer.

	DR. NELSON:  Jay.

	DR. EPSTEIN:  It may be helpful to realize that the terms of debate
have shifted over time.  The source plasma standard has caused the need
for whole blood collectors to implement parvovirus testing, so that they
can sell recovered plasma.

	This has caused the FDA to consider what is going on in the whole blood
scenario, and the way we looked at it is, well, if you are now testing
whole blood donors, shouldn't you have a proactive position to interdict
the at-risk unit.

	Now, what has evolved is that it can't be done immediately, don't allow
it, priorities may not allow it, but we are sort of looking ahead and it
was not clear some months ago whether there was, in fact, an industry
intent to do what is called real time testing, which basically means
testing as a release test.

	There is this gray zone where you may not be testing as a release test,
but you have an in-date unit, and you might or might not get the
opportunity to interdict it. That is an unpleasant place to be.

	So, you know, you sort of have these three scenarios.  You have testing
of outdated units for the purpose of screening and recovery of plasma. 
There is nothing further you could be doing about transfused units, and
a lot of time has passed with regard to any value notifying a donor.

	At the other extreme you have testing within 48 hours compatible with
other release testing, and then you have this gray zone in between where
you have some delay in testing, but you still have some in-date units.

	So, what the FDA is looking for is, first of all, an opinion whether it
is important to interdict these units because they are not being
interdicted now when there is no testing.  Then, we are looking for a
direction whether we should be pushing from a regulatory standpoint that
all the testing should, in fact, in whole blood, become pre-release
testing eventually.

	If testing is feasible and if ultimately, it is feasible as release
testing, shouldn't that be what happens.  So, that is where we are
coming from, and we recognize that you can't necessarily have it
overnight.  I mean I think we understand that point.

	DR. FITZPATRICK:  To me, that is a different question.  What Dr. Klein
said was if we know, we should interdict.  What you said is should we
test to interdict, and that to me says does this represent enough of a
risk to the patient population that we should advocate pre-release
testing.

	I didn't see presented today any more information than was available in
1999 on cases of transmission by transfusion.  So, those are two
different things to me, and I am not sure where you want us to go with
that.

	DR. EPSTEIN:  Well, my feeling is that if whole blood donors are to be
screened, that we should work toward pre-release testing for the purpose
of interdicting potentially infectious units and that, as you say, you
have already heard that high-titer units are almost certainly
infectious.

	We know that they are a serious threat to some recipients.  We don't
have good data on the frequency of clinically significant events.  Now,
we don't have any more data than we have previously prevented.  I think
that is part of the problem, but we were trying to focus today's meeting
primarily on the issue of benefit or lack of benefit of donor
notification.

	I appreciate the discussion of Question 1 has raised the additional
dimensions of that issue.  I think we could split it into two questions
if you like.  One is, is there a benefit to interdicting
parvovirus-positive units, and the other is, if testing is done now,
should it become pre-release testing.

	Is your feeling that you can't vote this question or you don't know
what it means?

	DR. FITZPATRICK:  I am concerned about the phrase, "Are risks to
transfusion recipients sufficient to warrant withholding high positive
titer units?"

	You know high-titer units are infectious.  To me, if we answer yes to
the question, we are advocating an effort by the industry to engage in
an effort to do pre-release testing to protect the recipient.

	DR. EPSTEIN:  Well, I would say that there would likely be an evolution
of policy and that the first step would be retrieving in-date units
found to have high titers and a vote in the affirmative would encourage
FDA to push toward ultimately pre-release testing, yes, but it doesn't
all have to happen at once.

	But, yes, a vote in the affirmative would put us on that course to
progress from retrieving in-date units potentially with lookback
notifications to an ultimate pre-release testing scenario.

	DR. FALLAT:  We have data from the blood banking industry that there is
106 titers in perhaps 1 in 15,000.  We give 1.5 million units of blood a
year.  That translates into quite a number of people getting that titer,
of which a certain percentage will be in the high-risk group.

	For me, it is no greater to vote yes on No. 1.

	DR. SIMON:  Well, for me, it is unfortunate, the implications in No. 1,
because I think if somebody told me they had tested, and it was a
high-titer unit and should they remove it, I would have to say yes.

	I mean I can't imagine a different answer, but on the other hand, I
would not want to encourage the FDA to move towards requiring this
testing as a donor test, in other words, because I think it takes us off
the track of the rationale for it, and I don't believe, as Dr.
Fitzpatrick said, that data in the past have suggested a need to look
for this virus or to prevent this virus transmission in whole blood,
platelets, and so forth, but rather as a problem in pooled product
particularly to hemophilia patients.

	So, I think it is unfortunate that there is that implication with a yes
vote, but I agree with Dr. Fallat.  I mean I don't see how one cannot
vote yes to No. 1 if you have that information.

	DR. EPSTEIN:  Could I suggest that we add a question to give you the
opportunity to clarify this, which would be:  Has a value for screening
of whole blood donors for parvovirus B19 been established?

	In that way, if you wish to vote 1 in the affirmative, you can still
vote 2 in the negative.  I think that would clarify things if I
understand the issue here.

	DR. KLEIN:  I would like to have that first part that is now split off,
I like the wording, because I think the wording is very important.  We
really have never looked for this, so we don't know whether it is a
problem or it isn't a problem, so you really don't want to exclude that
any more than you want to press forward with it in the absence of data.

	DR. NELSON:  Great.  Certainly, parvovirus B19 infections are a
significant problem in patients with AIDS and sickle cell, and all the
rest, but we don't know how much of it is transfusion transmitted, and I
guess that is the real issue.

	DR. SIMON:  We can vote on No. 1, I think, while he is writing No. 2.

	DR. NELSON:  Let's vote on No. 1.

	DR. SMALLWOOD:  Question No. 1(a).  If donations of whole blood are
tested for presence of human parvovirus B19, are risks to transfusion
recipients sufficient to warrant withholding high-titer positive units
greater than 106 genome equivalents/ml from use for transfusion?

	Allen.

	DR. ALLEN:  I think the data aren't certain, but I am convinced that
the answer probably is best yes.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  Yes.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  Yes.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  Yes.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  Yes.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  Yes.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  There is unanimous yes for Question 1(a).

	Question No. 1(b).  Has a value to blood transfusion recipient been
established that is sufficient to warrant donor screening for human
parvovirus B19?

	DR. SCHMIDT:  I am sorry.  Would you read that again?

	DR. SMALLWOOD:  Yes.  Has a value to blood transfusion recipient been
established that is sufficient to warrant donor screening for human
parvovirus B19?

	DR. FALLAT:  It that for whole blood transfusions or are you separating
out transfusions?  Yes?  Okay.

	DR. SMALLWOOD:  Roll call.

	Allen.

	DR. ALLEN:  I think most of the discussion I heard was really to the
absence of data although we agree that there certainly is a potential
risk out there especially from high-titer units.

	I am going to have to, in terms of the way the question is worded, Has
a value been established, the answer is no.  We need studies.  I think
there is a potentially very significant risk out there to certain
populations.  It is a real concern.  I don't think we have the data now.
 No.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  I would vote no for the same reasons.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  I am going to abstain.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  No.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  No.

	DR. SMALLWOOD:  Lew.

	DR. LEW:  No.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  No.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  No.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  No.

	DR. SMALLWOOD:  The results of voting:  2 yes votes, 9 no votes, 1
abstention, and the industry representative agreed with the no vote.

	Question No. 2.  Is temporary deferral of positive donors warranted in
the setting of:

	(a) whole blood donation?

	(b) apheresis donation?

	DR. SIMON:  Is apheresis here plasmapheresis?  I am getting the word
that it is, (b) is plasmapheresis as of the plasmapheresis industry.

	DR. KLEIN:  (b) could also be plateletpheresis.

	DR. SCHMIDT:  Do we know what a positive donor is?

	DR. NELSON:  No.  It could be 102 or 1040.

	DR. ALLEN:  Remind me again.  With the plasma industry, my
understanding is we are really talking weeks or longer between the time
that the testing is done and any results are available, so
plasmapheresis, I mean it's a moot question.  Plateletpheresis is the
testing is done reasonably rapidly.

	DR. SIMON:  That is an interesting question.  Ordinarily you wouldn't
do it on plateletpheresis since there is no recovered plasma.  I mean if
you take our vote on 1(b), go with the majority, you wouldn't do it on
plateletpheresis unless you are making recovered plasma with it, but
ordinarily you wouldn't be.

	DR. FITZPATRICK:  With 1(b), to me, until you resolve 1(b), you can't
move on to 2(a) and (b).

	DR. KLEIN:  I don't really agree with that.  I think if you have got a
positive unit, then, what do you do with that donor?  You have got a
high-titer positive unit sitting here, and you have a donor, someone who
is going to come in 56 days later.  Then, I think the answer is pretty
obvious, but someone who might come in, in 48 hours, you have to think
about it.

	DR. SIMON:  Well, 56 days later, I assume you are saying you would not
defer, and 48 hours you would except you won't know that for three or
four weeks.

	DR. KLEIN:  Not in plateletpheresis, should you be doing it for
plateletpheresis, and I guess there are protocols where plasma and
platelets are collected, are there not, Jay?

	DR. EPSTEIN:  I am thinking.  Again, it is a case where splitting
rather than lumping.  FDA brought it forward this way because we were
thinking about frequent collection, and we were neutral about how long
could it take to do the whole cycle of testing, because there is so much
variation going on.  We are not making the assumption things stay the
way they are.

	But I think for the moment it would be helpful to split out apheresis
from plasmapheresis.  So, basically, the two scenarios come down to the
whole blood apheresis donor to make transfusible components where that
donor may indeed come back in 48 hours to give platelets again.  Part
(c) would be the scenario of source plasma donation.

	So, if we would say whole blood and apheresis donation from whole blood
donors, in other words, the donors who meet the whole blood standard, so
apheresis donation to make transfusible components, and then (c) would
be source plasma donation.

	DR. FITZPATRICK:  So, Jay, in following Dr. Klein's, would you consider
this the same as 1(a), if you had the result and knew the result in time
to make a decision, would you make one?

	DR. EPSTEIN:  Well, I think having posed and heard the vote on 1(b), we
are not now really thinking in terms of the scenario where it is all
pre-release testing, so we are back to the scenario where you might be
learning later.

	On the other hand, the donor, even though you learned later, even
though maybe it was 14 days, the donor could still be coming back, in
other words, they are not on a 56-day cycle.

	DR. NELSON:  This all one question that includes (a), (b), and (c).  Is
that right?  Vote separately?

	DR. SIMON:  Yes, separately.

	DR. NELSON:  Let's do the first, 2(a).  This is a whole blood donor
with an interval of 56 days?

	DR. SIMON:  2(a) would be, I believe, a whole blood donor with interval
of 56 days.  2(b) would be potentially plateletpheresis, which could be
twice in a week.  2(c) would be plasma donor, which could be twice in a
week, but you don't have the results for three to four weeks.

	DR. NELSON:  Right.

	DR. DiMICHELE:  The nuance of this, the time of notification I think is
critical to answering this question because I think in answering
Question 1, I mean I think we were sort of looking at the data that was
presented by the FDA and the American Red Cross, and the possibility of
getting this information out in two days, which is very, very different,
I think, given the period of viremia of someone who is determined to be
positive.

	So, given the overlapping period of viremia and basically the
identification and notification time, I mean I think those two things
are very, very critical.  If the notification time extends past the
period of viremia, the question is a moot point.

	If the notification time is included in the period of viremia, then,
you are absolutely right, then, we vote maybe the same or differently on
2(a) and 2(b).  I mean I think that this is an issue that has to be
clarified before we can vote rationally.

	DR. SIMON:  I think the information we were given is that the whole
blood segment could at some time move to having the data available
within 48 hours.  The plasmapheresis situation would not.  Those units
are all shipped to central testing laboratories that take longer to do
it, and also we have to keep in mind with (c), the level of antibody in
the final product because the donors that are then forming IgG are
people you would want as donors for IgG.

	I think that suggests that (a) and (b) you would probably say yes, and
(c) you would say no, and that would be my view of it.

	DR. NELSON:  (a), you would say yes with the 56 day?

	DR. SIMON:  I am sorry, I am getting confused.  (a), I would say no
because of the 56-day interval; (b), I guess you would have to say yes,
if you had it; and then (c), I would say no for source plasma.

	DR. DiMICHELE:  If that is what the question is.

	DR. STRAMER:  I just wanted to clarify time frames.  I said we would
have products tested by 10 hours to 48 hours, which is about two days at
the longest time, but for donor notification, by the time the donor gets
the test results, we may owe them a letter that is going to be two to
three weeks.

	DR. SIMON:  But if you wanted to defer a plateletpheresis donor, you
could put that in your computer.

	DR. STRAMER:  Right, that's true.

	DR. SMALLWOOD:  Question 2(a).  Is temporary deferral of positive
donors warranted in the setting of whole blood donation?  Vote.

	Allen

	DR. ALLEN:  No.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  No.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  No.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  No.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  No.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  No.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  No.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  No.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  No.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  No.

	DR. SMALLWOOD:  The results of voting for Question 2(a), unanimous no.

	Question 2(b).  Is temporary deferral of positive donors warranted in
the setting of apheresis donation from whole blood donations for further
components?

	DR. EPSTEIN:  Apheresis donation to make transfusible components.

	DR. SMALLWOOD:  To make, okay.

	Corrected 2(b).  Is temporary deferral of positive donors warranted in
the setting of apheresis donation to make transfusible components?

	Allen.

	DR. ALLEN:  Yes, and that's based on the assumption that the test
results are known within a short period of time.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  Yes.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  Yes.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  Yes.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  Yes.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  Yes, assuming it's not two-unit red cell apheresis in which
case it's 112 days.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  The results of voting for Question 2(b), unanimous yes.

	Question 2(c).  Is temporary deferral of positive donors warranted in
the setting of source plasma?

 	Allen.

	DR. ALLEN:  No.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  No.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  No.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  No.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  No.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  No.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  No.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  No.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  No.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  No.

	DR. SMALLWOOD:  The results of voting for Question 2(c), unanimous no.

	DR. NELSON:  Question 3.

	DR. SIMON:  This is the notorious contact question.

	DR. YU:  Do potential medical benefits to contacts of parvovirus B19
infected donors warrant identification and notification of positive
donors?

	DR. CHAMBERLAND:  My take on Dr. Brown's talk and when there was a
little bit of discussion about this, is that in terms of medical
benefits, if you stratify it by prevention of secondary transmission,
that just given the time frame, it is unlikely to happen.

	So, in terms of potential medical benefits, you are unlikely to prevent
secondary transmission to a contact simply because of the time
considerations and the type period when there is likely to be high-level
viremia that could be transmitted via the respiratory route.

	However, I believe he did hold out the possibility that in selected
situations, probably fairly rarely, that you might be able to have a
benefit in terms of potential treatment with modalities, such as IVIG
for some of the more severe manifestations of parvovirus B19.

	That was my take on it.  People are nodding their heads, they had a
similar--

	DR. BROWN:  That was my intention.

	DR. CHAMBERLAND:  Okay.  It is late in the day and I wanted to make
sure I (a) heard it correctly; and (b) restated it correctly.

	DR. NELSON:  We have already voted yes on 1, didn't we, notify, or was
that just defer?  Remove the product.

	DR. GOLDING:  Basil Golding.  Sorry, I will add it very quickly,  I
know it's getting late.

	A clinical benefit I see for people who have HIV and are getting
parvovirus infections and are getting anemia, aplastic anemia, and it is
going to last for a long time, and they are going to get stem cell
transplants, the doctor needs to know, so that they are not giving the
wrong treatment, instead of giving stem cell transplant, as an example,
where IGIV would have been much better.

	The same thing, if you have a pregnant woman who had a contact early in
the pregnancy, and the fetus is then getting into trouble, I think it
would be helpful to know what the causation was, and an intrauterine
transfusion would also be helpful.

	Also, the question of arthritis where you get long-term arthritis in
some woman, it would be helpful to know that it is not rheumatoid
arthritis, so there are diagnostic and other modalities that are
involved.

	DR. SCHMIDT:  I gave three reasons before why I thought no.  I would
just like to add to that.  When we started testing for HIV, we told
people not to come in just to find out if they were positive.  Those are
the bad guys, and we only wanted to be nice to the good guys, I guess.

	An interesting situation in the UK now, they are worried about if they
find a test for mad cow disease, that people will stop donating blood
because they don't want to know that they are positive for this.  I mean
it's a switch in the other direction.  But that might change if there is
some therapy for mad cow.

	We have fights about whether we are doing the wrong thing by giving
away free T-shirts, but also free cholesterol examinations.  I just
think we ought to stay out of the whole business and just do what we are
supposed to do.

	DR. BIANCO:  Dr. Nelson, I am very concerned about the consequences of
what is being discussed today.  We started and actually, Dr. Simon
presented it very well, with a process that was to try to make a product
for patients that receive those plasma derivatives better.

	Now, when we move to another way, that we created a complexity where
maybe one or two contacts a year in the country will benefit from a
process that will drive an entire community in the way they collect
blood.

	My concern is that those regulatory requirements will simply inhibit
us, so instead, people get it, contacts happen at home, they happen in
bed, husband and wife with a wife that is pregnant.  It is rare that we
have an event like Dr. Klein described.  It is possible, it is
plausible, but it is rare.

	If those requirements are imposed, this is only going to delay the
adoption of measures that could help make patients, certain patients
receive or allow certain patients to receive a safer product, because we
simply are not going to do it.

	It is so involved in so many requirements.  Unless there is a
regulation that tell us to do, and we know that that will take four or
five years at least to have a pre-release screening test that would
allow screening of all donors for testing for CMV, that would be an
equivalent model here, is testing that is voluntary and is done in a
relatively small number of units, which is what would probably be the
approach to deal with those patients at higher risk.

	I am just concerned about the implications that these will inhibit
progress because of fear of the impact of the regulation.

	DR. KLEIN:  I am going to disagree with that point of view.  I don't
know whether it will stop testing of single units or not, but it seems
to me that if you have tested individual donors, you have a test result
that could, in fact, impact on health.

	You (a) have a moral obligation to notify the donor of their test
results; and (b) you have a moral obligation to indicate what action
could be taken to prevent some infection, whether that is 100,000 of
them or whether it is three of them.

	If you just test pools, it becomes a moot point, but if you are testing
individual donors, and you are not giving the donor that result when it
may, in fact, impact upon either his or her health or someone else's, I
don't think that is the appropriate thing.

	Now, we are not talking about the ethical issue, we are talking about
whether there is medical benefit.  I think there might be a small
medical benefit, but I think if you are thinking in the patient's
interests, for those of us who are hospital based, I would want to do
that.

	DR. BIANCO:  I am sorry, Dr. Klein, I agree with you 100 percent.  We,
in our proposal, and unfortunately, the discussion, we did not, AABB did
not have a chance to present our joint program, our proposal has been
for minipool testing, it has not been for individual donor screening.

	If we come to the individual donor screening, even if we were doing
this limited number like we do for CMV, I think it has to be
communicated to donor on the basis of ethics and on the basis of
medicine, and I agree with you.

	But minipool is the issue today.  We are discussing an issue that
actually is going to impede the implementation of minipool because there
is a question can we test in minipool without resolving to the single
donor.

	What we heard today in the summary from Dr. Mei-ying is that the
understanding is that we should resolve those to the individual donor.

	DR. SIMON:  Maybe we should sort of divide this into what is and what
may be, and I think right now the plasma industry does have widespread
testing in order to provide safer product.

	As I understand it, in order to avoid throwing out units that are
perfectly good, they have in many cases gone down to the individual
unit.  They get this information about 20 or some days after the donor
has donated and by the time you would have notification, and so forth,
you would be talking about a month or so.

	I think at that point, the utility of transmitting this information is
extremely low, so I would hope that they would not be encumbered with
this obligation for an action they have taken to make the product safer
and for an in-process control because they happen to identify which
unit.

	I think if the blood banking organizations ultimately move to doing
this, like was reported by Dr. Stramer in her Phase 2, where they are
doing it along with HIV and hepatitis B before release of units, then,
it becomes another factor, and I think Dr. Klein's arguments would carry
much more weight.

	DR. FALLAT:  We are putting the scenario in Question 3, making the
scenario very different from Question 1, and now we are saying it's
minipools, and so we really don't know that the person that has the
positive viremia, therefore, it is not going to be possible to remove
that blood.

	It seems to me if the only thing that is going to be done is minipools,
then, we need data to find out just how big of a problem this is if you
did it on single donors or resolved it to single donors perhaps more
rapidly, because again if you go back to those figures, if you have 1 in
15,000 that have a high titer, and you are giving out in a year and a
half, you have got 1,000 donors that are receiving high-titer B19.

	I would guess that at least 10 percent of those will be people in high
risk groups perhaps, but this is all guess work.  I think we need that
data before we can press forward with single donor identification.

	DR. HEALY:   Dr. Nelson, this is Chris Healy with PPTA.  I just wanted
to make the committee aware of a point, and that is that the issue of
minipools and going down to the individual donations is really kind of a
red herring here.  The way that the testing is performed, the companies
do have unit identification bleed numbers.

	That can be accessed, whether you are down at the minipool level or
whether you are down to an individual donation.  What they do not have
is donor identification information, but information about an individual
unit, a bleed number, a unit identification number can be found out at
any point throughout the process, whether you are looking at a minipool
or whether you are looking at an individual donation.  There is complete
traceability throughout the entire process.

	So, the distinction between minipool and individual donation is really
immaterial here.  The critical distinction is do you have a donor's
name, do you have a donor's identification number, do you have the
center where that person donated, and are you in a position to contact
them.

	That information does not exist in the current strategies used for NAT
testing of parvovirus at the fractionator level.

	DR. FALLAT:  Would you clarify that then, is Dr. Simon correct in
saying it would take 20 days before you would identify that single
individual?

	DR. HEALY:  Yes, that is correct.  It takes quite a bit of time because
what we look at is from the time the collection is made to the time the
individual donation is identified, the confirmation testing is done, the
center is contacted, the donor's file is pulled.  They are identified.
Notice is sent out to them.

	By the time you add all that up, in addition to the inventory hold that
is in place, and all these other measures, by the time you add that up,
you are looking at quite a span of time, yes.

	DR. EPSTEIN:  I would like to ask Dr. Bianco a question.  If testing is
done on a minipool and you get a positive pool, will there be an effort
or will there not be an effort to notify hospitals that they may have
transfused a high-titer unit?

	DR. BIANCO:  That was not part of the program for the minipool,
stopping at an average of 20 units.

	DR. EPSTEIN:  So, you would have knowledge that out of a pool of, say,
16, or 16 to 24, however the case may be, there was a high-titer unit,
and the plan is not to tell the hospital?

	DR. BIANCO:  In that Phase 1, as we had planned, the intent was not to
notify the hospital or the donor, and these would be done after the
expiration of the cellular products, after 42 days of the collection.

	In Phase 2, that is what Sue presented, that is a pre-release testing,
and then it would be done like HIV or HCV.

	DR. EPSTEIN:  So, where does the scenario arise where there might be an
in-date unit?  It would not.

	DR. BIANCO:  In the minipool, in the way we proposed, it would not.  If
we resolve to the individual donor, then, the scenario that we are
discussing here certainly would apply, but that is not the intent.

	DR. EPSTEIN:  Part of the issue is that there has been a moving target.
 You know, we hear different plans at different times.  That is why the
agency is focused on the question of whether we should be proactive and
say that if whole blood donors are being screened, that we should be
pushing toward interdicting the high-titer units either in an interim
phase where it's product retrieval and lookback notification or
ultimately pre-screening and upfront interdiction.

	DR. BIANCO:  That is appropriate.  Let's say in this pool of average
20, there may be a frozen red cell.  Certainly, that frozen red cell
would be interdicted, but for all 20 units, not knowing which one of
them is the positive one.

	DR. EPSTEIN:  I think that what is being overlooked here is that when
you are transfusing units and you have knowledge that they may be at
high titer or that they were, that drives toward a situation of
lookback.  You know, you want to tell the doctor that you used a
high-titer unit, and it drives toward a scenario of product retrievals
from inventory.

	That is the phenomenon that is driving you to work back toward the
individual unit.  So, you end up there, you end up either doing upfront
screening as a release, in which case you identify individual units, so
that you don't have to throw out dozens of units, or you end up
identifying individual units because you are engaging in product
retrievals or lookbacks.

	DR. BIANCO:  But we will do that for all the 20 units in the minipool
regardless.  We will lose the product.

	DR. EPSTEIN:  You will lose?

	DR. BIANCO:  The 20 frozen red cells.

	DR. EPSTEIN:  I am sorry.  You would pitch 20--well, 20 frozen red
cells, yes.

	DR. BIANCO:  That is correct.

	DR. EPSTEIN:  But in the upfront screening scenario, if you use
minipools--

	DR. BIANCO:  Then, that is different.  If it is upfront, if it is for
release, it would be treated like NAT today for HIV or HCV with
resolution to the individual donor and all the actions taken.

	DR. EPSTEIN:  The whole idea of going from Phase 1 to Phase 2
implicitly strikes me as affirming Question 3.  Question 3 is whether
you should work toward identifying individual units.  Now, we are asking
if you do, should you also notify, but the commitment to go from Phase 1
to Phase 2 is a commitment to break down to individual units.  You are
already there.  The question then is should you notify.

	DR. BIANCO:  Oh, if we are in Phase 2, yes, I would be sitting there
and saying yes.

	DR. SIMON:  Then, Dr. Epstein, should we then divide this also between
the transfusible unit and the source plasma?

	DR. EPSTEIN:  Well, yes.  Again, I think Questions 3 and 4 were
intended to work together, and the answer for source plasma is really
that it's impractical under Question 4.

	DR. BIANCO:  Under the scenario of the minipool for the whole blood, as
Phase 1, would you include it under Question 4?

	DR. EPSTEIN:  Yes, I think if you in Phase 1 and you are in a scenario
where you have delayed identification, then, it becomes under Question
4, yes.  Again, the underlying issue is whether the goal here is to
screen units for transfusion.

	Now, Question 1(b) said we are not there yet, we shouldn't be taking
that position, and I am saying that if, in fact, you move to "real-time"
testing at any point, you are faced with the scenario of Question 3.

	DR. BIANCO:  But that I think is as I affirmed even to Dr. Klein, is
the scenario of all tests that we apply.  I personally and my
organization would have no objection.

	DR. EPSTEIN:  But it is not true, Celso.  In CMV, you do not notify a
donor.  The one-time ALT, you don't notify a donor.  With the one-time
anti-core, you don't notify a donor.  It is not automatic that we think
you should notify a donor.  It needs to be asked.

	I have only been pointing out that to argue that we never get there
because we only test pools is wrong thinking.  We will end up, at some
point, testing individual units at least for whole blood, and then the
question becomes material whether we think we should notify.

	Again, I would suggest that we do not always notify.

	DR. BIANCO:  I agree with you.  I think that we are not distinguishing
here clearly the minipool testing with no resolution of the minipool
versus the individual unit testing in any scenario for the whole blood
donor.  Even if you have a very delayed testing for a whole blood donor
to resolve to the individual unit, you certainly would come with the
ethical questions that Dr. Klein raised.

	DR. EPSTEIN:  I think we can disentangle this if we change it to B19
infected donors and just strike the word "identification," in other
words, if you have found an individual donor is the point here.

	DR. DiMICHELE:  Are we talking about the donor, though, or the contact?

	DR. EPSTEIN:  No, no, no.

	DR. DiMICHELE:  The way it is framed, it is about the contact, and not
about the donor.

	DR. EPSTEIN:  That is correct, but the issue is do the benefits to the
contact warrant notifying an individual positive donor.

	DR. NELSON:  Right.  So, you would notify the donor,

	DR. EPSTEIN:  What I am doing is I am removing the identification of
because that is the whole issue of breaking down a minipool.

	DR. NELSON:  Right, exactly.

	DR. EPSTEIN:  So, I am splitting the issue out.  If you find yourself
in the situation of identifying an individual positive donor, should you
notify based on potential benefit to contacts.

	DR. BIANCO:  If I were sitting there, I would vote yes.

	DR. NELSON:  Can we vote on that?  Let's vote.  Linda.

	DR. SMALLWOOD:  Question No. 3, as modified.  Do potential medical
benefits to contacts of parvovirus B19 infected donors warrant
notification of positive donors?

	Vote.  Allen.

	DR. ALLEN:  Yes.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  Yes.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  No.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  Yes.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  I am going to abstain because I think that
notification is a due process of medical ethics and when you have a
result, you need to notify the donor, and it is not because of the
medical benefits to contacts.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  No.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  Yes.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  Yes, under the assumption we will get to several weeks in
No. 4.

	DR. FALLAT:  Dr. Nelson, are we going to consider the question of
should we notify the recipient of a high titer B19?

	DR. NELSON:  That is a question we weren't asked, but theoretically, if
you identified a high-titer specimen, you wouldn't transfuse it.

	DR. SMALLWOOD:  Results of voting for Question 3. There were 8 yes
votes, 2 no votes, one abstention, and the industry representative
agreed with the yes vote.

	DR. NELSON:  No. 4.  I am ready to vote.

	DR. CHAMBERLAND:  Question 4, there is a lot of wiggle room.  It says
should donor notification be limited to settings where testing and
notification can be completed within several weeks of donation.

	What are people's view of what "several weeks" are?

	DR. SIMON:  I think the intention here would be to split--I hope I am
interpreting correctly--the transfusible product situation where they
are going to be doing this rather soon after donation and discriminating
down to minipool individual unit versus the source plasma situation
where it is going to be several weeks.

	It might be clearer to say, if yes to Question 3, should this exclude
the source plasma donation situation, or if that is how it is
interpreted, I would say yes to Question 4.  I know that several weeks
is kind of questionable, but I think that is the intention, to
discriminate between those two situations.

	I would hope we agree that in the source plasma situation with this
passage of time, that it would not be appropriate to notify.

	DR. DiMICHELE:  I think you could interpret that question in a
different way.  I mean the way you could also interpret it would be, you
know, if it is past the two-week period of viremia, is it going to make
any difference to the contact, if you notify them or you don't notify
them.

	I think based on some of the information that has been presented by Dr.
Brown, I guess in some circumstances, it might still benefit the
contact.  It becomes a tricky issue again the way it is stated and
depending on how you interpret it.

	DR. NELSON:  If the contact is an AIDS patient who is now on
erythropoietin, yes, it would make a difference.

	DR. CHAMBERLAND:  I think you are really stuck here because you can
always come back to that.  For some people, the rare person, there might
be a potential medical benefit.

	I totally share the concerns that have been raised about the
implementation of this and the communication of these messages is just
really extraordinarily difficult to think about, but I am no sure in all
honesty that you can say, or unless people have--I mean there are ways
to go about trying to model this and do all those sorts of things, these
medical decision analyses, and things like that.

	I don't know whether this is one of these situations where it is
potentially amenable where you can try and put a quantifiable handle on
it, although oftentimes in the setting of questions that relate to the
safety of the blood and plasma supply, people are somewhat averse to
reducing it to quantifiable estimates, but that is where I continue to
just kind of get stuck at.

	DR. SIMON:  I think we haven't, in this discussion, talked about down
sides of notification, and we are talking about a very rare benefit
here, the AIDS patient, the immunocompromised who might get IVIG, which
could still be considered experimental therapy, versus people are going
to have consternation for no reason, get a lot of medical testing and
evaluation for no reason, see the doctor, and accumulate bills they can
ill afford.

	So, there are significant down sides and when we are out several weeks
and the contacts have already been made, it seems to me we have such
elusive possible benefits that the down sides become--to me, they
outweigh the benefits.

	DR. NELSON:  I am not sure about the down sides.  A person could get a
hemoglobin and if it's okay, or a reticulocyte count, if it's okay,
then, the infection is over.

	DR. SIMON:  Those cost money.  Often people don't have money for that,
and there is medical-legal risks or people who don't think they have
been notified appropriately.  So, I mean I think there are down sides
and I think the benefit here is so elusive and so minimal that I
personally don't feel that, at this time level, that it is reasonable to
ask the industry to make a contact.

	DR. BIANCO:  I would like to suggest a solution. It is not just the
source plasma, Toby, it is also the minipool testing where we did not
resolve to the individual donor.

	What Dr. Epstein has suggested, remove identification from No. 3, I
would transfer identification to No. 4.  If yes to Question 3, if the
donor is identified within several weeks of donations, or should
notification be limited to settings where the donor has been identified
within several weeks of donation, because then we focus on the
individual that would be the object of that donation, can we notify the
donor within a certain reasonable period of time or we miss the boat, or
we did not resolve the minipool.

	DR. DiMICHELE:  It seems to me that Question 4 actually still refers to
the contacts, which is what we answered in Question 3, you know, whether
we should limit it to contacts is one issue, but I believe it refers to
Question 3.

	I just wanted to make one other statement, and that is, you know, when
we try to resolve this on medical-ethical issues, it becomes very
complicated, because the question involves expectation of the donor,
expectation of donor contacts.  It involves social good and ultimate
making decisions on the basis of good to society or making decisions on
the basis of good to individual patients.

	You know, do we develop a policy that protects the least among us or
the greater good.  I think this becomes a very, very complicated
question.  Certainly, the testing and notification policies that have
gone on heretofore have certainly focused on the individual and the
expectation of an individual and an individual donor, which is sort of a
very individualistic approach to this philosophy, but it is a tricky
question and I think we have to decide on which basis we are going to
answer that question.

	DR. FALLAT:  Could I get a clarification?  If you find something like
this, do the blood banks consider that they have to go directly to the
patient, and not through their physician?  If you go through the
physician, isn't that kind of helping resolve a lot of these ethical
issues?

	DR. SIMON:  No, you go to the donor.  The blood bank has a relationship
with the donor.  You have no idea who the physician is, and some
organizations have the center physician assume that role, but you are
definitely going to the donor.

	DR. SCHMIDT:  The question was about patient, not donor.

	DR. SIMON:  These are donors here.

	DR. NELSON:  The other big problem is this might be pretty frequent
from some of the data that was presented.

	DR. SIMON:  It is only the high titer.  I mean this is a hit or miss
thing, which is the other thing.  If somebody donates right before they
hit their high titer, and their brother has AIDS, they are not going to
be notified.

	DR. NELSON:  Right.

	DR. DiMICHELE:  Are we answering the question about the contacts,
though, or the donors?

	DR. NELSON:  You are not notifying the contacts.  It is the donor's
responsibility if his wife is pregnant or if his roommate has AIDS, or
something like that, in other words, you would educate him about what
this means.  The donor would almost always be healthy by the time you
got to this.

	DR. DiMICHELE:  But we are notifying the donor based on potential
medical benefit to the contact, even if it's beyond several weeks after
donation.  That's the question we are answering.

	DR. NELSON:  Right.  That is the issue.

	DR. KLEIN:  This says within several weeks.

	DR. CHAMBERLAND:  Right, and so in Toby's shorthand, I mean it includes
both the whole blood donors, as well as the source plasma donors.  That
is what the shorthand here is for.

	DR. SIMON:  You are answering no, right?  Yes would not include the
plasma donors as I interpret it, and no would.

	DR. NELSON:  Because of the word "limited to."

	DR. SMALLWOOD:  Question 4.  If yes to Question 3, should donor
notification be limited to settings where testing and notification can
be completed within several weeks of donation?

	Vote.  Allen.

	DR. ALLEN:  Yes.

	DR. SMALLWOOD:  Chamberland.

	DR. CHAMBERLAND:  No.

	DR. SMALLWOOD:  Davis.

	DR. DAVIS:  Abstain.

	DR. SMALLWOOD:  DiMichele.

	DR. DiMICHELE:  No, on the basis of a slightly different interpretation
of the question that Toby has sort of iterated.

	DR. SMALLWOOD:  Doppelt.

	DR. DOPPELT:  Yes.

	DR. SMALLWOOD:  Fitzpatrick.

	DR. FITZPATRICK:  No.

	DR. SMALLWOOD:  Klein.

	DR. KLEIN:  Yes.

	DR. SMALLWOOD:  Schmidt.

	DR. SCHMIDT:  Yes.

	DR. SMALLWOOD:  Fallat.

	DR. FALLAT:  Yes.

	DR. SMALLWOOD:  Harvath.

	DR. HARVATH:  Yes.

	DR. SMALLWOOD:  Nelson.

	DR. NELSON:  No.

	DR. SMALLWOOD:  Dr. Simon.

	DR. SIMON:  Yes.

	DR. SMALLWOOD:  Results of voting for Question No. 4.  Six yes votes, 4
no votes, 1 abstention, and the industry representative agreed with the
yes votes.

	DR. NELSON:  I guess that's it.

	[Whereupon, at 7:00 p.m., the meeting was adjourned.]

- - -

ajh

MILLER REPORTING COMPANY, INC.

735 8th Street, S.E.

Washington, D.C. 20003-2802

(202) 546-6666

