UNITED STATED OF AMERICA

P R O C E E D I N G S

(8:34 a.m.)

CHAIRPERSON OVERTURF: I'd like to call the meeting to order and turn it over first to Christine Walsh.

MS. WALSH: Good morning. I'm Christine Walsh, the Executive Secretary for today's meeting of the Vaccines and Related Biological Products Advisory Committee.

I would like to welcome all of you to the 101st meeting of this Advisory Committee. Today's session will consist of presentations that are open to the public. Tomorrow's meeting will consist of both open and closed sessions.

I would like to request that everyone please check your cell phones and pagers to make sure they are either in the off or silent mode.

I would like now to read into the public record the conflict of interest statement for today's meeting.

The following announcement is made part of the public record to preclude even the appearance of a conflict of interest at this meeting. Pursuant to the authority granted under the committee charter, the Director, Center for Biologics Evaluation and Research, has appointed for the discussions on February 16th the following participants as temporary voting members:

Dr. Robert Couch

Walter Dowdle

Theodore Eickhoff

Pamela McInnes

Arnold Monto

Stephen Phillips

Benjamin Schwartz

Melinda Wharton

For the discussions on February 17th, the following participants have been appointed as temporary voting members:

Drs. Pamela McInnes

Stephen Phillips

Benjamin Schwartz

Melinda Wharton

Based on the agenda for February 16th, it has been determined that the topic being discussed by the committee on the strain selection for influenza virus for the 2005-2006 season is a general matters issue. The committee will not be providing advice on specific firms or products on this day.

To determine if any conflicts of interest exist, the agency reviewed the agenda and all relevant financial interests reported by the meeting participants. The Food and Drug Administration prepared general matters waivers for participants who require a waiver under 18 USC 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 discussion before the committee, these potential conflicts are mitigated.

We would like to note for the record that the agency is in the process of selecting a nonvoting industry representative for this committee. On February 17th, the committee will hear updates on FDA's critical path initiative and will hear a presentation on the Laboratory of Biophysics and the Laboratory of Pediatrics and Respiratory Viral Diseases.

Meeting participants were not screened for potential conflicts of interest for these updates and overviews.

We would like to note for the record that Dr. Nancy Cox is serving as a consultant for this meeting, any speaker making a presentation. She is Chief, Influenza Branch, Center for Disease Control and Prevention in Atlanta, Georgia.

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

Ms. Linda Canas is Chief of Diagnostic Virology, Epidemiological Surveillance Division, U.S. Air Force, San Antonio, Texas.

Dr. Keiji Fukuda is Chief, Epidemiology Section, Influenza Branch, Center for Disease Control and Prevention, Atlanta, Georgia.

In addition, Mr. Albert Thomas is an industry speaker making a presentation. He has financial interests associated with his employer and regulated firms. He was not screened for these conflicts of interest.

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

With respect to all other meeting participants, we ask in the interest of fairness that you 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.

That ends the reading of the conflict of interest statement. Dr. Overturf, I turn the meeting over to you.

CHAIRPERSON OVERTURF: The entire first day will be contributed to the issue of strain selection for influenza virus for the forthcoming season, 2005 through 2006, and the first speaker is Dr. Roland Levandowski.

DR. LEVANDOWSKI: Thank you, Dr. Overturf.

Good morning, everybody. Welcome to Bethesda. Actually I see there's lots of room left at the front. So those of you who are sitting at the back are welcome to come up a little closer and the slides will be a little bit better, I think, for you.

I have been reminded by a friend that this is the Year of the Rooster, and if I can get this up here, this is the rooster family, and they're all smiling because they know it's time to get started making influenza vaccine again this year. So I don't really know who to attribute this picture to. It was sent to me by a friend who found it on the Internet, and if you like it, you may be able to find it, too. I couldn't, but I got the picture from the friend.

DR. COUCH: I thought you were going to say it's our new susceptible population image.

DR. LEVANDOWSKI: Well, it could be. This is the red jungle fowl here.

All right. Let me get down to business though. Okay.

As Dr. Overturf said, the reason that we're here today is for the committee to make recommendations for selection of the influenza virus strains, the A(H1N1) and A(H3N2) and B viruses that should be used for the influenza vaccines to be prepared for the 2005-2006 influenza season in the United States.

Why do we change the strains in the influenza vaccines? We do that because it's really important for vaccine efficacy. We know that vaccine efficacy relates to a couple of things, one of which is the potency of the influenza vaccines, but from a lot of experience, it has become very clear that the match of hemagglutinin and neuraminidase of the vaccine strains to the wild-type circulating viruses is important for vaccine efficacy.

And the first evidence of that for reduced vaccine efficacy was apparent two years after the first vaccines were licensed for use in 1945. Within two years, it became clear that antigenic draft in influenza viruses could reduce the vaccine effectiveness.

The questions that the committee needs to consider answering in order to make the recommendations are listed here and we'll be presenting information that covers all of these areas during the course of this meeting.

The first and most important questions is from Surveillance and Epidemiology: are there new influenza viruses that are circulating that have hemagglutinins and neuraminidases that appear to be different from the current vaccine?

And if the answer to that question is yes, we also want to know: are these new viruses spreading in people? Are they in wide geographic locations or are they just from one location?

Occasionally we see viruses that look extremely different, but it turns out that they're one off, and they don't seem to spread anywhere. So this question two, if it's answered yes, are those viruses spreading, we also want to know whether current vaccines can induce antibodies that will recognize those new viruses.

And if the answer to that is no, then we want to know further are there vaccine strain candidates available that would be suitable for large scale manufacturing of inactivated and live attenuated influenza vaccines.

I'd just like to go through a review of what the committee considered last year and what the questions and the sort of resolution to the questions was. First of all, for the current vaccine that we have now last year, the question was were there new strains of Influenza A(H1N1) circulating, and at that time you might remember we also had some reassortant viruses that were H1N2. Last year there really weren't strains that were antigenically different from the current vaccine strain. All of them were very much similar to what was in the vaccine.

The same question for the H3N2. Last year the answer to that question was yes. There were A Fujian-like viruses that we had known about since February of 2003. As you might recall, the first season that those were identified it was not possible to make a change in the vaccine, but those strains continued to circulate widely around the world in people.

And by 2004, although early on those were more or less in the minority or fairly quickly they became the majority and by 2004, those were the main strains that were circulating in the world.

For Influenza B, the question was asked: are there new strains present? And the answer was yes, and in 2004, the majority of the viruses were similar to a strain called B/Shanghai/361/2002, which is from the so-called B/Yamagata/1688 hemagglutinin lineage.

That lineage was not the one that was being used in the vaccine that was current last year. In a minority of the strains that were found during the epidemiologic studies were similar to the strain that was in the vaccine for last year, which was B/Hong Kong/330/2001, which belongs to the HA lineage that we represent with the strain B/Victoria/287.

In answer to the question were these new viruses spreading, the answer, of course, is definitely yes. The Fujian-like viruses had become widespread around the world and were predominant everywhere, and these B/Shanghai-like strains at the time we were holding this meeting in February were predominant not only in North America and the United States, but also in Asia and Europe.

Were the new viruses that were identified and spreading, were those inhibited by the current vaccines? And this question, as it sometimes is, was not a very definite no or yes. It was a little bit difficult to interpret, but it seemed like man of the A/Fujian-like viruses were not well inhibited by the current vaccines, although some of them were.

For the B/Shanghai-like strains, of course, we've known for a long time that these two divergent hemagglutinin lineages are not that well inhibited one by the other, and as time has gone on and antigenic drift has occurred in these strains, that has become truer.

Generally we also know that for the B/Yamagata-like strains and the B/Victoria-like strains, that very young children and people who haven't been immunologically primed, exposure to one of these does not seem to immediately give antibodies that cross-react with the other HA lineage.

So were there strains that were suitable for manufacturing? And the answer was yes. Of course, we all know that for inactivated vaccines and for live attenuated vaccines manufacturing depends on having egg adapted strains, either the wild-type or reassortant, and in the case of the live vaccine, of course, it has to be a reassortant for the attenuation phenotype.

But there were A/Fujian-like strains that were available, and there was a high growth reassortant that was being used in manufacturing for the Southern Hemisphere already, the A/Wyoming/3/2003 X 147 reassortant.

For the B strain, there were a number of wild-type isolates that seemed to be suitable for manufacturing, including B/Jilin/20/2003 and B/Jiangsu/10/2003, in addition to the B/Shanghai/361 strain itself.

So based on that, the strains that were selected for this year include A/New Caledonia/20/99-like strain, which in this case really is A/New Caledonia/20/99.

For the B/Shanghai/361/2002-like recommendation that was made, there were all three of these strains, B/Shanghai, B/Jilin, and B/Jiangsu.

And for the A/Fujian/411/2002-like recommendation that was made and the A/Wyoming/3/2003 strain was chosen or is the one that has become widely used for vaccine preparation.

Now, the implications of the strain selection were that preparation of the vaccines was on schedule throughout the year. All of the strains seemed to be typical and easy to adapt for manufacturing purposes, and going into the summer, the supply of vaccine was expected to match the demand predicted by previous years' experiences.

But what happened was that we ended up with a vaccine shortage at the end of the summer, and just to try to put that into a little perspective, from January until August, manufacturing had been progressing on schedule even including these two new strains that were recommended for use in vaccines, and it was anticipated there were going to be about 100 million doses of vaccine from all of the manufacturers combined for this year.

In August of 2004, Chiron notified regulatory authorities about a sterility issue and indicated that investigation to identify the cause and the implementation of corrections was underway, and at that time Chiron made a public announcement indicating that there would be a possible delay in distribution and possibly a reduction in the amount of vaccine that would be available.

You also probably all know that in early October of 2004, the MHRA, the U.K. regulatory authority, announced that they were suspending Chiron's license to manufacture inactivated influenza vaccine for three months, and that was based on the issues that have previously been identified and were in investigation and correction by Chiron.

Subsequently, over the next few weeks and certainly by November of 2004, it became clear after consultation between FDA and MHRA that the vaccine that Chiron had planned to make was not going to be available for us in the United States.

In response to that, there were a number of things that happened within the Public Health Service, and I'll just very briefly indicate some of those. At FDA there was a lot of work done to evaluate manufacturers who were not licensed in the United States to identify whether their vaccines could be used under IND.

There was consultation with manufacturers to discuss regulatory mechanisms going forward from this time for getting approval of new products in the United States. That includes accelerated approval, fast track and priority reviews to facilitate those new licenses, and all of these things actually have been continuing.

CDC had a number of roles to play, and I'm not indicating everything here, but certainly there are some very prominent roles in the public health response to what was happening with loss of some of the vaccine that was anticipated.

CDC immediately reviewed and communicated the use recommendations that would be appropriate for this reduced amount of vaccine that was anticipated. It worked very diligently in terms of coordinating distribution of the existing vaccine supplies and were very closely linked and working with manufacturers and FDA in terms of acquisition and use of vaccines under IND in the United States.

National Institutes of Health also as part of the Public Health Service responded to this and were able to provide support for a number of clinical studies that might be done for vaccines under IND-made commitments to help manufacturers and their interests in doing clinical studies that would be useful for IND and possibly later on for vaccine license approvals.

And of course, they've been giving continuing support for development of new vaccines. This is something that was ongoing already at NIH, but have continued to try to facilitate development of new vaccines, tissue culture vaccines, recombinant DNA, and also adjuvanted influenza vaccines.

And the Public Health Service and HHS generally underwent a global consultation with other partners, including national regulatory authorities, but certainly also with manufacturers to try to find where there might be additional vaccine supplies and to acquire those for use here.

Currently there still is vaccine that's available, and I guess the most recent recommendations that I have seen on use of vaccine for late season use are on the CDC Website, and I would anticipate that those are continuing to be updated. So anyone who is interested in them should certainly check -- interested in vaccine supply and maybe how to obtain additional vaccine -- should continue to check the CDC Website and see what the latest information is.

There still, as of the time I was putting these slides together, was inactivated and live attenuated vaccine available for use, and I think we would all like to make sure that the vaccine that has been produced is used effectively.

And finally, there are also IND vaccines that are available during this year from both GSK and Berna Biotech.

Now, switching gears a little bit and thinking about where we are and where we're going from here for this year. We're in February so there are a lot of things that are happening. So there are a lot of things that are happening. Obviously I've put this together as a kind of a pyramid, and the most visible part of the influenza vaccine, of course, is the vaccine use that occurs in the fall months and into the winter, but you'll see that there's a lot of activities that have to go on before that can happen, including preparation of the vaccine shown in blue here and all of those bars, and all of the support activities that are required for the manufacturers to know what strains that they're going to be using, acquire the reagents and the materials that they need to permit them to go ahead with manufacturing.

So we're here in February. I don't have a pointer. I don't know if I can get the arrow to show up here on the -- I don't see one, but we're here in February, and we're right at the point of recommendations being made both by this committee and by other national health authorities.

Obviously surveillance continues throughout the year. Development of seed viruses and reagents and reference materials, that goes on throughout the year as things become apparent and become available, and it's based on all of those underlying activities that the preparation of the vaccine can start.

Now, I showed this slide last year, but I wanted to show it again to talk about what happens when there are new viruses that are added to the vaccine. It's quite a challenge for manufacturers to get everything together in the relatively short period of time that they have, and even a simple strain change, we're talking about work that requires many, many weeks, somewhere in the order of between 12 and 20 weeks in order to accomplish all of the tasks that need to be done.

In order to have any change in the vaccine, of course, there have to be reference viruses that are obtained. Those come from surveillance, and it is not always easy to get those. As you might recall, it was difficult to isolate some of the H3 viruses in eggs, and as you're going to hear, CDC has been putting in a lot of effort into making sure that egg isolates, appropriate egg isolates, are available for manufacturing. I think Nancy Cox will talk about that a little bit later.

Once those reference viruses have been acquired, then that's not the end of the job because each of the manufacturers has to take the reference virus and develop a working seed virus from that. This is not something that is done in a day. It takes several weeks' worth of work to identify the strain that seems to be appropriate for the manufacturing process and also to make sure that all of the quality control issues that need to be handled and addressed for those new viruses have been done.

Thank you.

So once that's accomplished there still need to be reference reagents that are produced for the reference virus as well, and that can be a rate limiting step. Once the virus has been identified or it's in our hands, then we can start to work on getting those reagents made, but it takes a period of about three months actually, well, six weeks to three months to have everything prepared because it requires both an antigen and an antibody that's made in sheep, which is a biological system that's not always readily controllable.

Generally though manufacturers are already for the strains that aren't changed can start working and keep working on preparation of materials for vaccine, and when it's time, they can start manufacturing the third strain so that they can formulate the trivalent vaccine, fill it, and then distribute that vaccine hopefully in time for use in the fall.

There will be a presentation later by our industry representative who will go into more detail about this, but I wanted to mention it also that it's a fairly complex set of activities that need to be undertaken for implementation of any strain change, and it is a kind of a stressful situation for all parties that are involved to first make sure that their reference virus is present, make sure that the reagents get produced and make sure that the manufacturing can go forward.

I'll end up here with some bits of information from WHO recommendations during the past year. These are the recommendations that WHO made for the Southern Hemisphere, and you can see that the strains that were selected for the Southern Hemisphere, which were based on information much like what we'll be discussing here this morning, it was to keep the A/New Caledonia/20/99 strain as the vaccine virus and it actually was to include now a B/Shanghai/361/2002-like virus for the first time in the Southern Hemisphere, but it was a continuation of what had already happened with recommendations in the Northern Hemisphere in February last year.

However, there was a recommendation for a change in the A/Wyoming strain to an A/Wellington/1/2004-like virus, and that was based on the fact that the H3N2 viruses were undergoing antigenic drift.

And finally, these were the recommendations that have been published on the WHO Website for this coming year, and you'll see that although the H1N1 and the B strain are the same as what they were for the Southern Hemisphere, the WHO recommendation for the H3N2 virus is for another different strain, the California/7/2004-like virus. And we'll be presenting information that will probably make that understandable.

So the question for the committee that we would like to have addressed, the specific question that we're asking the committee this morning is what strains should be recommended for the antigen at composition of 2005-2006 influenza virus vaccine, and this recommendation should be based on the epidemiology and antigenic characteristics of the circulating influenza viruses.

On serologic responses, people have been immunized with current vaccines and the availability of candidate strains, and I guess I can stop there and see if there are any questions or comments.

CHAIRPERSON OVERTURF: Other questions or comments for Dr. Levandowski? Dr. Couch.

DR. COUCH: Haven't the WHO recommendations for the Northern Hemisphere for 2005-2006 already been made? Haven't they all? They were on the Website at any rate.

DR. LEVANDOWSKI: That's right. That's what I showed.

DR. COUCH: But you didn't include them. Won't you tell us what those are?

DR. LEVANDOWSKI: Right there. Isn't that it? Do I have it mislabeled?

DR. COUCH: No, that's fine.

CHAIRPERSON OVERTURF: Dr. Markovitz.

DR. MARKOVITZ: Yeah, thank you.

I wanted to ask. You showed something about the Berna Biotech vaccine in IND, and we're going to hear about the GSK tomorrow, but what is the Berna Biotech vaccine?

DR. LEVANDOWSKI: It's an inactivated influenza vaccine.

DR. COUCH: Aren't there other candidates for the IND or there was or what's the status of other manufacturing candidates for interest in our country? I guess that's the question.

DR. LEVANDOWSKI: Okay. Well, of course, I'm not going to talk about any INDs that are, you know, confidential information, but there is a lot of interest; there has been ongoing interest. There was interest even before this from a number of manufacturers to bring their products to the U.S. market, and those are all things that were in the works and are continuing.

So I guess what I can say generally is that there are a number of manufacturers that are interested who are pursuing avenues toward getting approval for their products in the United States, and they're multiple. It's not just one or two. It's multiple.

CHAIRPERSON OVERTURF: Any other questions?

(No response.)

CHAIRPERSON OVERTURF: We'll go on then. Dr. Fukuda is going to give us the U.S. surveillance data.

DR. FUKUDA: Good morning. I see that I'm allotted more time than I really need. I'm only going to spend a few minutes talking about surveillance in the United States, and then Dr. Cox will be covering some of the events going on in Asia related to H5N1.

So normally I just talk about the activity that's going on in the United States, but I thought I'd take a minute or two and go over some of the changes affecting how we're doing surveillance in the U.S. because they really have been quite substantial over the past year or two, and I think it's changes that the committee should know about.

So over in the left-hand column you can see how we've done surveillance in the United States for several years, and for quite a long time we've monitored viral activity through the WHO nerve laboratory system in the U.S., which is largely a group of Public Health laboratories, plus university laboratories. We've monitored influenza-like illness visits to a group of sentinel physicians scattered throughout the United States.

We've monitored mortality from influenza using two different systems. The 122 cities systems collects data from vital registrars' offices in 122 cities, and then the NCHS data set is the large data set reflecting all deaths in the United States which are analyzed a couple of years afterwards.

And then we typically get state activity assessments from the state epidemiologists every week, and so this is basically the information that you've seen for year in and year out.

Now, there are a couple of things which are really driving changes in surveillance. One of them is that since really the mid-1990s we've been trying to strengthen surveillance as much as possible, recognizing that there are a number of limitations.

A second thing is that there has been a great deal of concern about pandemic influenza, and I think this concern continues to rise, and so that's another driving factor for enhancing surveillance in the U.S.

And then the third thing is that there has been directives from the Director's Office at CDC really to strengthen surveillance so that the data comes in a little bit more quickly and so that it's more broadly representative of the country geographically.

So based on that, there have been a number of things done to enhance the systems on the left, and I'll go over one example, which is the sentinel provider system. There have been a number of new systems which have been added.

We are now monitoring pediatric influenza related hospitalizations through two different networks, the NVSN and the EIP Programs, and this largely comes out of last year's experience where we had so many pediatric deaths and so many severe illnesses reported in that age group.

And then also as part of that, we have begun -- we worked with CSTE, the Council of State and Territorial Epidemiologists, to institute national pediatric death reporting for death related to influenza.

So these are new systems, and then because of concern of H4N1 and the initial cases that were reported back in 2003, we have been, in essence, at a state of heightened alert in the United States where state health department and hospitals have been on the lookout for H5N1 cases among travelers returning from Asia. So these are new systems which have been added over the past year or two.

Then in addition, there are a large number of systems, and I'm just giving two examples here which are under evaluation. Biosense is a large conglomeration of data sets which are collected from groups such as the Veterans Administration, DOD, pharmaceutical industry, and so on. These are being evaluated for the potential value when conducting influenza surveillance.

And then we also will be talking with the Council of State and Territorial Epidemiologists later this year about whether influenza laboratory confirmed illnesses should be a national reportable disease, which would probably profoundly affect how we do surveillance. It would probably be the biggest change of all if we go in that direction.

So this is the effect of some of those changes. You can see on this slide here that this represents the numbers of sentinel physicians in the country and then the number of visits that are made to those physicians for influenza-like illness.

So back in 19 96, which is the column over on the left, you can see that we probably had somewhere between 50 and 100 sentinel physicians reporting these data, and then if you go up to 2004, we have over 1,000 physicians reporting on a regular basis, and so this represents an increase of less than a million patient visits to about five to nine million patient visits per year.

Now, these physicians are scattered in the United States and these dots represent where they're located, and you can see that, in general, they're distributed in the way that the population of the U.S. is distributed, and typically when we've analyzed these data, we've shown you curves like this, which shows you the percentage, the cumulative percentage of visits for influenza-like illness on a week-by-week basis, and you can see in different years that percentage increases as we go into the influenza season, then comes down.

What we're trying to do, we're testing a couple of other ways of analyzing these data, however. This map here represents the application of the so-called outbreak detection algorithms to each physician, and so, in essence, each physician is treated as a sentinel for detecting outbreaks or increases in activity.

And so using certain statistical methods, what we do is look for an increase in visits for influenza-like illness for each of the physicians, and you come up with maps like this, this sort of speckled map where the red represents increases and the black do not.

I think right now what we're mostly struggling with is whether this kind of analysis adds anything substantial to what we already have, but anyway, so that work goes on.

So let me go into the current season now. So you can see, and this updates the report to the committee. I think the committee has surveillance data up through week four, and this goes into week five. So these numbers will be a little bit different from what you have.

So basically you can see that this year it has been a mixed Influenza A and B season, but predominantly A. To this point there have been about 65,000 specimens tested by the laboratory system and about 11 percent of those have been positive for influenza.

And of those which have been positive, about 85 percent of the isolates have been Influenza A and about 15 percent have been Influenza B.

Now, of the Influenza A viruses, about a third have been subtyped, and you can see of those that have been subtyped almost all of them have been Influenza A(H3N2) viruses, with a few H1N1 viruses or a few H1 viruses.

And so Dr. Cox will be going much further into these data in a few minutes, and so this graph here represents the same data shown somewhat differently, and so these stacked bars represent Influenza A viruses and B viruses, and the numbers that have been identified as we've gone into the influenza season.

Now, one of the important things to see here is this black line. This represents the percentage of specimens which are positive for influenza, and in many ways it's often the earliest indicator of how the season is going.

So right now you can see that approximately 23 or 24 percent of specimens coming into the system are positive for influenza viruses.

Now, if you look in the past of the past several seasons, you will see that this percentage typically peaks somewhere between a quarter and a third of specimens testing positive for influenza viruses when we reach the peak of the season. So right now this curve looks like we've reached the peak, but this was probably somewhat of an artifact, or we're not sure if it's an artifact right now. It may represent a bit of a reporting lag.

So based on this curve right here, it still looks like that we're going up in the season and we haven't quite peaked yet.

Now, this slide here represents the visits for influenza-like illnesses to sentinel providers, which I showed you a few minutes ago. The red line represents the pattern for this year, and the green line represents the pattern that we saw last year when we had that early season.

Now, these are curves which you wouldn't have seen last year. So these are laboratory confirmed hospitalizations, and these are hospitalizations coming into the NVSN system, and so these represent hospitalizations of children zero to four years of age, and you can see the blue line represents last year when we were hearing about so many reports, and the red line represents what we're seeing this year.

Now, these are similar data coming into the EIP system. I won't go into the details. The NVSN system and the EIP systems identify hospitalizations in somewhat different ways, and so the absolute rates are somewhat different, but what they show, in essence, is fairly comparable.

And I want to point out one thing though. The blue lines here, again, represent what we saw last year. The solid blue line represents hospitalizations that we saw in children zero to four years of age, and then the dotted or the broken line, blue line, represents the hospitalizations that we saw in children five to 17 years of age.

And so you can see that there was a quite large difference in rates of hospitalizations depending on age.

Again, the red line represents the hospitalizations that we're detecting this year. So, again, you can see there's a substantial difference between the experience this year and last year.

This curve here is the familiar pneumonia and influenza mortality curve which comes out of the 122 cities system, and so far this year we have not detected an increase in excess mortality.

I was looking at this slide this morning. So I'm struck that we have red states and blue states, and there are more red states than blue states.

(Laughter.)

DR. FUKUDA: But all things change. So anyway, these are activity levels represented by the state and territorial epidemiologists. The red, in essence, report reveals or indicates the highest level of activity in a state, and then the blue levels represent a somewhat lower level of activity.

So this is the last slide here. I think that what we can say is that in comparison with last year, this season has been relatively moderate. It has been dominated by Influenza A(H3N2) viruses. I didn't go over these data, but so far there have been six reported pediatric deaths associated with laboratory confirmed influenza. This is in contrast to 153 laboratory confirmed deaths reported last year for the entire year.

But then, again, this season clearly has started later than last year. We are in a period of ongoing activity. We cannot say that activity has peaked in the country yet, and so we still don't know what the full impact and what the full numbers will be.

So I'll stop there.

CHAIRPERSON OVERTURF: Just one question. You had that slide that looked like four curves or four seasons with pediatric hospitalization rates.

DR. FUKUDA: Yes.

CHAIRPERSON OVERTURF: What is the quality of the data for the two prior seasons?

Obviously we really didn't have a surveillance system that was looking at pediatric hospitalizations that I know of prior to last season; is that correct?

DR. FUKUDA: Well, this system, the NVSN system, has been in place since 2000, and so this actually represents now five years' worth of data.

I would say that the quality of these data are excellent. You know, this is a system which was set up by the National Immunization Program in Rochester and in Tennessee, Rochester, New York, and then in the Vanderbilt area, and then more recently they've added a third site.

And in essence, it's an active system where all children coming in meeting a certain case definition are then tested for influenza and other viral respiratory illnesses, and so it's a pretty labor intensive system, but the data themselves are quite excellent.

CHAIRPERSON OVERTURF: Other questions? Dr. Couch.

DR. COUCH: I wanted you to go ahead if you would, Keiji, and contrast those two systems a little bit because they don't exactly say the same thing.

DR. FUKUDA: Sure. I think one way to look at the NVSN system is that it's close to an ideal way of trying to look at what children are getting sick with and to identify rates of hospitalizations associated with various pathogens.

By contrast, the major limitation I would say of the NVSN system is that it's restricted to a small number of sites and it's expensive. The EIP system is a program intended to look at a wide variety of issues, and so the ABC system looking at bacterial infections comes out of that system. FoodNet comes out of that system, and this is a population-based surveillance system in 11 sites in the U.S. right now.

And so what this system does is take existing data. It takes how physicians handle children or other people coming into hospitals and looks at the virus detections and so on as they're currently done, and then takes that information and makes it available and makes it available in a way which is population based.

And so I think the strength of this system here is that it's a much larger system, and like NVSN, it represents population based data, and it reflects practice as it's actually done right now.

So it's probably comparatively less labor intensive, but I think the sensitivity of this system -- there will be an article coming out on this -- is less than the NVSN, yeah. So in a certain sense they are pretty complementary. They try to do different things, but in fact, you can see that the overall picture of the data is pretty similar. The absolute numbers are different, but I think that you both get a good sense of the rates going on, and certainly I think that these systems are going to be very helpful for looking at differences in seasons, particularly in children right now, you know, and this has been a major question. You know, how much does it vary in children and what is the impact?

CHAIRPERSON OVERTURF: Mr. Phillips.

COL PHILLIPS: Keiji, it was mentioned last week at ACIP, but I can't recall. Can you comment on the percentage or the numbers of children six to 24 months that received immunization this year compared to last year?

DR. FUKUDA: Yes. I think that actually Melinda or Ben may remember better, but I think that when we first started out, you know, the rates in that age group were very low, I think, less than five percent, and then within a year it went up to about 45 percent somewhere; is that right, Melinda? I'm not quite sure.

DR. WHARTON: I think the most recent data for six to 23 month olds from the BRFS for this year was 57 percent.

DR. FUKUDA: Oh, 57? Okay. Sorry.

So it has really been an astounding increase in that age group.

CHAIRPERSON OVERTURF: Yes, Ted.

DR. EICKHOFF: Keiji, two things. In the very first bar graph you showed about the sentinel physician providers I missed something. A straight line had turned straight down. What was that straight line? Is that the number of sentinel physicians?

DR. FUKUDA: This is the number of patient visits. So if you take the sort of cumulative number, and I think that the downward turn just reflects that we're still going on through the season. So I think that at the end of the season that line will be going up.

So I think that the downward line is just an artifact of where we are in the season.

DR. EICKHOFF: Okay. Thank you.

The second part of the question related to the state epidemiologist reporting. Now, I know there are definitions that go with each of these categories of reporting, like regional and widespread and so forth and so on, but yet I can't escape the feeling that there may, in fact, be a great deal of observer variability in these reports.

So you have any sense of how variable these may be within a specific definition, such as regional?

DR. FUKUDA: I think there's probably a substantial amount of variability. I mean, clearly, how each state decides to report their activity varies. I mean some states look more at their laboratory data. Other states look at perhaps what they're hearing about hospitalizations and so on, and in a sense it represents a gestalt from that state.

Nonetheless, I think it's funny, but I think that it actually pretty well represents what we see in the other parts of the system where, when we look at increases in visits to sentinel physicians, for example, in the northeast or in the southwest, and it correlates pretty good.

And what it does, you know, we really are not at a point yet with the other systems where we can break the data down to a state-by-state level and feel that they're robust enough that we can report on a state-by-state level. I think we're getting to a state where we're feeling pretty good that in a lot of the regions the data from the other systems are pretty good for those regions.

But as we get into smaller and smaller cuts at the data, you know, it becomes a little bit more -- the confidence intervals become a little bit too wide. So this really represents our way of trying to get at what are the states themselves feeling like they're seeing and how, you know, are they responding to that and reacting?

DR. FUKUDA: There's another question. Yes.

DR. DOWDLE: First I'd like to congratulate CDC for continued expansion of the surveillance system. It's quite gratifying to see that, and I'm also really interested in your discussions, upcoming discussions in making influenza a national notifiable disease, which brings up the question: in your discussion with the states, what do you see is going to be the major challenges to get this done?

I mean, this has been discussed before, but there are many challenges to doing that. So have there been any changes? Is there different attitudes? And what do you think are the real challenges this time?

DR. FUKUDA: Walt, I think it represents a couple of things that are changing out there. One is that, in fact, it turns out that there are more states in which laboratory confirmed influenza already is a reportable disease within the state than we really, I think, suspected.

I mean, when you look at it, it's probably around 20 states or so. So that's one difference than some years ago.

A second issue is that I think that influenza has gotten so much attention over the past couple of years that on the political agenda in a lot of states there is now a recognition that they really want to keep on track of what's going on with influenza in their states much more, and that's probably a big change over, say, five or ten years ago.

And then the third thing is that, you know, the State and Territorial Epidemiologists really pull together with CDC when there's kind of a crunch going on, and I think that there has been a big push over the past few years with the vaccine supply disruptions, the push from the Director's Office at CDC really to strengthen surveillance in a way that data is coming in a little bit more quickly. It's a little bit more specific, and perhaps eventually can be used really to respond to emergencies a little bit more quickly.

And so based on that, this idea of moving to notifiable diseases really came out from the states. It's something, I think, that we would have hesitated to approach because of all of the practical and feasibility issues, but basically a couple of the states came to us and said, "You know, we really think it's time that influenza surveillance begin to be treated like other diseases in the U.S. and that we begin to look for confirmed cases and try to track those, and so I think moving into that direction is really the biggest hurdle is going to be to get all of the states which aren't doing that right now to agree that it should be a notifiable disease.

I think if that is done then I think all of the other issues are relatively simple to deal with, and I think for the states it's really just a feasibility issue. You know, they're dealing with bioterrorism activities and so on, and there are so many things going on that everyone is trying to respond to, but that's the real hurdle.

CHAIRPERSON OVERTURF: Dr. Monto.

DR. MONTO: Coming from a state where we've been discussing making influenza notifiable disease and in a situation where there is concern that there would be significant under reporting or different reporting from different states using different criteria, is there discussion about standard methods that would be used in different states to get away from the situation which we have right now where the state epidemiologists basically makes a seat-of-the-pants decision about the level of influenza activity?

DR. FUKUDA: Well, Arnold, I think that if we do move to a situation where laboratory confirmed influenza becomes a reportable disease throughout the country, then the first issue is going to be laboratory confirmed what, and it will probably focus on something like hospitalizations because it's relatively restricted in numbers.

So we haven't entered into the discussions with the states about the nitty-gritty of how this might be done because I think first there has to be discussion about whether the other states think that this is how they want to go.

But I suspect that if we get past those discussions, it will focus on how do we start off relatively narrow and then do we expand out later on?

CHAIRPERSON OVERTURF: Dr. Farley.

DR. FARLEY: Well, this is in many ways a follow-up to that question in that the issue of testing, the type of testing, rapid testing versus virologic testing, and its sensitivity/specificity issues, but also the word is that some insurance companies cover the test and some don't, and I guess if they're hospitalized things may be much different in terms of coverage for testing, but where do the policy makers fit in that equation of it we're going to a laboratory diagnosed surveillance system that's reportable, will there be recommendations on whom to test and whether it should be covered?

DR. FUKUDA: Yes, Monica, that's a big issue. If you look at everyone who is tested for influenza right now, it's clear that a majority of them are now being tested using the rapid detection kits, and we all know that the sensitivity and the specificity of those kits is not at a level where any individual test result, particularly in the off season or when you have odd results is, you know, so solid, but you know, it's also the increased usage of that kind of testing which has made the whole discussion about moving to laboratory confirmed influenza possible.

You know, without that kind of testing we wouldn't be having this discussion with the states, and I think that some of the things that we'll have to come to grips with and which I believe will probably change over the next several years is that there are undoubtedly going to be regional differences, individual physician differences in terms of how often and how they're willing to use those tests, and that will change.

And so I think that all of those will be somewhat problematic. Nonetheless, I think that they're all addressable issues, and I think that when we look at the data coming in as a large lump of data, it will be pretty analyzable. I mean, that's what I suspect.

CHAIRPERSON OVERTURF: Other questions?

(No response.)

CHAIRPERSON OVERTURF: If not, Dr. Cox, are you ready to present now?

Okay. Dr. Cox will present.

DR. COX: Okay. Good morning, everyone. I'm very pleased to be here presenting the virologic data once again and shifting from a domestic perspective to a global perspective.

I thought I'd spend just a few minutes at the beginning of my presentation talking about the H5N1 situation in Asia, and then we will be able to focus exclusively on the task at hand, which is vaccine strain selection for this coming year.

This slide actually shows the countries that have reported H5N1 outbreaks in poultry since December 2003. The countries shown in light purple are the two countries, Japan and South Korea, that have their outbreaks under control and the H5N1 virus as far as we know has been eliminated from their borders.

The countries shown in purple have outbreaks in poultry, but no human cases have been report, and the three countries shown in red have outbreaks in birds and at least one case in humans. The countries in cases with humans are Cambodia, which has reported a case recently; Thailand, which has reported 17 cases of which 12 have resulted in death; Vietnam, which has had a lot of activity recently and a lot of publicity recently. There are 37 H5N1 laboratory confirmed cases with 29 deaths, for a total of officially reported cases of 55 and 42 deaths.

Now, of course, the case fatality rates are very high. We know that case ascertainment is not perfect. It's far from perfect, in fact, and so this probably represents an over estimate of the case fatality rate. Nevertheless, it's a very sobering picture.

We've noted that there's a high case fatality rate regardless of age, although the illnesses, the detected cases have tended to be in children and young adults.

The clinical symptoms are similar to the earlier cases in 1997, and lymphopenia is a prominent feature.

Diarrhea has been reported as being a prominent feature in some of the recent cases.

There was a second wave of infections that began in August of 2004, sort of tailed off a bit, and then increased again during late January and February.

And then, of course, I'm sure many of you have heard in the press and perhaps have even read the paper in the New England Journal about the Thai family cluster, where because of particular circumstances it was possible to document probable human-to-human transmission from a child to her mother and to her aunt.

Now, we have been looking very carefully at the antigenic properties of these viruses, and I probably showed you this slide before, but this slide shows that the viruses have actually drafted antigenically quite dramatically from 2003, where we had the Hong Kong/213/2003 virus, which had been used in the -- the wild-type virus had been used in Rob Webster's lab to produce a vaccine reference strain, and we had hoped that it would be appropriate to use for pilot lots for the situation that was developing in Asia.

Unfortunately, you can see that the ferret antisera against the Hong Kong/213 virus has a very nice homologous titer. However, that antiserum covers the Vietnam/2004 and Thailand/2004 viruses very poorly.

Likewise, we were able to see distinct differences between the Hong Kong/97 (H5N1) viruses here. We see the homologous antiserum titer, and this antiserum covers the 2004 viruses very poorly.

These viruses themselves do induce a good antibody response in ferrets that have been infected intranasally, but by using these sera we can also see distinct differences between 2003 and the 1997 H5 viruses.

So it became very clear that new vaccine candidates needed to be developed for the ongoing situation. Candidates have been developed and perhaps Pamela or someone else from NIH will give an update on the current pilot production situation for candidates that have been developed with these two particular strains if there are questions.

I'd just like to give a very brief summary of the highlights of points I'd like to get across to the committee. Obviously avian influenza viruses including all of these subtypes, but particularly the H5N1 viruses, can pose a major risk to global public health. Early detection of human to human transmission of novel influenza viruses is essential. It's difficult in Asia, but surveillance has been ramped up, and if there are questions about what the U.S. has done to help improve surveillance in Asia, I would be happy to field those questions.

The 2003 through 2005 Asian viruses are heterogeneous both in their antigenic properties and in their resistance to influenza antivirals. They are also heterogeneous with respect to pathogenicity, and the current strains are more lethal in mammals by the current measurements we have than the 1997 strains, and I'm talking about animal models here.

There is ongoing vaccine development, ongoing antiviral stockpiling, and pandemic preparedness activities at many levels within our own country, within other countries, and within WHO.

Surveillance in animals, including birds, swine, and other susceptible hosts, such as felines is critical, and a research agenda needs to focus on enhancing our understanding of the genesis of pandemic influenza viruses.

We have a unique opportunity to view how new pandemic strains may or may not develop, and there is now a very broad global recognition of a need for better communication between human and veterinary health authorities, and we are all working very diligently on improving communications.

So now I will be shifting gears and talking about the current influenza season, which is as Keiji has just shown you increasing and really getting going.

I'd like to provide a bit of an overview by way of introduction to the global situation. Now, we have compiled information from all of the four WHO collaborating centers, which are located in London, Atlanta, Melbourne, and Tokyo.

And we have also included information from the European influenza surveillance scheme, which is a very comprehensive influenza surveillance system in Europe, and have also included the very good data from our Canadian counterparts who report on a weekly basis their analysis of influenza viruses that have been isolated in Canada.

Generally speaking, between October 2004 and January of 2005 through the current time, influenza activity has been reported in Africa, the Americas, Asia, Europe, and Oceania. In general, influenza activity has been relatively low compared to the same period last year globally as well as nationally.

The influenza season began in October in North America where viruses were first detected, and it has increased quite gradually in countries in the Northern Hemisphere, including countries in Europe and Asia.

As you can see here, Influenza H3N2 viruses have predominated worldwide and were responsible for the majority of outbreaks. Influenza B viruses from both the Yamagata and Victoria lineages have continued to circulate globally and have been responsible for a few outbreaks.

Influenza A(H1N1) viruses have been detected less frequently and have been reported to be responsible for only one outbreak so far.

I would like to note that of the Influenza B viruses, those on the Yamagata lineage, which is represented in our current vaccine, have predominated. If you add up the 303 and the 74 and look at the proportion of B Victoria viruses on a global basis, it's about 20 percent in the United States. You can see clearly here it's also roughly 20 percent, in the same ballpark at least.

We have relatively few viruses from Central-South America because they haven't had much Influenza B activity, and you can see that in Africa and Oceania B Yamagata lineage viruses really did heavily predominate.

So now we'll move on to Influenza A(H1N1) viruses. Now, I would encourage anyone who really wants to see these numbers to move forward because I know it's very difficult to see the HI tables on the screen. You do have copies, but they're not color copies. So it's sometimes a bit harder to see.

I mentioned that we had relatively few H1 isolates, but we do have some relatively recent strains, some December strains from Florida represented here as test antigens.

We also have quite a number of Asian isolates. There was a fairly large outbreak of H1N1 in Thailand. We received quite a few of the viruses. They were isolated mainly in September and October, and we also have some viruses from Hong Kong down here at the bottom.

We do, of course, distill the information that we receive and try to present representative data to you. We couldn't possibly present all of the HI tables to you.

I gave you a bit of an orientation, well, when we looked at the H5N1 antigenic table, but I'll remind you that what we're looking for is a fourfold or greater difference between the homologous titer with the vaccine virus, in this case a New Caledonia/20/99, which is in our current vaccine. We have a homologous titer of 640 here with the New Caledonia antiserum, and we're looking for differences, fourfold or greater differences with the current viruses.

Now, you can see very clearly from our own CDC data -- and this was confirmed by data from the other four collaborating centers -- that the New Caledonia antiserum covers the current viruses very well.

this is reflected in a frequency table. We have only in our collaborating center had a total of 14 viruses, H1N1 viruses. They were all H1N1. We detected no H1N2 viruses at CDC, and 100 percent of them were New Caledonia-like.

And if we look back at the previous period from April to September when influenza viruses were circulating in the Southern Hemisphere, we picked up only one virus that was low to the New Caledonia antiserum.

Now, I'm going to go on and remind you that we sequence the hemagglutinin genes of geographically and antigenically representative strains. So we tended to sequence the majority of H1N1 viruses that came into our laboratory simply because we had so few and we were trying to track exactly what was happening.

Our New Caledonia vaccine strain is located down here at the bottom of this evolutionary tree for the hemagglutinin genes, and it may be just a little bit difficult to see, but we've color coded the dates of isolation because we were very interested sine we have two distinct clades or sublineages of (H1) HA genes. We wanted to see which of these two clades had the most recent viruses, and it's this clade at the bottom which has the most recent viruses.

Now, Roland mentioned that we had focused much more of our efforts on obtaining egg isolates, and we have shown in all of our evolutionary trees the viruses for which we have egg isolates. So you can see for H1N1s where really there hasn't been very much change. We were prepared in terms of having egg isolates ready should there be a surprise.

Now we're looking at the neuraminidase genes. As I mentioned, we have only detected H1N1 viruses over the past few months, no H1N2s, and you have a similar sort of pattern generally speaking with two different sublineages or subclades, but there haven't been that many amino acid changes associated with the ongoing evolution of the neuraminidase genes.

I forgot to mention for this previous slide that, of course, we are unable to distinguish antigenically the viruses in this sublineage versus this sublineage. So there are no antigenically detectable differences with post infection antisera between these two groups.

I'm going to skip the serologic responses. There are a number of tables in the CDC package. We did a lot of post vaccination human serology this year. We actually had two panels of serum from children, one panel from children zero to 23 months of age and another panel from children five to eight years of age.

And so if you have specific questions about those tables of serologic results, I'll be happy to answer them, but Roland will be doing a summary talk in which he compiles all of the serologic data accumulated by all of the collaborators to the WHO Global Influenza Program.

So our H1N1 summary is as follows. Relatively few H1 viruses have been detected worldwide. The majority of the H1 viruses were closely related antigenically to the New Caledonia vaccine strain, and no significant variance of H1N1 viruses were detected during recent months.

No H1N2 viruses were detected, and that is true with respect both to the U.S. strains and the strains analyzed globally.

N1 neuraminidase genes of recent H1 viruses were similar to those of viruses isolated prior to October 2002. So you can see the H1 situation is fairly straightforward.

Okay. We'll move on to H3N2. H3N2 viruses always cause us a lot of headaches. They are responsible for more severe influenza seasons, generally speaking, including a higher numbers of hospitalizations and deaths.

I'll walk you through this table which includes our reference panel up here. Here's our Wyoming vaccine strain right here with a homologous titer of 640.

Here is the Wellington virus here, Wellington/1/2004, which was recommended as a vaccine strain for the Southern Hemisphere. It has a homologous titer of 320, and then we have some relatively new variants which will be mentioned later on. We have the North Dakota/1/2004 virus.

And I would like you to note that the Wellington, North Dakota, California, and Singapore/37 viruses are all egg isolates, and therefore, I'm concentrating on data generated with viruses and antisera to these viruses that are potential vaccine strains.

Now, at the beginning of this season we were seeing that the majority of the viruses were similar to the Kentucky and New York/57 strains, and they were really quite well inhibited by antiserum to Wyoming. If you see a twofold difference, it's not considered significant because that's within the error of the test, but as the season progressed, we began seeing more viruses with titers of 80 and even a few viruses with titers of 40 against to Wyoming serum.

Once we got the California egg isolate and product a post infection ferret antiserum to it, what we found in this test and which has been borne out in our laboratory in other tests as well as in the other collaborating centers is that the antiserum to the California egg isolate covers recent strains better. It has a lower homologous titer, and there are not reductions or there are no more than twofold reductions compared to the homologous.

The same is true to a great extent with the North Dakota strain, although the California strain did seem to cover viruses slightly better, and we found out that the North Dakota egg isolate did not grow particularly well.

So I'll show only one more table, and I'd like to mention that this particular hemagglutinin inhibition test was done using getting pig red blood cells. The H3 viruses, the current ones grow relatively poorly, and sometimes it's necessary to use guinea pig red blood cells to detect high enough titers to do HI tests, and we do all of our screening for H3N2s using guinea pig red blood cells.

And if we have a virus with a low titer, it's too low to test with the turkey red blood cells, which were the red blood cells used for the previous HI test. Then we use guinea pig red blood cells to do the HI test.

Now, turkey cells are the standard cells used in all of the WHO collaborating centers.

What we have found, and it has been very, very interesting, indeed, that with guinea pig red blood cells, we often see that there's greater differentiation between strains, and so if you look at the Wellington homologous titer, it's 640 here, and it's dropped even lower against a couple of current strains, Victoria/500 from the Southern Hemisphere and the California/7/2004 from our recent season.

It doesn't matter whether you use guinea pig red blood cells or turkey red blood cells. The antiserum to the California strain covers recent isolates and even those that are difficult to quantitate using turkey red blood cells very well. In contrast, you see a homologous titer of 640, and a number of titers of 80 here with recent strains. And that's true no matter which continent you're looking at.

So in summary, I'd like to try to explain this frequency table which is a little bit more complex than the frequency tables that we normally have simply because we didn't have the ferret antiserum to the California egg isolate until January 5th.

Since that time, since the time we've been using that, we've been able to characterize 30 percent of the total 261 H3N2 strains that we've looked at from global sources as California-like, but we also had prior to the introduction of the California ferret antiserum been characterizing a number of strains that we haven't had a chance to go back and test retrospectively that were low to the Fujian.

Our guess is that until you do the studies you don't know for sure. Our guess is that these will look California-like when we eventually are able to test them retrospectively.

Last season, which isn't shown here, we had a bumper crop of H3N2 viruses, and we had well over 1,000 strains that were analyzed during the period preceding this, which was our influenza season last year, and the majority of those viruses were Fujian-like, as you will recall.

Okay. Now, you see the evolutionary tree for the H3 hemagglutinin genes. Our vaccine strain, Wyoming/3/2003, is right down here shown in red. Once again, the blue strains are our egg isolates. The Wellington strain, which was recommended for use in the Southern Hemisphere is here. You can see that we've moved up the tree from Wyoming for the Southern Hemisphere vaccine recommendation.

The one change that does not appear, the one amino acid change that does not appear in the Wellington strain that appears in the majority of the currently circulating strains is this change at amino acid 145 that you'll note. It's a K to N change, and it has in the past proved to be significant in terms of antigenic variation.

Our California reference virus is shown right up here. It is designated as a low reactor. Some of the other egg isolates that have been sent out to other collaborating centers and to vaccine manufacturers include New York/55/2004 shown here and New York/40/2004 and Wisconsin/19/2004, and these are all quite representative of the currently circulating strains.

The Singapore/37 strain, which was shown on the previous HI table, is right here.

When we look at the pattern of evolution for the N2 neuraminidase genes, we see that the neuraminidase genes are clustering in a fairly tight group. Of course, for those of you who aren't so accustomed to looking at the evolutionary trees, it's really not -- the distance between viruses is measured like this, not like this. These are spaced out.

So the vertical distance is not the important distance, and so these are clustering fairly tightly.

Here is the Wellington strain. Here is the North Dakota strain, and here is the California/7/2004 strain. New York 55 is right here, and of course, its neuraminidase is right on the backbone and very close to the consensus sequence for neuraminidase genes.

So in summary, Influenza H3N2 viruses have circulated in many countries, in the Americas, Asia, Europe and Oceania. In HI tests, H3N2 viruses were antigenically heterogeneous. Viruses isolated early in the season were often more closely related or most closely related to Fujian/411 and Wyoming/303 viruses, our two reference strains.

But an increasing proportion of recent isolates were antigenically distinguishable from these vaccine reference strains, and as I have shown you, were most closely related to the California/704 reference virus, both antigenically and genetically.

And sequence analysis of N2 neuraminidase genes of recent H3N2 viruses indicates that neuraminidases of recent viruses are genetically distinguishable from the Wyoming virus with these changes, but were very similar to the neuraminidases of Wellington and California.

Okay. Well, now I'll move on to Influenza B viruses. As Roland mentioned, there are two distinct genetic and antigenic groups of influenza viruses circulating globally. As I mentioned before, the Yamagata lineage viruses which are represented here in yellow have predominated both in the U.S. and worldwide. The Victoria lineage viruses are shown here represented in green. They are still circulating. The viruses that you see here are from Asia, but we had a number of viruses on the Victoria lineage from Florida as well -- from Hawaii as well.

We can see here from this table that the Shanghai/361 reference vaccine virus antiserum covers the current Yamagata lineage viruses quite well. That's true also in this test for the Jilin/20, a little bit less true for the Jiangsu, but there's the very high homologous titer here, and you'll note that there's a recent virus from Florida. We also did have some B activity in Florida earlier, and we had an egg isolate, which we were able to put into ferrets, and that particular egg isolate covers the current strains very well.

With the B Victoria lineage viruses, you'll remember that the previous vaccine strain was Hong Kong/330-like. Hong Kong/330 was used by some manufacturers. Hong Kong/1434 was used by others, and we're seeing that there's drift away from the previous vaccine strain. We've worked very, very hard to get a vaccine strain that would be suitable on the Victoria lineage, and what we found rather disappointingly is that as soon as we put the B Victoria lineage viruses into eggs, as soon as we isolate them in eggs, they lose the glycosylation site, and they tend to produce ferret antisera which don't uniformly cover the currently circulating B Victoria lineage viruses.

So we put a lot of effort into this, and have been relatively disappointed with the results. Nevertheless we're continuing to pursue this.

This is an updated summary of the Influenza B isolates characterized by CDC. Remember the previous table I showed as a compilation from all the WHO collaborating centers. As I mentioned, we're seeing quite a number of the B Hong Kong low, the viruses that are fourfold or greater reduced in titer compared to the homologous Hong Kong virus.

But the Victoria lineage viruses are a minority compared to the Yamagata lineage Shanghai-like strain, and the majority of the Shanghai lineage viruses that have been isolated recently are well covered by antiserum to the current vaccine strain.

Okay. I'll be showing evolutionary trees separately for the Yamagata lineage and the Victoria lineage. This is an advantage because you can actually see the strains better on the tree, but it's a disadvantage because you can't see the rather distant relationship between the Victoria and the Yamagata lineage viruses because they're not both on the same tree.

But I think for the purposes of our discussion today, it's best to do the presentation this way.

Here we have our Shanghai/361 reference virus. You can see that there are a number of amino acid changes that have occurred, but that we don't see a consistency in terms of viruses which are low reactors.

Again, shown in blue we have egg isolates designated. We have a large number of egg isolates, and I think I'll move on to the B Victoria lineage, HA genes.

Our previous recommended vaccine reference strain was Hong Kong/330/2001. You can see that the viruses have moved on. Here are some of the Hawaii strains, the egg isolates that have lost the glycosylation site and haven't produced good antiserum in terms of covering the currently circulating strains.

Again, I'd like to emphasize that the minority of viruses are on that lineage.

Now, if we look at the evolutionary relationships among the Influenza B neuraminidase genes, you can see that there are two distinct subgroups here, but the majority of viruses have neuraminidases in this group here, which indicate that they are reassortants between the two lineages.

So anyway, the neuraminidase genes are being tracked, and we can see that there are some differences, but we have representative strains from both lineages.

So in summary, Influenza B viruses continue to circulate in many countries. The majority of analyzed Influenza B viruses belong to the Yamagata lineage and are closely related antigenically to the B Shanghai/361 reference vaccine strain.

Most B Victoria/287 lineage viruses that we analyzed were reassortants bearing Hong Kong-like HAs and the Szechwan or Yamagata lineage neuraminidase genes. B Victoria lineage viruses were antigenically distinguishable from the previous vaccine strain from this lineage that was used a few years ago.

And then I just put up the summary table one more time in case there are any questions about the circulation of the different groups of influenza viruses that have caused problems around the world.

Okay. I think I'll stop there and entertain questions.

CHAIRPERSON OVERTURF: Are there questions? Dr. Monto.

DR. MONTO: Given the diversity between the Yamagata and Victoria lineage occurring in Asia at the WHO meeting last week was there any concern expressed in making a global recommendation for continuing with the Yamagata lineage?

DR. COX: We discussed that at length, and I think that what you can see here is that at least for the Asian viruses that we've had, there is approximately a 50-50 split. However, Japan is just experiencing the beginning of its influenza season, and it's predominantly B, and all of the viruses that they had obtained so far were B Yamagata lineage viruses.

So you know, our most current information was that B Yamagata was continuing to predominate in Asia even though the numbers we have are relatively small here for Asia and indicated about a 50-50 split.

DR. MONTO: Do we know anything from China? Because they've diverged from the recommendation, as we all know, in the past.

DR. COX: China has, as you probably know from press reports, had to close down their Institute of Virology due to a SARS incident, and that led to delay in analyzing and shipping influenza strains from the National Influenza Center in the Institute of Virology in Beijing to the WHO collaborating centers.

So we have not yet received a recent shipment with viruses from December and January.

CHAIRPERSON OVERTURF: Dr. Markovitz.

DR. MARKOVITZ: Yes, I wanted to ask a couple of questions. One, this is just for my information. With the two different strains of Influenza B, in the past when there has been serious illness in kids with Influenza B and deaths, is one strain more likely than the other, you know, one lineage, I should say, Victoria versus Yamagata more likely to cause serious illness?

And then the second question I had is if you could just tell us a little bit more about efforts to develop vaccines for avian flu. I know it's a long story, but if you could summarize a little bit about what different institutions are doing about that.

DR. COX: Okay. With respect to the first question about whether more serious illnesses are caused by Victoria or B Yamagata lineage viruses, we really don't have enough data to say definitively, but based on my knowledge of the characterization of viruses from children who have died or had serious illnesses, I would say that both lineages are capable of causing serious illness in children.

I would like to offer the opportunity to Pamela McInnes to really talk more about the vaccine development issues. I think I mentioned that vaccine reference viruses had been produced in three different laboratories around the world, one in the U.K. and two in the U.S., and a couple of those reference viruses have been given to manufacturers for production of pilot lots and so on, and Pamela has much more recent information than I have. So perhaps she could make a few comments.

CHAIRPERSON OVERTURF: Dr. McInnes, do you want to do that?

DR. McINNES: Do you want me to do that now or you want to finish any questions for Nancy? Whatever your preference.

CHAIRPERSON OVERTURF: Are there any further questions for Dr. Cox? Yes, Ben.

DR. SCHWARTZ: I just have a question about interpreting some of the HI data, and specifically with respect to the Type B. You've emphasized in your presentation that what one should look for is a fourfold difference between the various strains with a particular antisera, but you don't really emphasize at all the absolute height of the titers.

If you look at the B data and you look at the Shanghai 361 and compare it with the Florida/7, and both of them seem to be very good across the whole Yamagata lineage, but the Florida/7 titers are higher compared with the Shanghai, and I was wondering if that has any meaning whatsoever and whether that has any predictive value in terms of which may be a better vaccine strain.

DR. COX: We haven't noted that titer -- there is some ferret to ferret variation in terms of the height of titer, and sometimes when we get a very low titer with the particular ferret, we'll inoculate the same strain and we'll get a higher homologous titer. So there is ferret-to-ferret variation.

We have not noted a correspondence of the homologous titers that we obtain with ferret serum, and a corresponding either enhancement or diminution of titers in humans using those strains for vaccines, but it's a good question, and we have -- there are some factors about the hemagglutinin which we don't understand which makes some strains inherently more immunogenic than others.

We have been discussing and thinking about ways to get a better handle on what those factors may be and how to predict which strains might be the best vaccine strains. So far we don't have a handle on that, and we've often been very, very limited in terms of the number of egg isolates that we had available. I think we've put after we faced the Fujian situation where we didn't have an egg isolate. We've put an enormous amount of effort and had partnership with industry in this effort to obtain more egg isolates. So we really may have more options in the future, and it may be more important to really have a handle on predictors for immunogenicity in humans.

CHAIRPERSON OVERTURF: Dr. Eickhoff.

DR. EICKHOFF: Nancy, a question about the H3N2 data. If I'm reading the dendrograms correctly, and I may be stumbling on them, is the drift represented by the recent A/California-like isolates in the same direction as that started by the A/Wellington strain, or does it go off in a wholly different direction?

DR. COX: You are correct. The A/California virus has simply moved on from the Wellington strain. So it's just an advance. It's just Wellington progeny with a few more changes.

DR. EICKHOFF: And I also get the impression that the shift, the degree of drift, rather, is much less dramatic in this instance, in the A/California strains than the drift of the A/Fujian strain was from its predecessor. Is that correct?

DR. COX: That is correct.

DR. EICKHOFF: Up to a point.

DR. COX: Up to a point. I mean, you really have to look at the gestalt, and I think that when Roland begins discussing the human post vaccine serology, you'll see why I was greatly surprised by the human post vaccine serology.

It seemed to differentiate current strains even better than our ferret sera did, and that is the first time in my memory that that has been the case. So I was, shall we say, unpleasantly surprised by results for the H3 post vaccine human serologic testing that was done, and we'll get onto that later.

DR. EICKHOFF: Can I ask a further question then?

DR. COX: Certainly.

DR. EICKHOFF: And maybe this is a question that Keiji can answer also or can't answer also, as the case may be. Do you have any sense, considering that the A/California strains now at least in some parts of the country seem to predominate? Is this strain behaving any more aggressively epidemiologically and can we, therefore, anticipate that our season this year may go on further than it already has? Is that a fair inference to draw or not?

DR. COX: I think it is very difficult to say. I think we could have in areas of the country that were not so heavily affected by influenza last year, we could have continuing H3 and B activity. It's really difficult to predict how the strains are going to behave in the population. I think we just have to wait and see. We can't predict.

CHAIRPERSON OVERTURF: Yes.

DR. KARRON: Two questions actually about the H5 presentation. One was that you mentioned heterogeneity and antiviral susceptibility, and I assume that means a susceptibility to amantadine and rimantadine, and these are all susceptible to the neuraminidase inhibitors, or is that not the case?

DR. COX: You're right. I didn't go into a lot of detail. Not all of the H5N1 viruses that we received last year were resistant to the adamantanes, but all of those that were isolated from humans were resistant to adamantanes. So there were some in birds that were sensitive to adamantanes. We have tested all of the H5N1 viruses that we've received, and in laboratory tests, the viruses are sensitive to the neuraminidase inhibitors or we use oseltamavir. We don't use zanamivir.

And animal experiments done by Rob Webster and others have indicated that in vivo in animal models the viruses are also sensitive.

DR. KARRON: And then my second question was actually about I noticed that you had said with the ferrets when you -- obviously there's tremendous drift of the H5s, but they all, in fact, make antibodies to all of the viruses. I was wondering if you had any data from the survivors of H5 human infections about the quality of their HI responses to these H5 viruses.

DR. COX: A very good question. It has been extremely difficult getting serum from survivors. Oftentimes individuals are reluctant to give blood for a variety of reasons, and then it's often difficult to get the serum sent to us. We do have some serum in very limited quantities and have requested additional amounts and we're looking at the ability of serum from survivors to inhibit in neutralization tests a variety of viruses.

So we should have a better handle on that, and of course, one of the ideas for the pilot lot testing in humans is to look at the ability of the antibody induced by the pilot lot vaccines to inhibit a variety of the antigenic variance of H5.

CHAIRPERSON OVERTURF: Dr. Couch.

DR. COUCH: I want to pursue Arnold's comment a little bit, and I guess I really have a comment rather than a question for you, Nancy, and I was not here last year, but I'm sort of back in the same mode I was two years ago with the Influenza B strains.

And you look at these epidemiologically. Well, I think mostly people know they were dominant in Asia for a number of years before they began to show up in the rest of the world, and then they've been jockeying with each other for dominance would be the way I would describe what's gone on in the last few years.

And our approach to handling that dominance is to guess which one is going to be dominant in a coming year, and it has been amazingly successful that the Victoria derivative was guessed the right year and then went back to Shanghai. But I think it emphasized that we're guessing is the point I wanted to make.

And we've looked at the serologic data in the past that looks very much like the same data we have this year, and that is that adults and elderly individuals have a reasonable degree of cross-reactivity to vaccine responses to either one. You get the Yamagata derivative and you have a reasonable cross-reactivity to the Victoria and vice versa.

It's children where the differences are really distinctive. If they got the Victoria vaccine, they have very little immune response that you would say would be protective against the Yamagata derivative and vice versa, and if you wanted to look at some of that data, what's going to happen with some of our guessing, some of you remember an old term that Paul Gleason brought in a few years ago of herald wave. We're looking at a herald wave, you see, of the Victoria derivative, and here we're guessing that that's not going to be true. It's going to be the Shanghai.

I think I asked Walter the last time I did this if he could remember, but the two separate Bs have been used in the vaccine in the past. I meant to go back and check when that was done, and you see, I tried this a couple of years ago and it didn't fly. I'm going to try it again.

You see, based on what we know about these immunologically, if you took that 15 micrograms, and let's assume that at least for the time being is fairly rigid, and you split it between those two and you look at the responses in the elderly, adults and the elderly, that they would be pretty good to either one if you've split it, you know, based on what we know about the cross-reactivity of the 15 micrograms of each.

That might not give you as much response as you'd like in those children, but it would insure that you've got protection against both of these strains, which will be present, we would say, and hopefully when we only pick one we're guessing the dominance, but that would be of less concern.

Now, I know that has not been pursued, and you need immunologic data to go along with that kind of thing, but if we continue to see these two jockeying, I think we need to think about ways to approach that other than immunologically with vaccines rather than just guessing which one is going to be dominant.

So I didn't have a question. I meant it as a comment, unless you want to add to that.

DR. COX: I think that you have hit the nail on the head. As long as we have these two quite distinct lineages of Influenza B circulating worldwide, we are making an educated forecast for which virus is likely to predominate, and we could be wrong.

I think that in an ideal world, we would have a tetravalent vaccine, but we're not in an ideal world and we know that that would reduce the number of doses of vaccine in an environment where we already are concerned about vaccine supply. We know that young children respond relatively poorly to the influence of the B component of the vaccine. We have to think about the manufacturing issues involved and the standardization issues, the whole complex of issues that these changes bring up.

And I think we are in a dilemma, and we should talk about these things, but we should also recognize that there are many practical issues that would go along with the departure from the way we've been doing this.

Roland may wish to add something or others may wish to.

CHAIRPERSON OVERTURF: I'm going to give Dr. Couch the last say for this segment and then we'll take a break for 15 minutes, come back, and we can discuss that.

DR. COUCH: Manufacturers certainly would not like to hear adding a different antigen, but part of my point, Nancy, was that we have had a tetravalent vaccine. So we have the precedent of it being available and it having been circulated, and that tetravalent was with the B strains.

CHAIRPERSON OVERTURF: So I would propose we take a 15 minute break and be back at 10:45.

Thank you.

(Whereupon, the foregoing matter went off the record at 10:31 a.m. and went back on the record at 10:55 a.m.)

CHAIRPERSON OVERTURF: Dr. LaRussa.

One more comment about the B issue, and then we'll turn to Dr. McInnes.

DR. LaRUSSA: So a number of us pediatricians were talking on the side during the break, and before I make this comment, I want to emphasize that I'm not proposing we do this for this coming year, but just to plant this seed for the future.

If, in fact, it could be shown that if you put the two B lineages in the same vial and you could get good immunologic responses in children, one way around doing this dance we do every year about which B lineage we're going to pick is to make a separate pediatric vaccine, which you'll already do. You'd know that you'd have a stable demand for it every year. You'd make your eight million doses. It would be there because we do very well with immunizing kids, and you could get around this whole issue because one of these years we're going to do the wrong dance and pick the wrong strain.

So what I would propose is that we think about the kinds of serologic studies you would need to do to show that you could put two B lineages in the same vial and get a good response in kids, and then think about a separate pediatric vaccine for the six to 24 age group.

CHAIRPERSON OVERTURF: Dr. McInnes.

DR. COUCH: It's testable.

CHAIRPERSON OVERTURF: Yes, you can stay there.

DR. McINNES: Thank you very much.

This is a summary update. A year ago at the vaccine strain selection meeting, Dr. Lambert made a presentation on H5, and the initial responses of different agencies within the Department of Health and Human Services, and this is an update for you.

It is personally been one of the most gratifying experiences in government because I think the flu machine within government has always worked very, very well, but this has been really a marvelous experience of people really working extremely well together, as well as with manufacturers and other government contractors, working very hard.

Nobody should underestimate how seriously the department is taking the threat of pandemic influenza. It is the subject of a great percentage of our lives and of our time, and I will just summarize. I'm not going to provide lots of details.

The reference virus, you heard Dr. Cox talk about the reference virus for the H5. The particular reference virus that is being used to provide the pilot lots of vaccine that I will be talking about was produced under a government contract, and it utilized reverse genetics technology to make this reference virus, and it turned out to be a real test of the select agent rule.

And so the dry run of, in fact, going through the process of generating the data and the pathogenicity data on these genetically engineered viruses did go through the select agent rules and, in fact, the data were very compelling, and it resulted in the exemption from select agent rule.

And we still though were subject to the U.S. Department of Agriculture permitting process to move this virus. So the reason I am sharing these pieces with you is that many of the pieces that will be in play during pandemic response have, in fact, been tested in this past year and hopefully will facilitate the path for future journeys.

Pilot lot contracts were awarded to the two licensed and inactivated vaccine manufacturers in the United States, licensed in the United States, and that was Aventis Pasteur, and I'm not sure if I should be calling it Sanofi now, but when I say "Aventis," I hope I'm calling it the right name, and Chiron.

So the contracts were awarded around May of 2000, May through June 2004. The reference viruses were produced, characterized, exempted from select agent and moved to both manufacturers, and both manufacturers have been underway with pilot lot production of an H5N1 inactivated vaccine candidate.

The quantities of doses that have been procured for the pilot lot scale is less than 10,000 doses from each of the manufacturers, and they're in two different dose concentration formulations.

Aventis has completed its bulk manufacture, as well as its formulation and filing and finish and testing, and the IND. We have filed the IND for the clinical evaluation of this candidate.

Their clinical development plan has been designed and laid out and will be implemented through the NIAID vaccine and treatment evaluation unit contracts. We have proposed two different programs, one for the Aventis candidate, one for the Chiron candidate because they'll be available at different times, as well as we want to insure access to the appropriate number of individuals in each of the target population groups.

The clinical trial scenario will begin with a trial for safety and immunogenicity in healthy adults, and with those data in hand, we'll move to evaluation in the elderly and in younger children.

Coupled with this for the pilot lot scale of production, the department awarded a contract for the commercial scale production of an H5N1 vaccine and the intent of this was to test the commercial production capacity and the ability to respond, as well as handling of these candidates in the facilities. This is a much larger scale.

This has been completed. The vaccine bulk has been made. The formulation, finishing and vialing will be dependent on the data that come out of the clinical evaluation program of the investigational pilot lots.

In addition, the department has awarded a contract to secure egg supply year around to enhance our capacity to respond to produce pandemic vaccines. Couple a deliverable under this egg securing contract is, in fact, pilot lots of investigational candidate vaccines for pandemic preparedness, and so over a period of years, several of those will be made and will be evaluated, all of which is designed to build on our knowledge of safety and immunogenicity, the human response to these novel antigens.

That's really a summary of where we are on the response to H5.

CHAIRPERSON OVERTURF: Dr. Markovitz.

DR. MARKOVITZ: Yes, thanks.

I had a couple of questions. One is how did people get around the issue that these tend to kill eggs, chicken eggs?

And then second of all, how are the vaccines that you're talking about -- are they going to be able to deal with this heterogeneity issue that Nancy Cox was alluding to or is that going to mean that we're going to have to be, you know, similar to what we do with the other strains, sort of constantly revising them?

DR. McINNES: Sir, the issue around being able to get a meaningful yield by growth in eggs is dealt with by engineering the virus, the wild type virus so that, in fact, you're going from the tiger down to the pussycat that can, in fact, be grown in egg, moved to the manufacturers and dealt with under usual biocontainment levels. So that, in fact, happened very successfully and yield was quite good.

We would anticipate that the early clinical studies will, in fact, be generating sera in response to this particular reference virus candidate that can be the subject of investigation in terms of what sort of protection one might derive to strains that have some change in them or some drift in them.

DR. MARKOVITZ: Can you tell me more about how it was mutated without violating proprietary matters?

DR. McINNES: There is, in fact, a publication on this, which I'd be happy to share with you.

DR. MARKOVITZ: Yes, that would be good.

CHAIRPERSON OVERTURF: Dr. Eickhoff.

DR. EICKHOFF: Pamela, I know regarding the reverse genetics technique great concerns have been expressed in the past about intellectual property rights. Have those been resolved or is that being addressed?

DR. McINNES: It has been, I think, addressed extensively, and I think there are paths to resolution for this. In the investigational framework and experimental framework it's not an issue. The issue comes around the commercial area. So, yes, this is the subject of a lot of discussion, and I think there are solutions on the table to deal with it.

CHAIRPERSON OVERTURF: Dr. Karron.

DR. KARRON: Are there differences in this Sanofi and the Chiron products or are they essentially the same product?

DR. McINNES: Our goal was to go as close as possible to their currently licensed methodology and formulations because we felt that that would assist in a licensure process in an emergency situation. So you wouldn't be dealing with trying to deliver huge amounts of vaccine under IND.

So given that they produced them in a pilot facility, it's not identical, but the hope was that the process would be as close as possible and the formulation would be as close as possible to their license formulation.

CHAIRPERSON OVERTURF: Were there any other questions or comments anybody else wanted to make? Yes, Dr. Royal.

DR. ROYAL: Thank you.

One would expect that since this is a genetically engineered virus that one could introduce a series of mutations and establish panels that one could use to screen different strains. Is that sort of approach being considered?

DR. McINNES: I don't quite understand the question.

DR. ROYAL: Well, since you said that this is a genetically engineered virus, so one would expect that one could introduce mutations in the hemagglutinin and other regions that could be used to generate reference sera to anticipate, so to speak, strains that might arise instead of waiting for various strains to appear on the horizon and sera to be generated at that time.

Is that sort of approach being considered?

DR. McINNES: Along with this program, there is a large effort and CDC is involved in it as well, which is a genomic spaced program to actually characterize a whole series of viruses in the intent that you might select particular cannons for the particular tributes to manufacture vaccines.

The reverse genetics approach that I'm talking about here is specifically for generation of a relevant reference virus which is taking the two relevant genes and inserting them into a backbone, a worker virus backbone to facilitate the yield.

So, yeah, there is a lot of discussion about how one might generate hypothetical candidates, whether you would utilize genomics technologies, whether you would engineer technologies, whether you would use wild-type, classical reassortant methodologies. It is a great deal of discussion around what might be the appropriate candidates to try.

CHAIRPERSON OVERTURF: I think if there's no further questions we should probably proceed on the agenda with the report from the Department of Defense. Linda Canas.

MS. CANAS: Hello. Good afternoon or morning.

In 1942, the precursor organization to what is now the Armed Force Epidemiological Board began a series of clinical studies using concentrated inactivated Influenza A and B virus vaccine. In 1943, the next year, there was an Influenza A outbreak. In the control group, those people who had not been vaccinated, there was a three to six percent increase in illness compared to those who had received this trial vaccine.

The incidence of hospitalization was seven percent in unvaccinated individuals and two percent in vaccinated individuals. As Dr. Levandowski indicated, in 1945, the military licensed this vaccine, but in 1947, it was quite ineffective against the circulating virus, and thus we proved that the changing nature of the influenza virus compels us to match the vaccine each year with the virus that we expect to circulate.

In 1976, the Air Force began a surveillance program that was fondly called Project Gargle, which was based with a series of sentinel sites around the world where we had people stationed, and this was also public health, wanting to know what's going on with our active duty members and their families.

And in the process, we were in areas of the world where influenza was emerging and causing disease so that we could share that information with those organizations that would make the vaccine decision.

It is a mandatory vaccination for active duty military.

In the mid-'90s, there was a presidential decision directive establishing the global emerging infection system, GEIS, and their mandate is to be a force against microbial emergence of pathogens and to assure the biosecurity of the United States, and we had influenza already, a program that was well established. So it only made sense to expand this and make it a tri-service program.

So in 1997, it did become tri-service with Air Force, Army and Navy participation, and the influenza part of this program. In the GEIS Program there's two wings. I'm going to be talking today about the part that we do in San Antonio, etiology based, where we have sentinel sites, and whatever they send into our lab in San Antonio we work up for viral pathogens and report anything we find.

The other wing is Naval Health Research Center in San Diego, California, and their sentinel sites are all of the recruit centers of each of the services. They know the demographics. They have an established population. They keep track of the febrile respiratory illness going on in that population, and based on those demographic numbers, a certain number of samples are collected and submitted weekly to this lab in San Diego. So they can establish rates and know trends and tell when something is getting out of hand.

Now, I always think it's important that the board understand how this program work. We have a variety of layers of cooperation which make it work, make it effective, and also give us opportunities that we can respond to.

As I've indicated, DOD GEIS oversees the program, and laboratorians and epidemiologists from all of the services work together. Each year we do have an annual meeting. Last year it was in late May in San Diego, and 27 individuals from each of the services, again, laboratorians and epidemiologists, representatives of the central hub of GEIS.

We had people from Bangkok lab and the Cairo lab there, and also a representative from CDC, and the purpose is to analyze the program from the year before, much as we're doing here today, see what we expect to come about, where our opportunities will be, how we can cooperate and work together.

We choose the sentinel sites. The epidemiologists make sure that they have all the information they need to run this program. They put together a PowerPoint presentation that goes out to the Public Health officers of each of these sites that they can use to develop educational programs for their providers so that they understand what the program is about, what it's trying to accomplish and what their role is.

We're very aware that we're adding to their daily work load. So to try and involve them and make them benefit from it as much as possible helps everyone.

During the season, a weekly report is put out on the Web. Anyone who has access to a dot-mil Website can access this report, and this year, actually very recently, it is now being posted on the Epi-X Website.

Over in the laboratory we make sure that the sites all have the collection materials, they have all the information on what we expect from them, how they should collect the samples and preserve them, and of course, this is all best case scenario.

I need to emphasize that this is a full service virology lab working within a full service reference lab. This is not a stand-alone program, and that helps us because we have the infrastructure of a lab. We have FedEx contracts set up for deliveries from around the world. It's very easy and timely to collect two or three samples that meet a case definition, get the in a FedEx box going out the next day, and arrive in our laboratory in a timely manner.

We do operate this as a clinical program, and for the most part everything is set up the day it arrives in our lab.

When we do get flu isolated, we use conventional methods where tissue culture. We want the isolate. We're not interested in just knowing if flu is there. We want the isolate. So we do use a variety of different tissue cultures, and once we have those, then we become more modern and have various molecular tests that we can go on to characterize as virus.

And, again, I said it was a clinical program. So when we have it isolated, it goes back to the submitting site as a patient report. The epidemiologist notify the Public Health officer what's going on at their base for two reasons. First, we're giving back to them information that they have provided for us, again, trying to involve them in the program, but also so they can go into the Air Force reportable event surveillance system, AFRESS. Culture confirmed influenza is a reportable event, and this gives them all of the information they need to go in and update those records.

Meanwhile back in the laboratory, after we've identified as Flu A or B, we do go in and subtype selected samples, and some of those go on to be molecularly sequenced, and all of this information is then shared with CDC, and then I get to come here and tell you about it.

This is our map for this year. The red stars indicate those sentinel sites that are new this year. The blue ones have been around for a while. The green sites are representative from Army in Thailand and the Navy in Lima, Peru, where they've had longstanding research labs, and we've been able to hook on with their protocols. They have IRBs, and do surveillance in the local populations. This has proved to be very helpful.

There's other labs, Kenya, Cairo, and Jakarta, that are also supported by GEIS, but I do not have their information here today.

Reporting from the Air Force, we can say that 80 percent of the active duty Air Force have been vaccinated as of the 14th of February. When the vaccine shortage issue was first announced, the decision was made that the priority status in the active duty military forces would be the same as what had been suggested by CDC with the addition that people who were deployed or deploying would also be in that priority group.

Eventually the recruits got FluMist, and they are all being vaccinated now and FluMist is available for other active duty members. There is now vaccine for all of the active duty members of either the injectable or the FluMist.

We had a great many questions in the beginning about FluMist. It is an inactivated virus. Were we going to pick this up in the laboratory?

So we had the sequence information for what the seed strain, the A/Wyoming would look like. We did not look for this in every virus we got, but in those that were referred to us as being particularly of concern, we did look at those, and we have not found any that match that.

This slide was submitted by NHRC in San Diego. As I said, they are responsible for surveillance of the recruits. In the military there's a particular problem with adenovirus in the recruit centers.

So NHRC is to adeno what our laboratory is to influenza, but this year right around the time when they were just getting to vaccinating the recruits, they have isolated a total of nine Influenza A viruses. One of those was in a person who had been vaccinated for more than two weeks. All of these have been subtyped as H3N2.

Now, every season has its own personality. I'm preaching to the choir here, and this graph kind of shows dramatically the last two seasons.

One of the things we do in this program that we're particularly trying to get out that we are a resource for outbreak investigation, both laboratory support and epidemiology support, and this was dramatically exemplified this past summer when we received word from the lab in Thailand that there was evidence of an influenza-like illness in Napal.

So two of the researchers stationed from Thailand that were working in Katmandu traveled to southeastern Napal to a refugee camp. There were actually three. When I made this slide I thought there were two camps, but there were three camps, and between July 1st and July 3rd of this past summer, they collected a total of 64 samples from hospitalized patients.

They only went to the hospitals because of the political situation in Napal. It was not considered safe to be traveling through the countryside.

Within two weeks, we had those 64 samples in our lab in San Antonio and had isolated 42 Influenza A viruses. All 42 were subtyped to be H3N2, and I want to say here I don't have HI data, nice pretty charts like Nancy Cox does, because we do not have animal studies. We don't have ferrets. We use the WHO reagents, which basically just gives us a breakdown of whether it's H1 or H3. So more extensive studies we send off to CDC and also then we do some molecular work.

And that's what we did here. Twenty-seven samples of the 42 that were positive were randomly selected to see if we could look at a molecular basis for this outbreak, and we studied the hemagglutinin gene, and in the majority of those 27 isolates, there was a four amino acid substitution from the A/Fujian virus. All four of those were either within or near an antibody binding site.

And here we have the signature, K 145N substitution that Dr. Cox mentioned earlier as one of the samples that had drifted, and we do feel that this shows there had been a significant drift away from the A/Fujian viruses.

This is a graph of our season so far. Judging from the way samples are coming in, we have not peaked. It is an A season predominantly. We had received just under 1,300 samples as of the 9th of February. Thirty-seven percent of those were positive for any respiratory virus. Twenty-six percent were positive for Influenza A.

We have subtyped so far about 40 percent of those, and they have all been H3N2, and a portion of those then have had nucleotide sequencing and all of those have shown the four amino acid substitution that I mentioned earlier.

Overall three percent of our viruses were B, and it was a mix. It was a 50-50 mix between B/Yamagata lineage and the B/Victoria, but there was a very definite difference in the season. We got samples in from Hawaii late summer, early fall, and they were split with the Hong Kong and the Shanghai. That made us nervous about what we were going to face this year.

All of our other Hong Kong isolates have come from Peru. One of those was isolated in December, but all of the others had been from the past summer, June, July and August. Everything we've seen so far -- now, we're not talking bit numbers here, 38 total -- have been Shanghai.

Just this past week we already started getting more B viruses. So we haven't had a chance to subtype those yet, but so far everything late in the season has been B/Shanghai.

If we looked at a breakdown of our basis, of our sentinel sites compared to CDC regions, our season started much as everyone else's did with the early results coming in from upstate New York. We're still getting some samples from them, but it has moved much more down coast into New Jersey. All of those As that we have subtyped have been H3N2, and I think there's only one B that we've had in this region prior to last week. It was the B/Shanghai.

We've started getting a lot from Alabama and Mississippi, and of course, we just have specific bases. So it's very dependent on what they send us from those, but we had a sudden surge of A from these bases and a few Bs that came late.

This is probably the region that has the greatest number of sites, and it does include Lackland Air Force Base with the Air Force recruits. Now, the Air Force recruits are in the sentinel site for NHRC, but ours is a clinical lab, and it's Air Force, and we're in the same city, and we just handle their samples for clinical purposes, and because it is recruits, we've had a great deal of the adenovirus early in the season.

And, in fact, the big mystery right now is where is adeno at this point. We're just not seeing it. We're getting a lot of samples, and we are getting flu. Certainly there's a lot of flu in the area, both A and B at this point.

There's only one site in Illinois, but they're very enthusiastic, and I was concerned last week when we got three Bs from them because I hope they're not getting another wave.

Colorado has the Air Force Academy. We've always been very concerned about those. They are not in barracks, as the recruits are, but it is a training center, and for public health purposes, we're very interested in detecting early outbreaks, and of course, knowing what they have that's circulating there.

Overseas our early isolates came and still continue to come from Lakenheath Air Force Base in the U.K. They seem to have more flu than the surrounding area. So it's probably enthusiastic Public Health officers making sure that they are doing a good job of surveillance.

Most of what they have is A. It has been very boring I'm told to do their work because they look exactly alike. They are H3N2, again, with the four amino acid substitutions.

They are beginning to pick up a little more B, and the ones here, again, are B/Shanghai.

Italy and Germany are just coming up. We've got quite a few from there, and I should mention in Germany, in addition to our Air Force base there, there's the Army Landstuhl regional Army medical center. They have a very good virology lab of their own.

And the same is true with Tripler in Hawaii, Tripler Army Medical Center. They isolate the viruses there and then send us the positives. So we're only working up the positives from them.

Of particular interest are those areas that are deployed, and there are many challenges to getting samples out of the deployed areas. Influenza surveillance is not their primary objective, and even if they're concerned about it, they don't have supplies. They have limited storage space. They don't have any dry ice, but they are concerned.

This represents a site in Iraq. I reported last year on a site from Kyrgyzstan. We haven't gotten anything from this year, and we're told they're not seeing any respiratory illness, which of course is the good news.

We have 13 isolates here. Eight of the people had been vaccinated before they arrived in country. The other five were vaccinated when they got in country. We don't have any information on where they got the illness, if it was from the surrounding community, other individuals. I don't have any denominator status.

We don't understand that this is a particularly big problem, but these are the only ones we have received, and it has worked well. This is where you just do what you can do: collect the samples and ship them however you can. They do come FedEx, but they don't have dry ice. So we've given them gel packs to use, and we're getting very good isolation.

So we go for best case and settle for whatever we can get from some of these more esoteric sites, and of course, the deployed areas we continue to monitor for any of the respiratory pathogens.

I mentioned that in Hawaii we get from Tripler where they isolate their own. We are just beginning to see samples from Asia. They are a primary concern. They're very interested. We get a lot of queries about what's going on. Many of them are very aware of their location in regards to the H5.

But they just now started coming in. We did have reports a couple of weeks ago of an outbreak in Hokkaido, but it was an exercise. It wasn't an established base of any kind. They had no samples or no collection supplies. We tried to get things to them, but of course, there are limitations, and of course that illness had run its course by the time we got them there.

There was even an effort made to contact the local hospital where the patients had been sent to see if we could get isolates from them, but there was a language problem. So hopefully we'll start seeing more from them.

South America sends us a shipment of between 150 and 200 samples every few months. In fact, we just got a box in this past week, and they're very good. We get a very good percentage of isolation in what they send us.

We have seen a lot of flu A, more flu B from them than we have seen from some of the other sites, as I mentioned, most of it being the B/Victoria lineage.

This slide represents kind of -- there's always an interest. If it's not flu, what is it? So we put together this slide, actually added on the last year. So it's a fix full seasonal year review. It kind of represents seven different seasons on this slide.

But our case definition is specific for influenza. We have been stressing this year to make sure fever is in the case definition of the samples that are selected. So we look for flu and we get flu.

We do have a variety of other things circulating at the same time, and we have picked up several of the enteroviruses and parainfluenza viruses.

I think there are two unique things to our population that I need to point out, the first being adenovirus. Because of the recruits and what we pick up with NHRC and ours with Lackland, we do have a higher percentage than you would see in the normal civilian population, and just the reverse is true for RSV.

RSV is a very big problem this time of year for babies and immunocompromised, and they don't make up very much of our population. And even if it is there, the RSV virus is very fragile and is better detected on site with a rapid test. It doesn't survive travel very well. So a program like ours, we get one or two a year at most.

But this is one of our big concerns, and it is certainly a goal of the program because we're back to the same thing. If it's not flu, what is it?

And SARS really brought this to the forefront. When we really needed to know a sample was coming, and we didn't get any samples from SARS, it was just the intellectual discussion in the labs that were getting them, it's difficult to actually identify. So if you're in a SARS endemic area and you have a symptomatic patient and you get a negative, how do you know that it's not SARS? You'd much rather know what you do have.

We're working very hard to get tests. There are tests, but they're not easy and they're not cheap to do in the laboratory for things like chlamydia pneumoniae, mycoplasma, Legionella, pertussis, and of course SARS and the other coronaviruses and things yet to be determined.

We do like to think of ourselves as being in position to handle these kinds of outbreaks. It's a challenge because most of our work certainly with flu, we do large numbers and we emphasize safety all the time, and the laboratory technicians are very aware of safety issues, but I will tell you last February we got a sample in from our lab in Thailand, and the report sheet said, "From soldier sick after culling chickens in avian-endemic area," and that got everybody's attention.

Well, it's nice to have that information because we immediately took that sample into the BL-3 laboratory, and always in the laboratory it's the ones you don't know that you have to be concerned with. So it was a real eye opener to all of us that, oh, yeah, this is true. This can really happen.

We never know what's coming into our lab. We get so caught up in the Flu A and B that sometimes we have to step back and make sure that our procedures are going to be there to cover things that are more pathogenic.

If we summarize our season this year, it has certainly been predominantly A. All of what we have seen has been H3N2, and all of those have had this four amino acid substitution indicating to us a drift away from the A/Fujian last year and even the A/Wellington from South America.

The B, while it's in our population evenly split between the two, it suggests to us since the latter ones are Shanghai that that is the way it's continuing to go, and we have nothing to add to the H1 story. We know it's out there, but we haven't isolated even a single one.

Thank you.

CHAIRPERSON OVERTURF: Are there cases for Linda Canas? Yes, Dr. Farley.

DR. FARLEY: On the current slide that you have up, I'm just noticing that it seems in 2003 that things changed a bit in the distribution or the prevalence. Did anything change in the methodology. The influenza isolation rates seem to have gone up, and the adeno went down a bit from the previous years.

MS. CANAS: The adeno is difficult to understand because a lot of times they just quit sending us so many samples. If you look at the '99-2000, everything we got was adeno. The number of samples went from 2,000-some to over 6,000 because they were giving us so many. So adeno went up.

So they quit sending us so many. So adeno, it's hard to tell what that means because we weren't looking at them.

And the other one was Influenza A and what?

DR. FARLEY: Well, the Influenza A isolation rates, they now represent 25 percent in 2003 and 2004.

MS. CANAS: Flu has its own. A lot of times it's the people who are taking the samples. I think in the last two years, I don't have definitive data on this, but I think we put out information on these rapid flu tests about their sensitivity and their specificity, and they should culture the negatives. I think they send us all of their positives, too. I know some places do because I've talked to them. "Oh, but we want you to see what our virus looks like."

So I think the last two years that's been a big part of it, but we've also stressed a lot more the case definition. I reported last year that we did a records review at one of the sites that was overwhelming us, and of those patients that had fever in the diagnosis, 40 percent were positive for a virus. Of those that did not have fever in the diagnosis, eight percent were positive. So that was strong indication to us that this was a way.

This program keeps getting bigger and bigger, and our resources don't, so to try and target those that truly are ill with influenza or respiratory virus. So those two things, I think, have.

DR. COUCH: Just a quickie. You partly answered that, I think. Your case definition for sampling is just acute respiratory illness in the reports for health care?

MS. CANAS: Yes, they do visit, and it is ‑-

DR. COUCH: No fever criteria or syndrome criteria?

MS. CANAS: Fever of 100.5 or greater.

DR. COUCH: They must have fever?

MS. CANAS: Well, that's what we tell them, and we give them sheets and they fill them out, and you say fever and they say 98.7.

DR. COUCH: Otherwise your percent of rhinovirus, of enteroviruses and no rhinoviruses would be very --

MS. CANAS: Well, we don't have the right conditions for rhinovirus. So I suspect a lot of our negatives are rhinovirus. Our temperatures are around 36. You need to be 34 to 35 to isolate the rhinovirus.

DR. COUCH: But you are asking for fever.

MS. CANAS: But you're asking for fever, and that's what we're stressing with the FluMist also, when people are reporting after getting FluMist that they're sick. As long as they have the fever and submit it, then we can ascertain whether it is disease or not. Those that don't, the FluMist people feel bad afterwards, but they don't have fever.

CHAIRPERSON OVERTURF: Dr. Monto.

DR. MONTO: Two questions. First of all, could you update me as to the situation with adenovirus vaccine in the military? I know there was no vaccine for a while. Are you vaccinating now for adenovirus, number one?

And, number two, are you able to get anything about seasonality in other parts of the world from your data? This has become a major issue in trying to identify impact.

MS. CANAS: Right. I'll answer the second one first and then refer to Colonel Phillips for the adenovirus.

I one time made a comment about when South American first started sending samples, that, oh, good, we could get theirs in our off season when they had flu season because they're Southern Hemisphere. That just made sense to me.

And the reply back was, "How do you know that? We're trying to figure it out."

And, in fact, we get flu from them all year, and they say it's because they have so many different sites and their topography in the country where they have some around the coast, some in the mountains, and it varies more with that, and I can only comment. The rest of them tend to follow the same seasons that we see.

Colonel Phillips.

COL. PHILLIPS: Regarding adenovirus, and you had commented on the epidemiology on the slide there, 1999 was the last year that we had adenovirus vaccine and were able to give it, and have seen steadily cases since then.

We are aggressively pursuing the pursuit of the new adenovirus. Our researchers are working real closely with the folks from Bar Laboratories that has the contract for that.

Phase 3 clinical trials have begun and are underway currently, and we're anticipating FDA licensure of a vaccine by 2007.

CHAIRPERSON OVERTURF: Yes, Dr. Dowdle.

DR. DOWDLE: Thank you.

One of the real advantages, I think, of this system, global system, in terms of influenza surveillance, of course, is sampling sites throughout the world, and it's a real opportunity for sort of timely sampling in the sense of you know what's going on.

The question I would have is how does the timeliness relate to, for example, other shipments that Nancy would get in from other parts of the world and other WHO centers. Does this system lend itself to completing or being ahead of some of those or would it be lagging behind, depending on how the shipments are arranged?

MS. CANAS: Well, I can only comment on our system, and those that come in from South America and Thailand do have a lag, sometimes as much as five months, but the others are coming in a day or two after collection, and then within a couple of weeks, we try to get anything off to CDC, but I don't have any idea.

DR. COX: I can comment just a bit further on what Linda said and can concur with what she said. In some cases, we do get viruses through the military surveillance sites on a very timely basis, and it's really excellent.

She mentioned but didn't emphasize probably quite as much as I would like to emphasize that there's a lot of sequence work going on, which feeds directly into the sequence information that we have and, therefore, that goes into the WHO system.

In other cases, we get viruses, for example, from Thailand, directly from their National Influenza Center, in a more timely manner than we do from the military. So it varies depending on site and a variety of circumstances.

What we have found generally is that GEIS is a tremendous value added to the WHO global influenza program, and that we would very much like to see it continue, the GEIS activities to continue.

Along the same lines, we've had the opportunity this past year through efforts of the department to work in closer partnership with some of the military labs and will be providing some financial support for NAMRU-2 and NAMRU-3, to enhance the already existing influenza surveillance infrastructures that are in place there in the hopes of both expanding influenza surveillance into more rural areas in Indonesia and some other countries, and also having more timely shipments of viruses to our WHO Collaborating Center.

So I think there's very close interaction and real value added is the bottom line.

CHAIRPERSON OVERTURF: Any other comments?

One thing I didn't hear was whether there's any enhancement of this or has been an attempt to enhance the system recently to look for H5, H7, H9 strains, particularly from those areas where they used to come in the past.

What I heard was that you rapidly grow the viruses that you tend to characterize their hemagglutinins. Are there additional epidemiologic data in those sites where these viruses are likely to emerge that are collected that in any way changes your operation at that point?

MS. CANAS: Well, this is what we struggle with. As long as we can keep geography in that mix there, we feel pretty safe that those things that are coming in from endemic areas we can handle in the BL-3 laboratory.

We work again with CDC to know what's going on in the world. Other than that it's just trying to keep up with what's going on and be prepared.

CHAIRPERSON OVERTURF: Are there other questions or comments? Yes, audience.

DR. GAYDOS: My name is Joel Gaydos. I work with Linda and the other people on the GEIS program.

In response to some of the questions that have come up over the last few minutes starting with Dr. Dowdle, one of the reasons that we meet once a year is to make sure that the programs are complementary, and there are a couple of things that I think are important in addition to what has been brought up.

We have people who are permanently stationed at places like Cairo and in Lima, and we do a lot of bulk processing. But if something is happening, we can speed up the processing. If something happens at the CDC and they think that we should be getting more specimens from Napal or from Peru, then we get the word from them, and then we go out and collect the specimens.

We also have the opportunity because of having military people stationed there to get into some areas where other people aren't and where specimens would not be easily obtained, like some of the places you have heard about in Napal.

So I think the program has worked very, very well with the CDC. As Linda pointed out, specimens come to her laboratory. She works very closely with Atlanta and also with the Navy group in San Diego.

Not all of our specimens come in that way. Our lab in Cairo for a number of reasons works directly with the CDC. Our lab in Jakarta for a number of reasons works directly with the CDC or with the laboratory in Australia.

So I think it is, in fact, a surveillance system in that we are regularly collecting specimens. We're working together.

We also respond to situations, and we do have the epidemiologic capability to immediately jump on something if it's picked up in the laboratory.

CHAIRPERSON OVERTURF: Thank you.

Are there any other questions or comments?

(No response.)

CHAIRPERSON OVERTURF: We have 15 minutes extra for lunch, and the proceedings of the meeting will begin, again, at one o'clock sharp.

So thank you very much.

(Whereupon, at 11:46 a.m., the meeting was recessed for lunch, to reconvene at 1:00 p.m., the same day.)

AFTERNOON SESSION

(1:08 p.m.)

CHAIRPERSON OVERTURF: I'd like to call the meeting to order please.

Good afternoon. We now open our open hearing, and I'll ask Christine whether there's any members in the audience or otherwise who would like to address the committee.

MS. WALSH: As part of the FDA Advisory Committee meeting procedure, we are required to hold an open public hearing for those members of the public who are not on the agenda and would like to make a statement concerning matters pending before the committee.

Dr. Overturf, would you please read the open public hearing statement?

CHAIRPERSON OVERTURF: Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision making. To insure such transparency at the open public hearing session of the Advisory Committee meeting, FDA believes that it is important to understand the context of an individual's presentation.

For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationships that you have with any company or any group that is likely to be impacted by the topic of this meeting.

For example, the financial information may include the company's or group's payment of your travel, lodging, or other expenses in connection with your attendance at this meeting.

Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships.

If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.

MS. WALSH: I have not received any requests at this time. Is there anyone in the room who would like to address the committee?

(No response.)

MS. WALSH: I see no response. Dr. Overturf, I turn the meeting back to you.

CHAIRPERSON OVERTURF: So we will proceed with the agenda and talk about vaccine responses. Dr. Roland Levandowski.

(Pause in proceedings.)

DR. LEVANDOWSKI: Okay. While we're working out the technical end of things here, maybe I'll just give a little bit of background information about the serologic information that I'm going to be discussing.

The serological data that I'm going to be presenting actually will be coming from a number of different centers, and I'm going to try to summarize that information as best I can. I think you all are aware that we share serum panels between several different laboratories and test these same serum panels, each within those laboratories.

The whole point of doing this serological exercise really is to try to see whether the responses to current vaccines confirm what we have found already with the antigenic and genetic characterizations that have been presented this morning already by Nancy Cox.

And whereas you saw for the tables with the ferret sera where there were a few very well characterized ferret sera used to test an enormous number of different antigens, for the serologies we're really kind of reversing that procedure. We're looking at a relatively few antigens that have been selected to be representative of the current circulating strains that seem to be different to us, and we're using a much larger panel of sera.

Now, the sera that we use, those different serum panels that we have available to us are not all identical. They're not all collected the same way. I think that's something that everybody should be aware of also.

Some of these serum panels, you'll see the number of sera on each one of them. Some of these serum panels have been prescreened to select out people who respond well to the vaccine, and that's actually okay. That's good because if there's not a response of any sort, then we won't be able to perform what our primary purpose for this whole exercise is, which is to try to compare the responses between the different antigens.

We're not really trying to say that any of the vaccines used are immunogenic. We're not trying to say anything about the vaccine itself. Really we're focusing on trying to compare what kind of antibody response we see for the vaccine strain as compared to the test antigen.

And so what I'm going to be emphasizing as I have on some previous occasions are the geometric mean titers, the post immunization geometric mean titers. The tables that I'm going to show you are going to be in the traditional sense. They're going to have percent fourfold increases. They're going to have percent of people who are above an arbitrary cutoff point like one to 32 or one to 40, but I'm really going to be focusing on the geometric mean titers and how that relates the test antigen to the vaccine strain.

You're going to see also as I show you these results that there are going to be differences in the absolute titers between labs, and again, that's not something I want you to focus on. What I really want you to see are the differences that are shown between the labs.

And what I can tell you generally when we look at this, although there are differences in the absolute values from each lab, when you do a rank order of the way the different antigens fall out, the test antigens, they come out to be not exactly the same but pretty close to the same in terms of which is recognized as the highest response and second and third and fourth, and so on.

So these are the serum panels that have been provided shown on this slide and the next one, and you'll see here that the vaccine strains used in all of the centers were pretty similar. I just point out some slight differences.

All of the centers using A/New Caledonia and A/Wyoming as the H1N1 and H3N2 strains, but there are differences in the B strain. The vaccine used for preparation of these sera from Australia contain the older B/Victoria-like vaccine antigen, B/Brisbane/32, but the other centers have used more recent B/Yamagata-like vaccine strains, B/Jiangsu/10/2003 or B/Shanghai/361 and that should be 2002.

We also have a couple of panels of sera from children. Nancy Cox mentioned that they had tested two. There's only one that I'll be discussing, one panel from the CDC that I'll be discussing with the following information.

We had even a third panel of sera from children, and I'll be presenting some information from those. So these are not exactly the same two panels that Nancy mentioned, but they are two panels of sera from children that we have to look at.

So jumping right into it, these are the different antigens that are used for the H1 serologies, and all of them are H1N1 viruses. As Nancy mentioned, there really haven't been any H1N2 viruses circulating, but we have a good representation of strains from around the world, from Asia, from North America, from Europe, from Oceania.

And in terms of responses, I'll be showing you some tables that are not all the tables that were available or all of the data, but some that are just representative of results that came out from the different centers.

So here is a table of results for three different serum panels from the United States, from Australia and Japan. These are sera that were done in the lab here at Center for Biologics, and these are the numbers of serum pairs that we had from each of these sites. The antigens that were tested are shown here, and in blue on every one of the slides I'll have the vaccine antigen, and in red I'll point out the ones where there's a 50 percent or greater reduction in the post immunization titers as compared to the vaccine strain.

So from this you see that we tested not only New Caledonia, the vaccine strain, but also A/Florida/4/2004, A/Bangkok/1544/2004, and A/Okinawa/42/2004.

And for the most part the post immunization responses of people who were immunized with A/New Caledonia were very similar for the other strains that were tested. The one exception here is that in this particular serum panel with A/Florida, there was a more than 50 percent reduction against the A/Florida strain as compared to the vaccine.

So those were adults, and now elderly, from the tests that were done at National Institute for Biological Standard and Control, or NIBSC. They're ordered somewhat differently here and different numbers and so on, but again, A/New Caledonia was the vaccine strain. A/Netherlands/128/2004, A/New Caledonia/9/2004, and A/Okinawa/42/2004 were the test antigens.

And by saying that, I'd also point out that we didn't test all of the same antigens in every lab, but there is overlap between antigens that were tested in the different labs. They're not all exactly the same. So we have a somewhat broadening of the numbers that are tested and some similar or some identical strains that are tested in the different labs to give us some idea of where we're standing in terms of the overall results in comparison between labs.

But here, again, you see in the post immunization geometric mean titers for the most part the new strains were well inhibited by antisera that were raised against the vaccine strain. In this case, the one exception, it was for this A/Okinawa/42/2004 H1N1.

So those were adults and elderly. For the pediatric population, the two different groups of children were children who were six to 23 months of age. I don't know exactly what the mean age is, but they were all less than two years old, and another group of children who had a mean age of 21 months with a range of eight to 38 months.

And, again, the antigens that were tested are shown here. For the somewhat older children, we did not really see much of a difference between the vaccine strain and the test antigens, although it's a little bit low for this Florida strain, which is similar to what I showed on the first slide that didn't quite reach this 50 percent reduction

And here these younger children for these two antigens, A/Florida/4/2004 and A/Netherlands/128/2004, there were more than 50 percent reductions, and I'm just going to go right on to a table where we've tried to summarize all of these results looking at the percents that were greater or the reductions in geometric titer that were more than 50 percent.

Why are we emphasizing this? This is a way to try to get some idea of how severe or how the magnitude of the difference. It's a somewhat arbitrary way to do things, but it does give us a way to try to handle the data which are somewhat disparate between the different labs and for the different antigens.

And if you look at this, there are a number of different antigens that were tested. Many of them were tested in more than one lab, and I'm not including some of the information for strains that were just tested in one lab, although I have a little bit here.

But what you see generally is that for all of these H1N1 strains, for the most part on average there was not really a 50 percent reduction either by the total numbers or by the mean averaging the percent reduction for all of the studies all together. Actually they're quite low.

So the difference here wasn't very much, and you can see from the range also that for the most part, very few of these antigens when they were tested were there 50 percent or more reductions.

So now moving on to the H3 strains, A/Wyoming, of course, was the vaccine strain, and there were a whole range of representative current viruses. All of these, I think, can be categorized as California/7/2004, although I'm not entirely sure about the Singapore strain. That might not be truly considered California/7-like, but it's a more recent strain than Wyoming, and if anything, it would be somewhat like Wellington or out farther than Wellington on the genetic dendrogram.

But, again, we have a range of viruses that were tested, and here I'm showing some data from the CDC lab for adults, from the United States, Europe, and Australia. The antigens that I'm showing here, Wyoming is the vaccine strain, and here there were A/California/7/2004, A/Singapore/36/2004, and A/Tennessee/6/2004. Those strains were tested.

And I would call your attention to the fact that for every one of these serum panels and with all three of those California-like antigens, the reductions in post immunization geometric mean titer are really quite obvious. They're very large reductions. It's much more than 50 percent in each one of these cases.

And as you'll see on the next slide, that holds true for the elderly here as well. These are sera that were run at NIBSC for several serum panels. They tested California/7/2004, Singapore/37/2004, Oslo/807/2004, and Shantou/1219/2004. And here, again, you see it stands out very much that there's a big difference between the post immunization responses for the vaccine strain and for these newer California-like antigens, and it's a lot more than a 50 percent reduction.

And, finally, for the H3, here are some data for the pediatric population again, and again, just not to belabor this too much you see that even in these young children or particularly in these young children, you see that there's a big difference between the post immunization titers for the California-like strains and the Wyoming vaccine strain.

For a summary of all of this, in this table I just call your attention to the totals over here. You'll see that almost every one of the serologic tests that was done show that there was a 50 percent or greater reduction in geometric mean titer as they were tested, and the magnitude of this difference was really quite amazing and substantial. All of these are much, much more than 50 percent, and in many instances the percent reduction -- none of the tests were under 50 percent and most of them were almost 100 percent.

So now moving on to Influenza B, this is a little bit more complex. As I mentioned in the vaccine, the Australian vaccine has B/Brisbane/32 as the vaccine strain. The vaccine that we're interested in here in the United States would have contained B/Jiangsu/10/2003. So I'm really not going to be focusing on results that came from the vaccine that was used in Australia in this case.

There were both Victoria-like Influenza B viruses that were tested here. I've listed them as Hong Kong/330-like. That was the last vaccine strain that we used here in the United States.

And there are also a group of viruses that are more like Yamagata/1688 or our current vaccine is Shanghai/361-like. So there's a range of different antigens to look at here.

And, again, this table from data from the CDC in adults for a number of the different serum panels. In this particular instance B/Jilin/20, which is equivalent to B/Jiangsu/10, was used as the test antigen, but that's equivalent to the vaccine.

And I guess I should point out that in this case the antigen was ether treated. Not all of the labs do this, but many of the labs do ether treat the antigen before doing the serologic testing.

There are a number of antigens that were new viruses that were included in the test, including B/Colorado/4/2004, B/Florida/2004, and B/Fujian/2004, and again, all of those are in the same HA lineage as the vaccine, and you can see all of these in each of these instances were very well inhibited by the current vaccine.

The only thing that was not in this particular test was B/Hawaii/13/2004, which is representative of the other HA lineage that's not in the vaccine right now, and there were more than 50 percent reductions here, and this is very consistent with what we have been seeing in the past in the same situation. This is really not new. It's what we have identified in previous years when we've been comparing the responses of vaccine containing one HA lineages to viruses in the other HA lineage.

These are data from the Australian center, an elderly population, and here they were testing -- that should be B/Jiangsu/10. Sorry -- but they were using B/Jiangsu/10 as the equivalent of the vaccine strain, and then tested B/Brisbane/4/2004, B/Victoria/501/2004, and B/Hawaii/13/2004.

And here the results are not quite as clear-cut as in the previous slide. Here in this instance, compared to the vaccine response, we would say that the response for all of these test antigens, including the one that's not from the same HA lineage, the B/Hawaii/13, looked pretty similar.

But in this other panel from Europe, one of the strains that's in the same HA lineage as B/Jiangsu actually gave responses that seemed to be low in this particular center's tests, as did the Hawaii which we would have expected because it's in the other HA lineage.

And then the last in this particular panel of sera here. Again, there was not really a difference seen even between the Hawaii strain which was in the other HA lineage and the vaccine strain.

So these are results from children, these two different groups, and here it's a little bit more difficult to interpret because the vaccine responses to the vaccines themselves were on the low side. The sera tested at CDC using B/Colorado/4, B/Florida/7, B/Fujian/430, Hawaii/13, and the other similar strain to Hawaii, Phitsanulok -- I think I'm getting that right or close -- even though these results are low, it looks like for all of the strains that are equivalent to the vaccine there's a similar response, whereas it was a lower response for the non-vaccine antigens.

For these children, these slightly older children, however, although we had some response to the vaccine, we really weren't able to detect response to any of the antigens, including those that are similar to what's in the vaccine.

So trying to put this into a summary slide, looking at the 50 percent reduction between the different sera, I'd say overall it still sort of holds up that the strains that are in the HA lineage, not currently in the vaccine, Hawaii/13 and Phitsanulok/2053, that there are some reductions, but you know, they're not really high magnitude all the time. It's not really consistent between all the centers and all of the tests that there's a 50 percent or greater reduction, and in only about half of them, if I've done the math there right -- and I might have made a mistake, but it looks okay as I do it quickly in my head -- although it did reach like 50 percent for mean, there was a huge range here for the results.

For the viruses that are in the same HA lineage as the vaccine strain, here again there's some data that suggests that there may be some reductions for some of these viruses like Brisbane/4, and on the other hand, other data suggest that for new, more recently circulating strains similar to the B/Shanghai/361 that there's not so much reduction.

But, again, you see there's a wide range here in the results for these different serum panels, and the highest percent reduction actually would come from the children. So if I excluded those, you would see for adults and elderly that it looks like even less a difference between those more recently circulating strains that are like the vaccine strain, B/Shanghai/361.

So in summary, I guess what I would say overall is the studies were done with sera that were collected after immunization from a number of centers with a number of different vaccine products, and at this point it looks like the representative H1N1 viruses are very well inhibited by the current vaccine.

On the other hand, it looks very clear that the A/California/7/2004-like viruses are poorly inhibited compared to the current vaccine.

For the B strains, the B/Shanghai/361/2002-like viruses seem to be reasonably well inhibited, but those that are in the other HA lineage as represented by B/Hong Kong/330 are less well inhibited than the vaccine strains.

And I will stop there and see if there are any questions.

CHAIRPERSON OVERTURF: Are there questions for Dr. Levandowski? Dr. McInnes.

DR. McINNES: Roland, the infant data one presumes that some of these children are receiving their first dose of flu vaccine and some are maybe second, maybe, I guess, second. So what was the timing around the sera collection post dose, and for those who are receiving it for the first time, was this collected post two-dose?

DR. LEVANDOWSKI: Okay. I should have made that clear. Both of those serum panels were from children who had received two doses of vaccine. It was post the second dose of vaccine. I don't have information. I can't tell you at this point whether any of those children had been immunized previously or infected.

Some of the older children in the second serum panel more than likely were either immunized or infected. The younger children though you'd have to ‑- since these were studies done some time in the summer to early fall, late summer to early fall, this would have been happening in between flu seasons, but I can't really totally tell you how all of their exposure history fits in because I don't have those data.

And I think, you know, still these are really very young children, and it gives you an idea of what you would expect in a population of very young children.

CHAIRPERSON OVERTURF: Yes, Dr. Karron.

DR. KARRON: Two questions. the first is I actually first want to commend you on having these pediatric panels of sera. I think a couple of years ago we were discussing how important that was, and I think it's very useful to have them.

I guess my first question is now that we have two types of vaccines in this country -- we have an activated and we have live attenuated -- would it be useful to have data from individuals immunized with live attenuated vaccine to have that as a comparator, obviously not in the children under five, but you could have school age children and adults 18 to 50. I think, particularly when these issues of antigenic drift come up, that might be a useful comparator to have.

I'll let you say something, and then I have a second question.

DR. LEVANDOWSKI: I think it would be interesting, but I just would like to point out that the whole point of this, again, is to compare. It's not so much to say whether the vaccine is immunogenic or not. I don't think with the strategy that we're using here, I don't think we can really make any comment on that. So really we're trying to focus on getting a comparison between the different antigens, the vaccine antigens versus the newly circulating strains.

Having said that, I think it would be very interesting to have access to more sera all together from different populations, and it might make sense to explore that and see if that gives us any additional data compared to what we get from inactivated vaccines. I don't see any reason it wouldn't be useful information really.

DR. KARRON: And I guess the follow-up question really relates to something that Dr. Couch was talking about earlier, but you know, I guess my question is really we see very poor responses in very young children to the B strain. Is that consistent with what we've seen before, data that we have before?

And if so, I mean, I'm almost wondering whether it matters which B strain we choose because responses are poor. You know, it's hard to look at reductions when your geometric mean titer is ten.

DR. LEVANDOWSKI: Right. Well, those numbers are small, and I don't know how much of that is technical artifact and how much of that is significant in terms of what protection would be. Most people are used to thinking about titers of one to 40 or greater being representative of protection, but actually the data are fairly sparse in terms of indicating that there's a correlation between the titer and efficacy.

We do know the higher the titer of antibodies the more likely you are to be protected, but I don't think we know for sure that, you know, even at lower titers there might not be some level of protection. So I would hesitate to comment on what the meaning of those titers that we saw there are.

But we have for children, we have seen titers that have been very low. Usually it's for the H3 strains that we've seen that in previous years, but they vary quite widely, and when I said there may be some technical artifact, I guess we occasionally see differences that may be related to some aspect of the virus itself in terms of their avidity for binding to red cells, other things that are going on in the system.

But I'm rambling around here and not trying to answer your question directly because I don't think there is a direct answer to it.

DR. KARRON: But I guess my question is that I'd have to go back and look at these data, but the titers to be were not uniformly low across populations. You know, if you thought it was a technical issue, then presumably whether you tested adults or elderly or children, you would have uniformly lower titers.

My question is given how low the titers are to the B strain in these very young children, is that inconsistent with what we've seen before or is that consistent with what we've seen before in TIB in very young kids?

DR. LEVANDOWSKI: Well, I did answer that part directly. I think we've seen low titers for different antigens, not necessarily for B, but we have seen low titers for different antigens, not necessarily for B, but we've seen low titers with different antigens in very young children, and I don't think we always would predict what that would be. I think we've seen differences between children and adults for, you know, the same antigen and we also have seen if you look at a series of antigens like H3, for example. It seems that some of them are just more inherently immunogenic than others.

CHAIRPERSON OVERTURF: Dr. Couch.

DR. COUCH: Just for a moment, a lot of what we're looking at here is very test dependent, as Roland pointed out, and the split product is used for the B which, in order to give it the kind of sensitivity that will permit these spread out comparisons, and what he is giving us is really relative, not absolute, and when we're using the A strains, we're using it there in an entirely different HI test.

But in general, Influenza B responsive is when you do one type of assay, you see, which carries, as we know, the same kind of sensitivity. Neutralization tests are generally speaking not quite as good as they are to A, but that's all fairly relative. So it's difficult to compare B responses to A responses when we're looking at them here because it's so person and so test dependent. But he's going with relative findings for strain selection, you know, this percent of GMT, and then the same is true for the equal to or greater than 40, you know. What is the validation of a one-to-40 with a split product versus a one-to-40 with a whole virus antigen when you're doing Influenza A, and certainly a major difference is if you do B with whole versus the split product.

So it's all sort of a test-test nuance in that. So one of the reasons that I think that really caught my eye that you didn't comment on, Roland, was page 37 in the CDC surveillance where the pediatric population age is five to eight, 274 children by Kathy Neuzil, and only ten percent of them had a rise to be Hong Kong. There's a major disparity in that age group, both of which has inactivated and live recommended for.

You really would have liked to have seen the live to see how it does in that comparison in that regard, but as we go back to Roland's point, that doesn't influence our selection here. That's the kind of thing that the ACIP would want to be looking at when they talk about relative criteria for recommendations, but that data would have really been interesting to add to that one.

I only had one other just minor question, and I couldn't pick it up here, but in the individuals it looked like the elderly in the U.S. and Japan didn't respond as well as the elderly in Australia and Europe. That's got to be partly population dependent in testing. What do you --

DR. LEVANDOWSKI: No, no, no, no, please. I said at the very beginning that you should be aware that some of these serum panels have been prescreened to select out higher responders to begin with. Not all of these serum panels have been so prescreened.

DR. COUCH: I missed that.

DR. LEVANDOWSKI: So I think you have to be very careful. There are differences in how the sera were handled before they got to the individual labs to be tested. So I wouldn't draw any conclusions from that. I'd try to focus on the relative responses between the antigens.

CHAIRPERSON OVERTURF: Dr. Monto.

DR. MONTO: Just a further comment about B responses in young children. They don't even respond well to HI ether treated antigens post infection because we've had, when we were back doing our community studies in Tecumseh years ago, we could show that children under five were heavily infected when there was a B outbreak, but if you look just in terms of rises in antibody titer, a fourfold rise in titer, it looked like they were spared.

So there basically is a problem in just immune responses and young children measured by HI, we don't know how that translates to other things.

DR. COUCH: Just a quickie for general information. That same thing is true for the neutralization test, and I think our test is fairly sensitive. The same finding.

CHAIRPERSON OVERTURF: Yes, Dr. Dowdle.

DR. DOWDLE: Like Bob, when I saw that chart, the Australia sera were looking so good that I was ready to move to Australia, but --

(Laughter.)

DR. DOWDLE: So thank you very much for your explanation, Roland.

But this is just to say that I think that the evolution in sort of the use of human sera to interpret the strain differences has been really quite helpful and is moving forward quite a bit.

I wonder if there's further discussion on how this can be standardized even more because particularly I think there were core viruses, but could there be more core viruses, it might be used. In addition, the labs could use whatever other strains that they want, but I'm sure those discussions must be taking place.

DR. LEVANDOWSKI: You're right. There are discussions going on all the time about how to try to do this better. It's very difficult because of the logistics of transport of all of these materials, and although I did not get into that as part of the presentation, there are differences in the antigens, too, because in order to do the testing, it's done with a pool of virus that is grown usually in the laboratory where this is being done, and so there may be some subtle differences from passage levels in the different laboratories that have an impact on that.

But the logistic part of it is very tricky, and I'm not emphasizing that, but of course, these California viruses have only really been recognized as a separate group of interests since about mid-January. So to get these shipped around to the different labs and to be able to do the testing on those has been, you know, a challenge, quite honestly.

But in the ideal world, if we had more time and it wasn't flu, then maybe we could get it right.

CHAIRPERSON OVERTURF: Dr. Cox.

DR. COX: If I could just add a couple of more comments, we really have put a lot of effort into improving the testing and the consistency of choice of antigens and so on. Every year there's some different factor that's a problem, and this year, in particular, we had a rather slow start to the influenza season not only in North America, but also in Europe, and this year, as in other recent years, we have had two preparatory WHO conference calls, part of which are dedicated to talking about what strains we should include in the human serologies and where the serum panels are in terms of their availability for shipment and so on.

So there is really a lot of effort that goes into this. It's just a matter of, as Roland said, getting the viruses out there and then seeing what viruses grow best and how many you can actually put in test given the limited amount of antiserum that you actually have.

CHAIRPERSON OVERTURF: Dr. Eickhoff.

DR. EICKHOFF: Just a general comment. I think this year particularly illustrates the importance of having human serum panels available for study because as Nancy pointed out to me earlier in response to one of my questions, the difference between A/Fujian strain and the A/California strain wasn't really that clear in the ferret antisera data, but in the human sera data it's apparent that there's a real problem.

CHAIRPERSON OVERTURF: If there's no more comments, we can go ahead and progress to the availability of strain reagents. Dr. Ye.

DR. YE: Thanks.

In his presentation, Dr. Levandowski described the immune response to the current influenza virus vaccine, and I'm going to present this status of candidate virus vaccine that is used for production as a viral vaccine and the potency reagent that had to be used for standardize the antigen of vaccine product.

Influenza vaccine contains three antigens: H1N1, H3N2, and B virus. Current vaccine for H1N1 is New Caledonia/20/99. IVR/116 is a reassortant between New Caledonia/20/99 and A/Puerto Rico/8 or PR/8 high growth viruses in the lab.

This reassortant grows quite well in eggs. Currently we do not have antigenically divergent strains available for this particular subtype.

Wyoming is the current vaccine reassortant. NYMC-X-147, NYMC-X-149, RVR-134 are the reassortants between Wyoming and PR/8. All of them grow pretty well in eggs.

The candidate strain for H3N2 is a little bit complex. The first one is the A/Wellington/01/2004. There are three reassortants. RESVIR-20, IVR-139, NYMC-X-155 are the reassortants between Wellington/01/2004 and PR/8. All of them grow reasonably well in eggs.

And as Roland mentioned, Wellington/01/2004 is a strain recommended for the Southern Hemisphere in the 2005 season.

The next one is A/California/7/2004-like strain. Currently we do not have high growth reassortant for this particular strain. Although IVR-140 is a resortant between Singapore/37/04 with PR/8, this particular reassortant does not grow very well in eggs, neither grow well in eggs nor ideal strain represent A/California/7/2004.

However, we have at least six circulating viruses that represent A/California/07/2004. All of them are egg isolates. One of these strains, A/New York/355/2004, are the most among the six -- grow reasonably well in eggs, and this strain may be ideal strain to be used for generating high growth reassortant between this particular strain and PR/8, and the rest of them do not grow very well in eggs.

I'm putting up this slide to give you the sense of the possibility we can generate high growth virus for this particular subtype. There are at least the six laboratories work wholeheartedly to generate high growth reassortant for A/California/7/2004. There are three in the United States, one from CBER, CDC, New York Medical College, and three from overseas. They are NIBSC from New York (sic), CSL from Australia, National Institute, Infected Disease, from Japan, and hopefully, by mid-March we can have this high growth reassortant for A/California/704.

The reason I'm saying that is I mentioned one strain grows reasonably well. Then the possibility we have with high growth reassortant for this particular strain would be likely.

The current Influenza B strain -- current vaccine strain is B/Shanghai/331/2004-like. There are three strains: B/Shanghai/361/04 itself, B/Jilin/20/2003 and B/Jiangsu/10/2003. All of them grow moderately, at a moderate rate in eggs.

And currently we do not have a new antigenically divergent strain available for this B strain.

Now we move on to the potency reagents. We have antisera and antigen for standardization of the vaccine for both H1N1, which is New Caledonia/20/99 and Wyoming/03/2003. If the new strain is going to be chosen and the particular reagents needed to be prepared so that they wouldn't be available before May.

We also have antisera and antigen for both HA lineages, one for Shanghai/361/04, which is a Jiangsu/10/2003, and also we have antisera and antigen for Yamagata lineages, which are B/Hong Kong/330/2201 and Hong Kong/1434/2002 and B/Shandong/7/97.

Again, if the new strain are choosing the reagents, it would be available in May in the earliest.

Thank you.

CHAIRPERSON OVERTURF: Are there questions? Yes.

DR. BUCHER: I just wanted to mention that we had submitted a high yield B reassortant for Jiangsu, NYMC -- I'm from New York Medical College -- NYMC-BX-7. So it doesn't -- in our hands it grows about twice as well as the B/Jiangsu. So just to let people know that that's available. We did submit it to the CDC. It was also sent to your lab.

CHAIRPERSON OVERTURF: Maybe you can clarify something for a bacteriologist. What is it that you do to make the -- is it all trial and error or are there genetic or molecular elements known for egg growth?

DR. BUCHER: No, it isn't that random. We do use, although we do take advantage of a lot of possibilities here -- Ed Kilbourne had the main lab, and we're continuing on with his assistant as well and his reagents, and we've generated improved selection reagents.

So he developed the system for As, and that's what people have been using, using an old, well adapted egg strain as the donor, which is A/PR/8/34. So it's been in an egg since 1934.

For the Bs now, we -- and that's how we made X-147 and X-149, which were mentioned, and now we're working on the high yield A/New York/55.

For the B strains, we selected B/Lee/40, which has been in egg since 1940, and as I said, it doesn't give us the tremendous enhancement that we see for the As, but the Bs generally -- they generally are growing better than the new isolates of the A strains, but we did see an increase in yield.

So we know that it has the eight gene segments. Of course, we have to have the two from the current strain, the strain that's circulating currently. So we have to have the hemagglutinin and neuraminidase from the B/Jiangsu.

We know we have the M gene from B/Lee/40, and we're in the process of analyzing the rest of the genes.

CHAIRPERSON OVERTURF: Thanks.

Are there any other questions or comments?

(No response.)

CHAIRPERSON OVERTURF: Thank you.

We now have time for the comments from the manufacturers, and that's going to be presented by Albert Thomas of Sanofi Pasteur.

MR. THOMAS: Good afternoon. My name is Albert Thomas. I'm the Director of Viral Manufacturing for Sanofi Pasteur, the vaccine maker previously known as Aventis Pasteur.

I'd first like to thank the committee for the opportunity to present today, and I'd like to begin by talking about some of the critical factors that are involved with influenza vaccine supply and how the strain selection process can impact each of those factors.

First of all is the growth potential of the seed virus. Obviously there are many factors that can impact the number of doses of influenza vaccine that can be produced, such as the overall capacity that is available to each manufacturer, the average yield of all three of the monovalent strains currently in that formulation, but many times it is the yield of basically the lowest performer, the least productive monovalent strain that will ultimately impact or determine the number of doses that can be produced in that given season.

You may be successful in producing 40 million doses of your H1N1 monovalent component, 40 million doses of your B component, but if you can only produce 20 million doses of your H3N2 component, you will ultimately only be able to distribute 20 million doses of trivalent influenza vaccine.

Probably the most critical constraint is time. Given that the timing for influenza vaccine manufacturing is limited at the beginning by the timing of strain selection and is then limited at the end by the necessity to be able to manufacture, release, distribute, and administer the vaccine prior to the onset of the influenza season, your actual time is very limited, in which case you can actually manufacture the monovalent components and then formulate the trivalent vaccine, fill, package, release, and ultimately distribute.

Also, please keep in mind that the working seeds typically require about four weeks for development and release prior to them being able to be utilized in large scale manufacturing, and usually that four weeks is based upon the time that we initially receive the seed candidate.

The potency test reagents are also a factor that must be taken into account. Given that each monovalent component, the potency of that component or the amount of hemagglutinin must first be determined prior to formulation of the trivalent vaccine, and that's done via single radial immunodiffusion, which requires a strain specific reference antigen and antiserum.

Those two components, the potency test reagents must first be manufactured and calibrated or standardized for each new strain prior to the formulation. That can be anywhere from, say, an eight-week process to about a 12-week process, depending on when the seed candidate is first involved or first available.

The time line that I've got listed here has several assumptions built into it. I'd like to first describe those. This is based upon assuming that there is one strain change from one year to the current year, and the new strain here is listed as B or is listed as the blue component.

The way this is broken down is really first the upper half here is related to the production of the individual monovalent components. The lower half here is involving the formulation of the bulk trivalent vaccine, the ultimate filling, packaging, and then distribution of that vaccine.

Again, as I mentioned before, the time constraints are we are required or limited by the time that we need to begin distributing the vaccine, again, typically the August to the, say, early November timing. This past season was a bit of an exception. Obviously we've been able to extend the time, in which case we can distribute the vaccine.

This basically ends or defines when you need to begin distributing, and again, the actual beginning of the strain selection really determines when you can begin production.

Now, something you might notice here is is I've got this arrow here listed as essentially the timing that the strains are announced, basically the middle of February. If you're not familiar with influenza manufacturing, you may see that manufacturing is already underway.

Some manufacturers may choose to manufacture one of the strains at risk, and that risk is the fact that that strain may not actually be included in the vaccine the coming year. The reason why manufacturers may choose to do that is, again, because the time here is very limited, in which case you can actually produce those monovalent strains and be able to distribute that vaccine in time.

So, again, here, you know, assuming a model of one strain change for the year, assuming that manufacturers would have already had some of that production underway, manufacturers would typically be looking to begin production at that second strain very soon after the strain announcement. Basically the concern with the fact that they want to build too much of one strain without ultimately knowing the yield of all three strains that will be in the vaccine.

So typically following the strain selection, you know, assuming a working seat is available, manufacture of the second monovalent strain would begin. Once a working seed would be available for the strain, manufacture of that would also commence.

That depends now upon what is the timing or the availability of the reassortant. This time line here would assume approximately a mid-March availability, which again for this year may be aggressive, but at least it's something to start with.

Again, approximately the four weeks necessary for each manufacturer to produce the working seed to use in large scale production from that reassortant, and then the beginning production of that third strain.

Now, what then happens in parallel with that is the production and standardization of the potency reagents, again, the reference antigen and the antiserum. Until these reagents are available, the final or true yield to this third strain is not known. Once the reagents are available, manufacturers now have a very good idea of what their yields of the three strains are. So they then begin the strain balancing process here.

Again, each of the three strains are produced independently, and depending on what the yield is of each strain, manufacturers will emphasize one or the other with the ultimate goal at this time to have an equal number of doses produced of each of the three monovalent strain components.

And, again, once the potency reagents are available, the formulation of the trivalent vaccine can begin, which then would be followed by the filling, packaging, and ultimate distribution of the vaccine.

Current manufacturing status, as I alluded to in the previous time line, for some manufacturers the H1N1, New Caledonia, the 20/1999 manufacturing is underway, and again, that was initiated at risk, again, of that strain potentially not being selected to be in the vaccine formulation for the 2005 and 2006.

Again, the reason why manufacturers decide to produce that risk is, again, because the manufacturing time point here is very limited. We have a limited time. That's really our most critical constraint that we have to work with, and by manufacturing at risk with some knowledge of the surveillance data that's available allows us the opportunity to produce some of that product prior to that very critical time frame.

The H3N2 and B strain, obviously we're awaiting the strain selection process and are currently evaluating any new potential seed candidates that we're receiving, whether they're the A/California type of other type strains.

Basically, in conclusion, it's really cooperation. I think it's very clearly necessary for the successful influenza vaccine production and supply, and that really comes from all parties involved in the overall process. I would say the timely selection of the appropriate antigens, both consideration of the antigenic match, as well as the potential growth potential for that strain; the availability of the height of the seed viruses, especially the high growth reassortants.

Again, because time is so limited, the best yielding strain you have to produce each day will maximize the number of doses that can actually be distributed. And something that worked, I believe, very well last year and I think is a good representation of the availability or the greater availability of egg isolates is the opportunity for manufacturers to evaluate the growth characteristics of strains that are antigenically similar but may have very different growth characteristics and large scale production.

Again, the example last year, evaluation and selection of the B/Shanghai/361/2002 light strain. Manufacturers were given several strains to pick from, and actually it was the Jiangsu strain here which was received last but had by far the best growth potential, and that's what was utilized in production.

Again, just showing the fact that antigenically similar strains can have very different growth characteristics, especially in large-scale production, and the availability to have additional egg isolates is a key factor for a lot of the manufacturers to pick potentially what would be significantly a greater yielding strain which ultimately will yield many more doses for the marketplace.

And, again, the availability of the potency test reagents and the time frame or the time required to produce those must be taken into account in the overall campaign.

CHAIRPERSON OVERTURF: Thank you.

Are there questions for Mr. Thomas? Yes.

DR. SCHWARTZ: I'd like to ask you a question about your last point about the B/Shanghai-like strains. Two questions. One is: did all of the manufacturers come to the same conclusion or will one strain or a different strain grow better with one particular manufacturer's processor, a different manufacturer's process?

Secondly, I'm wondering if when those three different strains were supplied to industry whether there was an indication that the Jiangsu would be the best growing of those three strains or whether it was something that you had to test out in your manufacturing system before that recognition came about.

MR. THOMAS: Sure. Could I possibly defer the first question to someone on the committee? Again, I'm not -- typically each manufacturer is not involved with what the other manufacturers are selecting. I believe we all chose that same strain, but I'm not sure if possibly anyone from CBER would like to talk about that. Yeah.

CHAIRPERSON OVERTURF: Roland.

DR. LEVANDOWSKI: We've had two different strains based on preferences for manufacturers in past years, and I don't remember what year it was, but it was in the late 1990s we had -- actually it was the Hong Kong/330/2001 and the Hong Kong/1434/2001. We had different manufacturers producing different strains, and that went even further because in Europe the strain that was being used was the B/Shanghai/7/97 strain. So all three of those strains that Zhiping mentioned in his talk were used by different manufacturers based on what they saw as what was best for them.

It's a little bit stressful for the system to have that happening, but it can be accommodated.

MR. THOMAS: And then answering the second part of the question, I believe that's pretty much each manufacturer will probably come to that decision independently. I mean, there is data usually given to us when the seed candidate arrives, but eventually based on what each of our -- whatever process each manufacturer would follow to develop and evaluate that seed candidate, if it would be a good candidate for their process.

DR. SCHWARTZ: So that you actually tested all three of those before selecting that particular strain?

MR. THOMAS: Oh, yes, yeah, tested them in multiple passages. Sort of it's when the race begins. When you start receiving the seed candidates you begin very aggressively looking at each of these and try to quickly narrow the list down for those that you would continue to evaluate.

CHAIRPERSON OVERTURF: Kathleen.

MS. COELINGH: It's Kathleen Coelingh from Medimmune.

And just so people know, the live attenuated has the B/Jilin strain as the B component last year, and because we use a reassortant, the early growth of the wild-type strains is not necessarily a good indicator of how our reassortant will grown, but at the same time, I would agree with what Albert said. It's important to get these wild-type strains in early because we make our own reassortant. So we're sort of on the same time line with CDC and the FDA and Doris Bucher and so forth.

So it's the same. It's true we need to get these strains in early, but for different reasons.

CHAIRPERSON OVERTURF: Dr. Couch, did you -- oh, I'm sorry. Go ahead.

DR. HJORTH: Richard Hjorth from Sanofi Pasteur.

I'd just like to also say about yields that I don't believe CDC or CBER are that focused on assays that can really pick up differences between yields, such as the manufacturers are. I think Nancy has mentioned that before. You don't focus on early assays, especially, before the potency reagents are available.

So I think the manufacturers are probably in the best position to evaluate the yields coming out in the allantoic fluid, not that we necessarily run them all the way through our process, but we look at the allantoic fluid in a more quantitative way.

CHAIRPERSON OVERTURF: Dr. Couch, you had a comment?

DR. COUCH: Yeah, I had a question that's along that same line, and if you would confirm or perhaps deny the impression that many of us have that while we know you can balance a number of eggs you put in one strain or another, in terms of the actual yield of a single strain, is it true that the B is most commonly your limiting antigen?

MR. THOMAS: Historically, yes, the B strain has been limiting.

DR. COUCH: And that need for a high yield that Dr. Bucher referred to that's being worked on right now still carries a high priority that it has carried for a number of years now.

CHAIRPERSON OVERTURF: Comment from the floor?

DR. SUN: This is Wellington Sun from Walter Reed.

One question I had based on your time line, it's very stressful, and you said that manufacturers manufacture the first strain at risk before the selection. So what is the impact if the selection is the wrong one, and how often has that happened?

MR. THOMAS: Well, probably the first impact is probably a different person standing up here next year --

(Laughter.)

MR. THOMAS: -- based on the selection.

The second is I would say now that's manufacturing time that you just lost. So obviously you try to make the best educated decision to pick that strain as to which one you think is the greatest probability of being in the vaccine for the following year.

But, again, because your overall time is fixed and you're working within the time constraint of strain selection to really when do I need to stop production to meet the last available time to distribute vaccine. That time is fixed. So you try to find any way to potentially build up a bit of a hedge if you have a potential low yielder. Right now the yield, say, of the new strain is unknown. So by beginning that production, that gives you a bit of a cover factor knowing that you could potentially build up enough doses of the other strains in time.

But, again, obviously it is a decision made at a business risk of that product, potentially not being able to use in the vaccine.

CHAIRPERSON OVERTURF: Yes.

DR. FARLEY: If I remember correctly, the California strain is at an earlier phase of development from the previous presentation. Can you comment on concerns from a manufacturing standpoint about getting a California-like strain up and ready and going in the time frame this year?

MR. THOMAS: I think that's probably similar to -- I'll give you the analogy of last year ‑- the B strain. Again, we had two strain changes last year, both the H3N2 and the B strain, and it was actually the -- I believe the B/Jiangsu seed candidate we did not receive, I think, until some time in March, which means that we had very little time to pass that, to develop a seed candidate ready for large scale production.

Now, the issue is now currently with the A/California type, assuming that is what is selected for the vaccine. It's really how successful this process is going to be and how quickly a reassortant with reasonable growth potential can be completed, released and given to the manufacturers in time. You know, it's kind of obviously that's a strain that we're just seeing very quickly show up, and we're just beginning to receive candidates for that. It's really how quickly can that reassortant be put together and a working seed developed, and what will ultimately be the yield of that.

Probably no different than in previous years in which case, you know, a strain shows up near the middle or end of the season and it appears to be the right match for the next year's vaccine and how quickly can manufacturers put that into production.

CHAIRPERSON OVERTURF: Yes, Dr. Schwartz.

DR. SCHWARTZ: And just to follow up on that, in the previous presentation, it was mentioned that there were six different laboratories working on the reassortants for the A/California strains. Is it possible to tell us are they working on different strains or is it all the New York/55 or are there different California-like strains that are being investigated by these laboratories?

DR. YE: They are working on all six strains. So their own preference is to see which strain grows well in their hand. It's early in the development of the candidate strain.

CHAIRPERSON OVERTURF: Dr. Cox.

DR. COX: Just as a matter of information I'd like to mention that at CDC we are not using the traditional or the classical reassortment techniques, bur rather we have decided that in this particular instance we're going to actually use reverse genetics to attempt to make six, two reassortants for California and for one other California-like strain and we're choosing based on the knowledge we have about those particular viruses.

I think a number of other labs will be using the New York/55 because it does appear to grow better. So there will be some consistency in that, but you know, there will be a lot of effort and some diversity, I'm sure, as Zhiping has said in the candidates that eventually come out.

CHAIRPERSON OVERTURF: Dr. Markovitz.

DR. MARKOVITZ: Nancy, I'm happy to hear you're going to use reverse genetics, but I have a question. I thought with reverse genetics you have to go through a cell line before you go into eggs, and then that brings up all of the problems of certification of the cell line. Is that incorrect?

DR. COX: That is not incorrect, but we will be using certified cells, certified virile cells in laboratory conditions as close to GLP condition as it's possible to maintain. So we'll be making sure we have introduction of no TSCs and so on, and we're documenting everything so that we believe that the strains that will be made at CDC will be suitable than to go into eggs for production should the necessity arise.

DR. MARKOVITZ: Yeah, that sounds like a very good plan. I'm just curious in terms of who decides that a cell line is okay? I'm not speaking against that in any way. I think that's great, but I'm curious. Who gives you the green light to do that?

DR. COX: These are cells that have been certified through a long process of testing by one of the vaccine manufacturers, and they have provided those cells to us. FDA has the dossier on these cells, and so it's all very well contained and controlled.

DR. MARKOVITZ: So it's an FDA decision that the cell lines are okay?

DR. COUCH: But that approval is FDA, is it not?

DR. BAYLOR: This is Norman Baylor, FDA.

It's not an approval, but when we get information and, say, at a drug master file and what have you, we can review the characterization of the cell line, and then we can give that a green light. We don't approve, quote, unquote, cell lines, but we can evaluate whether they've been adequately characterized to our satisfaction to make products.

DR. COUCH: Just a clarification. I think what we're understanding is that you don't approve a cell line. You approve a vaccine.

DR. BAYLOR: Correct.

DR. COUCH: If it's made in a substrate of a particular type, that's a part of the review process.

DR. BAYLOR: Correct.

DR. COUCH: And that's what would be going on with this one also. So it's still an FDA review because the vaccine has to be approved if it had that substrate in its preparation.

DR. MARKOVITZ: Right, but I'm wondering who at FDA. I mean who finally says, "Yes, it's okay to do this"? I'm a little perplexed. It sounds very good. I'm not in any way against this. I think it's a great idea, but I just want to know who finally says yes.

DR. BAYLOR: CBER's office. if it's a vaccine, CBER's Office of Vaccines.

CHAIRPERSON OVERTURF: That's fairly nonspecific.

(Laughter.)

CHAIRPERSON OVERTURF: Robert. Yes, Dr. LaRussa.

DR. LaRUSSA: Just a clarification. Let's assume a worst case scenario, that you weren't able to make high growth reassortments from the A/California-like strains. Are you saying that the reassortment that's available for A/Singapore 37/04 would not be appropriate and would not work as a vaccine strain? Because that's the one you do have a reassortment.

DR. COX: Roland warned me that I would need to go first on this.

I don't know if you recall, but when I showed on one of my H3HI tables, what we found when we put the Singapore/37 strain into ferrets was that the ferret serum generated did not cover the current strains quite as well. We went back and very carefully went through the genetic data to see if we could find a reason for why this might be true, and we found that there were two potentially significant genetic changes right in the same neighborhood in the HA molecule, and although this neighborhood, this area of the HA molecule hasn't typically been associated with antigenic changes or major antigenic changes in viruses, there is right adjacent to an isoleucine to methionine change, and changes to methionine are fairly infrequent I would say overall.

We also see loss of a potential glycosylation site, which certainly could affect anogenicity. So these are changes relative to the majority of the California-like viruses.

So while the Singapore strain shares the change at amino acid/145, which we believe is critical, it also contains these two other changes which are not part of the consensus and may cause a slightly different antibody response in the recipient.

So we would view the Singapore/37 strain as less than idea.

DR. LaRUSSA: So that's a very gracious way of saying no.

(Laughter.)

CHAIRPERSON OVERTURF: Yes, Dr. Karron.

DR. KARRON: Just as a follow-up question to that, of all the other A/California-like strains, are any of those like the Singapore or none of the other California-like strains have those changes?

DR. COX: Correct. None of them do.

CHAIRPERSON OVERTURF: Yes, from the floor.

DR. BUCHER: I just wanted to raise the issue of our high yield A/Wellington strain, which was NYMC-X-155, which I had understood from the sequencing was a step ahead of the A/Wellington and I would presume toward the A/California type.

We had discussed this at the previous meeting, and there was a mention of that going into ferrets. So I wondered if that had been looked at any further. That is the highest yielding reassortant I think we've ever seen. So I wonder if that could be under consideration, as is the A/Singapore as fail safe.

DR. COX: If you can give me two minutes to look at our data again so that I'm sure that I give you the correct answer, maybe we could go on to another question and then I'll have the answer to that.

CHAIRPERSON OVERTURF: A question from the floor?

AUDIENCE MEMBER: Assuming all reassortants grew poorly, can a wild-type virus be used in a vaccine?

CHAIRPERSON OVERTURF: I'm sorry. Could you repeat that? It was not quite clear through the mic.

AUDIENCE MEMBER: Oh, yeah, that's regarding discussion on the H3N2 virus. So if the reassortants obtained not growing as well as the wild-type virus or even poorer, then can the wild-type virus originally isolated be used in the vaccines?

CHAIRPERSON OVERTURF: Yes, Roland.

DR. LEVANDOWSKI: The simple answer is yes. The wild-type virus can always be used in the vaccine, and that's what we use typically for Influenza B, and as you recall, even for Influenza A, the A/Taiwan/186 strain was a wild-type virus which was not a reassortant.

So, yes, we would make use of whatever seemed to be appropriate. If the question is related to the California-like strains, it doesn't seem very likely to me that that's going to be the case because the California-like strains as Zhiping showed are generally low to moderate growth, if any, so that it's unlikely that those would be suitable for large scale manufacturing, at least not within this decade.

DR. BUCHER: Since I'm not a manufacturer, in our hands the New York/55 is growing quite well. We've been passing wild-type along with our reassortant passes, and it's gaining. So it may not be so hopeless to use that as wild-type.

DR. LEVANDOWSKI: Thanks for that comment, but I wasn't saying it's hopeless. I'm just saying that generally the wild-type strains do not grow better than the reassortants. That's the whole point of doing the reassorting in the first place.

CHAIRPERSON OVERTURF: Dr. Cox.

DR. COX: Yes. So now I'm prepared to answer Doris' question with 100 percent accuracy. We had taken your new Wellington virus, your new Wellington reassortant, and we put it into ferrets and did the cross-test, and most unfortunately -- well, the antiserum that we generated had a very, very high homologous titer, but when you look at the relative titers to the current strains, it covers many of the current strains no better than the Wellington wild-type antiserum did.

CHAIRPERSON OVERTURF: Dr. Schwartz.

DR. SCHWARTZ: I just have one other question for you, Nancy, and that is that we talked this morning about some potential limitations to the commercial use of a vaccine that's made using reverse genetic reference strain, and so I'm just wondering if you can just describe the decision that you and your laboratory made to use reverse genetics for this California-like strain rather than to use traditional reassorting methodologies.

DR. COX: Well, it was really a very -- it's a forward looking strategy really. We do feel that it's very important for federal laboratories to be practicing reverse genetics on a regular basis for generation of vaccine strains, even knowing that there are some potential intellectual property issues associated with them.

We made this decision at a time when the North Dakota virus, which was one of our candidates, was growing extremely poorly, and we were getting reports back that California was also not performing well.

So we decided that we would go ahead and really work toward perfecting the system and moving viruses through making reverse genetics modified strains as quickly as possible just in case we were unable to generate one using classical reassortment.

It looks now like some of the additional egg isolates that have come on line since then are much more promising. So it may not be necessary, but nevertheless it's good practice in many senses of the word, and I think we'll be doing this in the future as well.

DR. LEVANDOWSKI: Yes, sir.

DR. COUCH: A very minor question. I would have guessed at least two and possibly three other sites that are also at least trying reverse genetics. Would you agree with that?

DR. COX: I don't believe so. I believe NIBSC is --

DR. COUCH: Leaving it alone?

DR. COX: Going to use classical reassortment.

DR. COUCH: How about Japan?

DR. COX: They didn't make any mention of using reverse genetics. At the WHO meeting I did discuss our plans.

CHAIRPERSON OVERTURF: Any other comments, questions?

(No response.)

CHAIRPERSON OVERTURF: We are about a half an hour ahead. So I thought we would just go ahead and take the break now, reconvene at three o'clock, and we'll try to answer the questions for the vaccine selection.

(Whereupon, the foregoing matter went off the record at 2:33 p.m. and went back on the record at 3:01 p.m.)

CHAIRPERSON OVERTURF: Okay. I'd like to call the meeting to order again.

First of all, I'm going to have -- Roland is going to give us a breakdown of the options for strain selection. He's going to answer all questions at this point.

(Laughter.)

DR. LEVANDOWSKI: Okay. If you're expecting me to boil this down to one number, I don't think I can do that.

Something has happened to my slides here. Where are we? Sorry for that.

Okay. So just to summarize a little bit before I go into what the options might be here, for H1N1 Influenza A viruses, as we heard this morning, there are relatively few Influenza A H1N1 viruses that have been circulating around the world, and at this point in time, there is no firm evidence that there is any H1N2 viruses that are still circulating.

There have been isolates sporadically from a number of areas, including Africa, the Americas, Asia, Europe, Oceania, basically everyplace, and there has been a single outbreak that was reported in Tunisia.

The HAs of the H1 strains are antigenically all very similar to the current vaccine strain, which is A/New Caledonia/20/99, and the H1 viruses that are currently circulating are also generally well inhibited by antisera from people that have been immunized with vaccines that contain A/New Caledonia/20/99.

The high growth reassortant of A/New Caledonia/20/99 is available. It grows well. It manufactures well, and it's a well know entity for the last several years. So for H1N1, of course, the first option is to retain the A/New Caledonia strain as the vaccine strain.

And in favor of that, as I just mentioned, most of the H1 viruses are the most recent H1 viruses, the most recently isolated H1 viruses are A/New Caledonia-like by their antigenic characterization. The current vaccines do appear to be well matched for the HA of the current strains, and manufacturing is very well worked out.

Against this might be only that there have been so relatively few strains to analyze at this point in time, and the influenza season isn't exactly over yet.

For the second option, the second option would be to use a more recent H1N1 virus for manufacturing, and this probe here is, I think, a fairly weak one. There might be a closer match with a hemagglutinin and a neuraminidase of contemporary strains. I think that's more likely in a genetic than in an antigenic sense.

Against this option would be that a new strain, of course, we're never really sure that it's going to provide any superior immunogenicity or efficacy compared to current vaccine strain, and certainly nothing has been done to investigate what might need to be done to support manufacturing for any new virus that would be chosen.

A third option would be to defer the recommendation, and in favor of that, that there might be or I guess there could be some hope that there would be some other contemporary strains that might be identified that would look closer to what we might expect for next year.

But based on what's happened so far, I don't think we have any true expectation that there's going to be any new forthcoming information on H1N1 viruses so that that would not point to a good option.

For the H3N2 viruses, the H3N2 viruses have predominated globally during this season, and that has actually been going on for quite a long time, the global predominance of the H3N2 viruses.

A new variant that's represented by the A/California/7/2004 strain has been identified already in many areas of the world, although it has only been recognized relatively recently. In January, I think, is when it really became obvious and clear from analysis of the strains that were being collected that there was this new variant that was developing.

The HA, most of the strains have been identified, not all, but most are antigenically distinguishable from either the A/Wyoming/3/2003 or the A/Wellington/1/2004 vaccine strains that have been currently in use.

And as you saw from the serologies, it's very clear that the overwhelming majority of new H3N2 strains are poorly inhibited by antisera from people who have been immunized with the current vaccines that do contain A/Wyoming/3/2003.

High growth reassortants, however, for the A/California/7/2004-like strains are not yet available, but on the plus side, there are a large number of egg isolates that have been recovered, and they're being evaluated and work is ongoing in many laboratories to prepare high growth reassortants that could be suitable for manufacturing.

So for the H3N2, the first option, of course, is to retain A/Wyoming/3/2003. In favor of that would be the manufacturing has been very well worked out. Yields are predictable, and that could all be accomplished easily.

Not in favor of that, however, is that, as mentioned, the HAs or most of the H3N2 viruses and certainly the great majority of the viruses in the last several weeks are antigenically distinguishable from the current vaccine strain.

And in addition to that serologic results from the current vaccines indicate poor responses against the more recently circulating viruses, and we also know that H3N2 influenza viruses are often responsible for significant morbidity and mortality, and so this choice should be made very carefully.

The second option is to change to use a more recent H3N2 virus, and here in favor of that, a more recent strain is likely to provide a closer match with the contemporary strains, and the ones that would be expected to be going forward.

The serologic results with the current vaccines do indicate, again, that most of the current strains are not well inhibited by current vaccine, and, again, the morbidity and mortality of H3N2 viruses is often quite significant.

Not in favor of this, however, is the fact that we don't at the moment have a high growth reassortant for an A/California-like virus in hand, and so yield potential for this, if that would be a choice is really there's no information to go on at all on this.

The third option is to defer the recommendation, and in favor of this, the choice does need to be made carefully because of the significance of the morbidity and the morality, and a more recent strain might be likely to provide a closer match for the HA and the NA.

But against this is that we already have a great deal of information from a number of sources about their current H3N2 viruses, and we don't really expect that there's going to be anything that we acquire that's going to change, significantly change or enhance our understanding of that in the next few weeks.

And then moving on to Influenza B, as has been pointed out and as has been continuing for some time, the Influenza B viruses and the two known HA lineages are co-circulating. Strains that are like the vaccine HA have continued to circulate, and they seem to be predominant everywhere, including in the United States.

However, there are strains that are more like the non-vaccine HA lineage, and they're making up approximately 20 to 30 percent of Influenza B viruses in the United States. Influenza B viruses haven't been predominant most places, but where they have been found, the majority of the strains have been B/Shanghai-like, and as was mentioned, in Japan so far with their influenza season, those are the only Influenza B viruses that are being recovered. It doesn't mean that that would be true in the future, but that's certainly true at the moment.

We didn't dwell on, but there does seem to be some antigenic drift in the non-vaccine HA lineage Influenza B viruses, and for the most part, the B/Shanghai-like viruses, B/Shanghai/361/2002 vaccine-like viruses seem to be pretty well inhibited by antisera from people who have gotten the current vaccine. It's not as clear cut for young children where responses may be less robust, and whereas we have information that suggests that although responses may be reduced against the non-vaccine lineage in adults and elderly, it's pretty clear cut that for very young children who haven't been immunologically primed and/or boosted with both of the different HA lineages that we can expect that immunization with one lineage is not going to produce antibodies that cross-react with the other lineage.

At this point no other vaccine strains have really been evaluated for potential for vaccine production, but there are some egg isolates that are available.

So having said all of that, the options here, again, would be the first option would be to retain the current vaccine strain, which is B/Shanghai/361/2002-like. In favor of that, manufacturing is very well defined. It's predictable now. The predominant strains continue to be in the same HA lineage, and they have been found in many parts of the world.

Against this would be that Influenza B viruses not in the HA lineage in the vaccine have been increasing in frequency a little bit in some places, and it's clear that they're not as well inhibited by either post infection or post immunization antisera, and in particular, in relation to the immunologically naive young children.

That brings me to the second option which would be to change to use a more recent B virus, and in favor of that we might get a better coverage for Influenza B viruses.

Against that, we don't really know that a new strain would provide superior immunogenicity and efficacy compared to the current strain. It's not clear that non-vaccine strains will increase any further infrequency, and again, they're not found in all areas of the world.

And adding another Influenza B strain may causae difficulties in manufacturing. As you heard, the B viruses are often the rate limiting ones these days and particularly if a wild-type virus needs to be used.

So the third option, again, is to defer the recommendation. In favor of that, there may be some additional information that comes out about what strains might be closer matches with the hemagglutinins and neuraminidases of the contemporary strains.

But against that, there's not any way really to know whether a new strain of either HA lineage would prove to be superior either in immunogenicity or efficacy compared to the current one overall, and it seems at this point, although there are still strains being collected, it's not clear that there will be any additional significant information to help to inform the recommendation.

So this is the question for the committee, just to put that up there. I'll leave that up there during the discussions. Again, the question for the committee would be: what strains should be recommended for the antigenic composition of the 2005-2006 influenza virus vaccine? And I would ask that that recommendation be based on consideration of the epidemiology and antigenic characteristics, serologic responses, and availability of candidate strains.

And I'll stop there.

CHAIRPERSON OVERTURF: Roland, usually it's intuitive in your recommendations what you're recommending, but I didn't catch it for the H3N2 strain. Could you be a little more precise there?

DR. LEVANDOWSKI: I'm sorry. I'm not sure I understand the question.

(Laughter.)

CHAIRPERSON OVERTURF: Did you have a specific recommendation for the H3N2 direction? You gave the pros and cons of three different decision pathways.

DR. LEVANDOWSKI: Right.

CHAIRPERSON OVERTURF: Did you have a preference for one of those pathways?

DR. LEVANDOWSKI: Oh, you're asking my preference.

CHAIRPERSON OVERTURF: Yes.

DR. LEVANDOWSKI: Well, I think what is very clear is that the current H3N2 viruses are antigenically distinguishable from what's in the vaccine, and furthermore, the current vaccines do not seem to produce antibodies that cross react with those viruses very well.

There is the difficulty of being able to prepare vaccine, but as compared to some previous years, we're in much better condition because CDC has been able to get some egg isolates very quickly for a lot of these strains, and it puts us in the position of being able to respond.

And as you heard from the manufacturers, they have some ability to wait to get the third strain. If there's a change in a strain, they have some ability to accommodate that in their manufacturing, although obviously, it would be ideal for them not to have any down time at all.

But there is the possibility of responding with a new antigen for the H3N2, and I would say that the data that's available to us points to that direction.

CHAIRPERSON OVERTURF: Yeah, I think I just want to clarify for the committee. I think the options actually include recommending a lineage reference for the H3N2 strain but could still defer, if we wanted, the final selection until we have more information about that.

DR. LEVANDOWSKI: Okay. So I really didn't understand the question. Yes, if you make a recommendation for a like strain, if it would be a California-like strain, for example, we would expect that it would be possible for us to meet that recommendation with whatever strains that we find that seemed to be appropriate.

And as the manufacturers indicated, we've done this in the past where the recommendation could be met by more than one strain if that's necessary.

CHAIRPERSON OVERTURF: Are there additional questions or comments from the floor? Dr. Schwartz.

DR. SCHWARTZ: The California-like seems to be the strain that's emerging, but if it were to stop emerging and the strains that we saw circulating were more like the Wyoming or the Wellington strains that had been used in the past, I think that the data that Nancy presented suggested that at least based on the ferret studies that the California-like strain would probably provide good coverage.

Can you go into anymore detail on that? Does that seem like a reasonable assumption based on the data? And are there any other data that you could share with us that might provide an indication that a California-like strain would still be effective against some of those Fujian-like?

DR. LEVANDOWSKI: I think that's actually a question more for Nancy than it is for me. I don't have the tables of data in front of me to look at, nor do I have an immediate response to that.

MS. COX: I think there are a couple of points I'd like to make. One is that that serologic data that we have from ferrets would indicate that California antiserum does cover a variety of viruses really quite well.

But the second point I'd like to make is that we really haven't in the past seen a situation where you have a new group of evolutionarily successful viruses which the California viruses really do appear to be now sort of falling off the radar screen and then going backwards to a previous strain. It might be that something even more successful and advanced comes along that we haven't seen yet, but we wouldn't expect to go backward.

But if we did, we would expect reasonable coverage, yes.

CHAIRPERSON OVERTURF: Any further discussion, questions?

(No response.)

CHAIRPERSON OVERTURF: What I would entertain then, I suppose, is a motion from somebody on the committee regarding perhaps the H1N1 strain, and then we can proceed with further discussions about the next discussions.

Dr. LaRussa?

DR. LaRUSSA: I make a motion that we retain the current H1N1 strain.

CHAIRPERSON OVERTURF: Okay.

DR. MONTO: Second.

CHAIRPERSON OVERTURF: We'll start with Dr. Wharton, and please vote yes or no.

DR. WHARTON: Yes.

CHAIRPERSON OVERTURF: Now, the motion actually, just to restate that, was that the current H1N1 strain would be retained.

Dr. Monto.

DR. MONTO: Yes.

DR. MARKOVITZ: Yes.

DR. ROYAL: Yes.

DR. FARLEY: Yes.

DR. McINNES: Yes.

DR. PROVINCE: Yes.

COL. PHILLIPS: Yes.

DR. COUCH: Yes.

CHAIRPERSON OVERTURF: I also vote yes.

DR. LaRUSSA: Yes.

DR. SCHWARTZ: Yes.

DR. WORD: Yes.

DR. DOWDLE: Yes.

DR. EICKHOFF: Yes.

DR. SELF: Yes.

DR. KARRON: Yes.

CHAIRPERSON OVERTURF: That was unanimous to retain the current H1N1 strain.

So I think we should proceed next to a discussion about what the committee's wishes would be for the H3N2 strain or the B strain for this year. Is there any discussion or suggestions?

DR. COUCH: I'd just like to make a comment that I think is important for us to keep in mind with our actions, and that is that we already have recommendations that are very specific from the World Health Organization. You see, we used to do this, and then they met after us. Now they meet, and we meet after them.

And harmonizing is a strong criteria for selection. If we don't agree with that, I would say that we're in the of having to need very strong evidence. We've got manufacturers who are already underway, and that is what they're doing, and their primary manufacturing sites are in Europe, not in this country.

If we want more vaccine in this country, harmonization, I think, has to be a part of our consideration.

Now, having said that, I would say that, you know, all of the data points to needing to change the H3N2, and they have selected A/California, and we've heard data all about A/California. So we vote for A/California, for H3N2.

DR. MONTO: Second.

(Laughter.)

CHAIRPERSON OVERTURF: I would agree. I think the clarification that was necessary was to know that all we're really doing is voting for a lineage. We're not really the ones. Obviously the virologists have to decide how to do that, and I think that's all we really are voting for at this point.

Yes, Dr. Eickhoff.

DR. EICKHOFF: A question for the maker of the motion. Is your motion to be construed as including A/California or an A/California-like virus?

DR. COUCH: A/California is never a specific recommendation. Any time you say that it will always be an A/California-like virus. I think that's implied. Oh, sorry.

It's always an A/California-like virus, never a single specific virus.

CHAIRPERSON OVERTURF: So there's been a motion for a change in the H3N2 virus to an A/California-like virus. I guess at this point we can start on the other end of the room and ask Dr. Karron to vote.

DR. KARRON: Yes.

CHAIRPERSON OVERTURF: You can make any comments you wanted to when you vote.

DR. SELF: Yes.

DR. EICKHOFF: Yes.

DR. DOWDLE: Yes.

DR. WORD: Yes.

DR. SCHWARTZ: Yes.

DR. LaRUSSA: Yes.

DR. COUCH: Yes.

CHAIRPERSON OVERTURF: That's right. Yes.

COL. PHILLIPS: Yes.

DR. PROVINCE: Yes.

DR. McINNES: Yes.

DR. FARLEY: Yes, and my only comment, and it's really a question that we can discuss after we finish, but the idea of coordinating this with WHO's decision making process and the timing of such meetings comes to mind here. I mean, we don't want to just be rubber stamping another organization, but if we're all doing the same thing, is there some benefit of working together in some way more directly?

So my vote is yes.

DR. ROYAL: My vote is yes.

DR. MARKOVITZ: Yes.

DR. MONTO: Yes.

DR. WHARTON: Yes, and a comment. My compliments especially to I guess my colleagues at CDC for having had the foresight to assure the availability of a number of egg isolates so that we're in a good position this year to move forward to development of a high yield reassortant.

CHAIRPERSON OVERTURF: Yes, Dr. Markovitz?

DR. MARKOVITZ: Yes. I'd also like to make the same or to second that thought, that really I was very impressed with the preparation this year, and I think that's very nice.

CHAIRPERSON OVERTURF: Well, sometimes I think we are endowed with good luck in terms of having isolates, and I think this year that seemed to be the problem, and I think actually deferring most of the recommendations at this point would probably have little influence.

And having been through this process now this is the fourth year for me, that rarely results in any major change. Two to three weeks more data just doesn't usually provide us with enough information. It has on occasion.

So the next issue the B viral strain for this year, and the suggestions and options for the B viral strains included retaining the B/Shanghai/361/2002-like viruses or to consider using a more recent B virus or to, again, defer that recommendation to another point in time.

So I open it up to the committee's discussion as to whether or not they would like to see more data or whether they think we should vote on the B strain now.

CHAIRPERSON OVERTURF: Dr. LaRussa.

DR. LARUSSA: Well, it seems to me that since most of the strains are still Shanghai and we're not going to get a whole lot more information and we're not in any position right now to recommend putting two B strains into the vaccine, but I think at least for now it would be wise to recommend retaining the strain, but moving towards the future to potentially providing a pediatric vaccine with the two Bs where it's most needed.

So my recommendation would be to retain.

CHAIRPERSON OVERTURF: Any other discussion? Yes, Dr. Farley.

DR. FARLEY: Well, I guess I'm wondering what would put us in the position of making such a recommendation. I mean, what do we need to come together for that recommendation to make more sense?

DR. LaRUSSA: You mean of having two strains? Well, I think the studies need to be done that you can actually get a good immune response in young children to two strains and that you could actually provide enough viral antigen to actually make that response.

Once that data is available, then it becomes a logistics problem of whether it's physically possible, but at least you'll know that if it works immunologically that you're going to have a constant need for it every year, and you'd feel much more comfortable about doing it, and then you could go ahead and do whatever you want for the adults.

CHAIRPERSON OVERTURF: Actually, this particular issue has been discussed almost every year that I've been on this committee, and I think there has been always a great desire on the committee's part to expand the B strains.

What is unique now is that we have a pediatric option, which we did not, and that may be a unique way to enter the problem and begin to accumulate the data.

Yes, Dr. Dowdle.

DR. DOWDLE: I think this issue about two circulating B strains or at least some concern about there being two types that are essentially two groups that seem to be around at different times and changing and prominent and so on is not new. This has been going on for a long time, and it seems that looking back, it seems that we've been more in this quandary than we have not had this quandary. I mean it has been more of a usual thing.

It would be very useful if I think next meeting that we could have a little paper on Influenza B and some of the issues over the years about how this, number one, has been dealt with, what are the issues in the past, and what are the advantages and disadvantages of going with two strains?

I mean, it would be nice to have some data for a change, and I think we've been faced with a lot of opinions and a lot of memories, but my memory could be completely wrong. It's just certainly what I remember in the past, and we've had this discussion, many, many times.

CHAIRPERSON OVERTURF: Dr. Farley.

DR. FARLEY: I fully agree, but I wonder what will motivate. Who will do? Who would produce the data and who's motivated to produce that data? Is that the manufacturer of the pediatric vaccine or is that FDA? You know, I guess I'm wondering about the process so that we won't be in the same place each year without some new data.

CHAIRPERSON OVERTURF: Yes, Dr. Eickhoff.

DR. EICKHOFF: I think the process or the discussion here has gone maybe a step beyond in what Walter commented on, namely, it's now at a level where are we going to talk about a whole separate pediatric product.

That raises a host of issues, and we would need manufacturer input on it, but a pediatric use of influenza vaccine we kind of anticipate is going to grow almost logarithmicly for the next several years. It's an issue that needs to be considered very carefully, but maybe not in this immediate framework of strain selection.

But, again, I think manufacturers will have lots to say about that.

CHAIRPERSON OVERTURF: The assumption I had from Dr. LaRussa was that he actually is talking about expansion of the or exploring the use of two B strains in young children, the current population in whom routine annual immunization is now recommended, the six to 23 month old children; is that correct?

For this year. We still have the difficulty of deciding which of the three options the committee would like to proceed with in terms of the option. My personal feeling is that I don't see it. From the data that Dr. Cox presented this morning, it's clear that the B/Shanghai lineage was about 80 percent of the isolates in the United States so that there seems very little reason right now to change that, although there are these disturbing data -- there always are -- of isolated settings where there is perhaps a 50-50 breakdown in certain settings.

PARTICIPANT: Dr. LaRussa made the recommendation to keep (speaking from an unmicked location).

CHAIRPERSON OVERTURF: Did you make that recommendation?

DR. LaRUSSA: Yes.

CHAIRPERSON OVERTURF: Do you want to make it as a motion?

DR. LaRUSSA: I'll turn it into a motion.

CHAIRPERSON OVERTURF: Okay, fine.

DR. LaRUSSA: I make a motion that we keep Shanghai this year.

CHAIRPERSON OVERTURF: Would somebody like to second that motion?

DR. WHARTON: (Show of hand.)

CHAIRPERSON OVERTURF: Dr. Wharton. Okay.

Well, we will start with you for the first vote. So the motion on the floor is to retain the B/Shanghai isolate for the current vaccine.

DR. WHARTON: Yes.

DR. MONTO: Yes.

DR. MARKOVITZ: Yes.

DR. ROYAL: Yes.

DR. FARLEY: Yes.

DR. McINNES: Yes.

DR. PROVINCE: Yes.

COL. PHILLIPS: Yes.

DR. COUCH: Yes.

CHAIRPERSON OVERTURF: Yes.

DR. LaRUSSA: Yes.

DR. SCHWARTZ: Yes.

DR. WORD: Yes.

DR. DOWDLE: (Speaking from an unmicked location) Yes.

DR. EICKHOFF: Yes.

DR. DOWDLE: Yes.

DR. SELF: Yes.

DR. KARRON: Yes.

CHAIRPERSON OVERTURF: This may have been done in record time this time.

(Laughter.)

CHAIRPERSON OVERTURF: Is there any further discussion or clarification? Does anybody want to speak? Dr. Eickhoff.

DR. EICKHOFF: Now that we have voted and we're not slavishly following WHO recommendation, the WHO recommendation for the Southern Hemisphere was just a little bit different than that for the Northern Hemisphere and the B component, and a question for Roland or Nancy, I guess: do you have any insight into why this recommendation changed slightly?

DR. LEVANDOWSKI: Actually, maybe I misled people. The serologies were done with an older Southern Hemisphere vaccine. The current vaccine is recommended by WHO for Southern Hemisphere includes a B/Shanghai/361/2002-like vaccine virus. The previous vaccine that was used in the Southern Hemisphere had B/Brisbane/32/2002, and that's the serum that was available for doing serologies because the new vaccine for the Southern Hemisphere, according to the current recommendations is only just now being produced and used. So it was not available at the time the sera were collected for the studies.

So maybe I confused people with that, but the recommendation, the current recommendation for the Southern Hemisphere is B/Shanghai/361/2002-like.

CHAIRPERSON OVERTURF: Dr. Schwartz.

DR. SCHWARTZ: I'd just like to briefly raise three issues. Just to get a sense, and this is the first time I've been at this committee meeting. So I don't know how the committee deals with these types of recommendations, but I think Monica mentioned the possibility of changing the timing of this meeting to perhaps more closely coincide with the WHO meeting or at least to have maybe better communications between the two groups.

I was wondering if that's a recommendation that the committee could make and how the committee would handle that particular idea.

It has also been discussed that we believe that a pediatric vaccine containing the two different B lineages should be studied, and I don't know if there is additional force if there's a vote and if that becomes an official recommendation of the committee.

Also I don't know if the manufacturers have candidate vaccines that include the B/Victoria lineage and whether there is material right now that could be used or whether new pilot lots would have to be produced and, therefore, there would be substantially increased time and costs involved with doing that.

The third issue that I just wanted to very briefly mention has to do with H5N1, and Pam earlier talked about some clinical studies that NIH is doing of investigation of lots of H5N1 vaccine, and I think it would be useful for this committee in some future meeting to discuss under what circumstances it might be considered to have H5N1 as a component of the influenza vaccine which would provide priming to those folks who were vaccinated in case H5N1 or H5 emerged as a pandemic strain.

That's certainly not something to talk about now, but I think it would be an interesting discussion to have in the future.

CHAIRPERSON OVERTURF: To me the pride -- and somebody from the FDA can correct me -- I would think that if there's a process that we really want to devote an entire meeting or a day to expansion of the B types, which would include presentation of data, and then an official recommendation and a vote, I think it would require me to -- I don't think we can do that here. I think this has come up over and over and over and over again, and perhaps it's something that the FDA needs to take into hand and perhaps plan a meeting for the VRBPAC specifically to address that single issue some time with some inclusion of data.

I think that's the process that has usually been used; is that correct?

Yes, Dr. Dowdle.

DR. DOWDLE: I'm going to change the subject. Go ahead, Nancy, please. I'm going to change the subject.

DR. COX: Yes. Well, I was going to change the subject as well.

(Laughter.)

DR. COX: And for those of you who haven't been involved in the vaccine strain selection process as long as I have, it may seem to you as if this meeting could be a rubber stamp and that may feel uncomfortable to you.

We used to have the U.S. meeting first. This caused a great deal of discomfort on the part of the rest of the world which felt that the U.S. was preempting the vaccine. It also put the U.S. at a disadvantage because our meeting was first. We didn't have access to the global data.

We had our WHO vaccine meeting in Geneva. It finished on Friday at about noontime on Friday, the 11th. We're here only a few days later really if you take the weekend into account. We were able to bring back all of those data, the global data, and compile them. And so some of the charts that I showed you really had incorporated the data from the other four WHO collaborating labs as well as data from some other national influenza centers.

And, of course, Roland was able to include the serologic data from the other four WHO collaborating centers in his presentation. So from the perspective of completeness, I think it is an advantage for the committee to actually see all of the data that's available globally, and so I think I just wanted to clarify that and make sure that you understand that there is extremely close coordination between CDC and FDA and WHO and the NIH and WHO as well, even if you don't see that specifically as a committee member.

DR. FARLEY: That's very helpful. Thank you.

CHAIRPERSON OVERTURF: Dr. Markovitz.

DR. MARKOVITZ: If I could just add to what Nancy said, from what I've seen -- I think this is my fourth year on the committee -- is that having the meeting earlier would not be good because we seem to get like a lot of our best data within the last few weeks right before the meeting actually takes place. So I'm less concerned about appearance of rubber stamp and more concerned that we have everything together to make the best decision.

CHAIRPERSON OVERTURF: I think the only thing I would add is that at the time the presentation is made about what the options are, I really think we should -- it was in the very first presentation that I think Dr. Levandowski made this morning, what the recommendations for the Northern and the Southern Hemispheres were, but it probably needs to reappear at the time just prior to the vote again because I think Dr. Couch --

DR. COUCH: I think Nancy and Roland could confirm it. The technical comment, as I understand it, the WHO is a global recommendation, and then it is up to each individual country and organization who deals with these decisions to then consider and make their own independent decisions.

But that doesn't negate the comment I made earlier. It's important for us to harmonize unless there's a strong reason not to.

CHAIRPERSON OVERTURF: Norman.

DR. BAYLOR: I just wanted to comment on the comment made about in a sense the vaccine development, like the H5N1, incorporating that into the current vaccine and whether this body would be the one to discuss that and make recommendations.

I think that process starts with the manufacturers and sponsors of INDs, and we have that discussion with FDA, and there's a process to have those type of meetings. And as those discussions expand and we start moving into clinical studies, that may be something that we would bring back to the VRBPAC, but the process will start with a discussion between the manufacturers, sponsors, and the FDA.

CHAIRPERSON OVERTURF: Any other discussion, questions? Dr. Markovitz.

DR. DOWDLE: If I understand your question, Ben, it really is how would an avian flu vaccine interdigitate with this one, right? So that's, I think, your question. What would be the answer to that? I'm curious, too.

No takers.

CHAIRPERSON OVERTURF: Dr. Levandowski.

DR. LEVANDOWSKI: Well, I think there are many precedents for going all different directions with influenza vaccines. Historically people have been telling me all day how we've had five and six antigens in the vaccine in very early days. In early times it would be something like PR/8 plus another H1 and B/Lee plus another B, and some of those uses, I guess we wouldn't go back to doing it exactly that way.

But we've had monovalent vaccines. We've had monovalent supplemental vaccines. Most recently the one that comes to mind was in 1986, the A/Taiwan/186 supplemental vaccine. It's possible to have a monovalent vaccine. I mean we're used to thinking about trivalent vaccine, and that's a convenient way to administer the product, but in reality that's just one way to do it, and monovalent is always available for us.

And this may be a consideration not only for an H5N1 vaccine, if it comes to that, but it could be a consideration for priming and boosting young children as well. If we need an additional component for a specific population that might be an alternative way to go.

But, again, thinking about how we get the different valencies of vaccines, before H1N1 reappeared it was an AB. It was a bivalent vaccine and nota trivalent vaccine.

So there are ways to get there, and I don't think there's anything that's so set in stone that it couldn't be evaluated and worked out

CHAIRPERSON OVERTURF: Dr. LaRussa.

DR. LaRUSSA: Just two quick comments. Can we work on some way to get some updates on what's going on with the California light strains once you see how the reassortments work out?

And then the second thing is I'm not sure how we left things with the discussion about the pediatric vaccine. Is that something you're going to go get back to us about, whether we could discuss that at a future time?

DR. MIDTHUN: Karen Midthun, FDA.

I think with regard to trying to explore bivalent B vaccines in a pediatric population that's certainly an excellent suggestion. I think what's needed though, and I think Dr. Baylor was alluding to this in the context of H5N1 is that you need to have a sponsor working in conjunction with a manufacturer who is willing to undertake such studied, and certainly we as FDA are there to work with them as they put these particular products into clinical trials to evaluate them.

So I think the question really is, you know, there has to be identification of an entity who is willing to make such a product, and then there has to be a sponsor who's willing to take that product into an IND. And obviously we're there to work with whoever would want to do this and think that would be very good information to be able to have for everyone's consideration.

But you know, we, FDA, don't have the ability to make let's say a candidate vaccine and then conduct a clinical trial. We're there to provide oversight as this goes forward, and I think that's certainly hearing the discussions that have been made here. You know, we can get together with the rest of the other agencies and Health and Human Services and discuss, you know, are there some ways that one could facilitate moving into that direction.

CHAIRPERSON OVERTURF: I think it's a problem a little bit of agency overlap. I think there's issues that a lot of who makes recommendations about research and what needs to go forward. Obviously the NIH is involved. NVAC and other advisory committees really have to make a decision about whether this is a viable alternative.

I don't know whether this committee can do anything more than get it on the table, and I think that's what we've done, but I still think that it could be carried as a topic, even as a portion of a meeting some time in the future so that we give it a little more airing.

Dr. Dowdle, did you have a question?

DR. DOWDLE: No, no, no, no, no. The point I was going to make was exactly the same point Nancy made about a plea not to change this meeting in relation to the WHO meeting, and just to point out that as Bob said, if there's a good reason to change, we certainly could change. And I would add that it wouldn't be the first time that we went against U.N. advice.

CHAIRPERSON OVERTURF: My experience has been the same. This is my fourth year on this committee, is that the data that are available from all sources seem to come together for this meeting, and usually it's very close to the WHO meeting. It would seem unlikely that it would be helpful to move it back much further. We would have even less data to try to develop our own policies and recommendations.

DR. COUCH: I have one more comment. I need to point out to this audience my name is not Ed Kilbourne, but we continue not to give the kind of emphasis that many of us think we should to the neuraminidase.

You see, we've gotten it from when Roland talked about selections, he didn't just say the hemagglutinin. He said the hemagglutinin and the neuraminidase, and when Dr. Cox presented her virus isolates, she gave us the evolutionary development of the neuraminidase.

It's an important antibody most of us think. I'm a strong believer. We really want the hemagglutinin as the primary antigen, but the neuraminidase is an important second antibody. I think that was one of Walter's terms one time, and we've never really come to grips with what kind of standards or that we could adopt to include the neuraminidase.

It does contribute and it is desirable, and we do know from some of the tests that have been done on vaccine marketed preparations that the quantity of neuraminidase activity varies all over the map, and presumably some of them lack immunogenicity for the neuraminidase at all, and others are probably pretty good.

This is not to anything other than make that as a point of record, that I think this committee needs to continue to think and to try and make attempts to how to address the neuraminidase in some sort of standards, and that's not easy in itself, for those of you who are worried about what would you say is your standard for the neuraminidase in each vaccine. I have my own ideas. That's a different discussion.

But I don't think we should lose sight of that antigen as an important one that we all recognize is a part of the vaccine immune response that we'd like to have and somehow get it in with not hopefully in the too distant future some criteria for its presence and appropriateness of immunogenicity.

CHAIRPERSON OVERTURF: Dr. Schwartz.

DR. SCHWARTZ: Let me just ask you a question about that. So we're talking about a number of potential California-like strains that we're developing reassortants for, and so let's say that there are reassortants for five different California-like strains that go to the manufacturers.

Some of them may produce more neuraminidase than others and some of them may grow better in eggs than others. Are you suggesting that their neuraminidase production be a criterion for which is the best California-like strain that could be selected?

DR. COUCH: No. I think it's a manufacturing product decision, not an antigenic seed. I think that what CDC would contribute to that and the FDA would be that if there are neuraminidase differences among the A/California-like strains, then they would select one that is more characteristic and more changed perhaps than A/Wellington or an A/Fujian, just for that hope that that gets the appropriate neuraminidase.

So far I don't think there have been any examples in which that's been necessary, but that would be the seed control on that. No, we're talking about manufacturing control on the presence and immunogenicity.

CHAIRPERSON OVERTURF: Any further comments?

Are there any further comments from the floor or the manufacturers?

(No response.)

CHAIRPERSON OVERTURF: Christine Walsh has an announcement.

MS. WALSH: I'd just like to ask all of the committee members to please take your red folders with you tonight. They do contain confidential FDA material that's in them, and we cannot leave them in the room overnight.

Thank you.

CHAIRPERSON OVERTURF: The meeting is adjourned.

(Whereupon, at 3:51 p.m., the meeting in the above-entitled matter was adjourned, to reconvene at 8:30 a.m., Thursday, February 17, 2005.)