UNITED STATES OF AMERICA

P R O C E E D I N G S

(8:35 a.m.)

CHAIRMAN DAUM: Good morning and welcome to the meeting of the Vaccines and Related Biological Products Advisory Committee. Let me begin by asking everyone in this room to consider some personal noise hygiene measures at this moment and to please turn off all beepers, all cell phones or at least reduce them to a vibrating mode where no one will be disturbed by hearing them.

The second measure, usually the first, I'd like to turn the floor over to Dr. Sachs, Jody Sachs of the FDA, for administrative matters, including a conflict of interest statement.

Dr. Sachs.

DR. SACHS: Welcome to the 93rd meeting for the Vaccine and Related Biological Products Advisory Committee. I welcome the members, and thank you all for coming, and also I welcome the public.

I'd like to read a conflict of interest statement for the record.

The following announcement addresses the conflict of interest issue associated with the Vaccine and Related Biological Products Advisory Committee meeting on December 17th, 2002.

The Director of the Center for Biologics Evaluation and Research has appointed Nancy Cox, Dr. Kathryn Edwards, Dr. Theodore Eickhoff, Dr. Bruce Gellin, Dr. Holli Hamilton, Dr. Martin Myers, and Dr. Dixie Snider and Dr. Mark Steinhoff as temporary voting members for this meeting.

To determine if any conflicts of interest existed, the agency reviewed the submitted agenda and all financial interests reported by the meeting participants. As a result of this review and based upon the FDA draft guidelines on disclosure and conflict of interest for special government employees participating in an FDA product specific advisory committee meeting, there were no meeting participants who required a waiver under 18 USC 208.

Dr. Diane Griffin, Peter Palese, Richard Whitley recused themselves from participating in this meeting.

We would like to note for the record that Dr. Michael Decker is participating in this meeting as a nonvoting industry representative acting on behalf of regulated industry. Dr. Decker's appointment is not subject to 18 USC 208. He is employed by Aventis. In the event that the discussions involve specific products or firms not on the agenda and for which FDA's participants have a financial interest, the participants are reminded of the need to exclude themselves from the discussion. Their recusal will be noted for the public record.

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

The following VRBPAC members will not be present today: Dr. Diane Griffin, Dr. Audrey Manley, Dr. Peter Palese, and Dr. Richard Whitley.

In your members' folder on the left-hand side there's a flyer that tells you at noon there are two restaurants that are open in the hotel available for lunch. So it's a little flyer and it shows you what two restaurants you can eat lunch at quickly.

And with that I'm happy to turn over the meeting to Dr. Daum, our Chair.

CHAIRMAN DAUM: Thank you very much, Jody.

The next item in our agenda is a bittersweet one for some of us, and that is the plaque presentation of retiring VRBPAC members, and for this purpose we call on Dr. Kathryn Zoon.

Dr. Zoon, good morning.

DR. ZOON: Good morning. It's a pleasure to be here this morning, and I think like Dr. Daum, it's a bittersweet meeting for me as well because it's my last VRBPAC meeting. So I just wanted to say a few words today.

It's a special pleasure for me to be here. Our FDA Advisory Committees are one of our most important vehicles at the FDA to provide us with the expert advice and scientific deliberations that help us ponder very important decisions that affect the public health. It helps us make the processes and our scientific analyses public in a way that we can share this information with you and understand the questions and also the importance of the decisions that will be made both for now and in the future with respect to products and policy for the public.

This particular committee, I have to say, has had just a tremendous amount of responsibility, and it has hung tough when it needed to. It has provided expert scientific advice on important issues, and I think the membership here should feel proud of its contributions both to public health and for dealing with the health and safety of the vaccine and related products that we have asked for your advice.

I'd like to ask Dr. Karen Midthune to please come up and join me. Karen is the Director of the Office of Vaccines, and it really for many reasons is a time when we all reflect on the contributions of this committee. In some ways it's a little ironic that we wait until you all leave, but I think you are appreciated all through the process, and certainly we officially recognize you at the end of your term.

And I want to say to Bob Daum and Walter Faggett and Barbara Loe Fisher -- and, Barbara, it's good to see you -- and also Diane, who I know couldn't be here, but each and every one of you during your term here has provided insight and a diverse opinion on the application and use of vaccines, and that's important. It's important for the process. It's important for moving these products forward and having them being used in the proper way to protect our children, to protect adults, and I think we are all terribly thankful at the FDA for everything that this committee and particularly the members here today that are leaving us after serving such a what I would say just chock full of challenges on this committee.

And if I could ask each of the members who are here today to come up, Bob, Walter, and Barbara, I'd like to present you with a plaque and a letter from the Commissioner.

I'll start out, and, Karen, while I'm getting this technically challenging activity, would you please like to say a few words?

DR. MIDTHUNE: Sure. I'd just like to echo what Kathy said. We really, really do appreciate your input, and we recognize what a phenomenal time commitment it is to really read through these materials, meeting after meeting, and really come and be so prepared as you are to give us the very important advice. So I would just like to add my thanks because we're asking a lot of you, and we really appreciate it.

DR. ZOON: Thank you. Thank you, Karen.

So I'll start out with Bob Daum, and I'd like to read the letter from Linda Skladany. It says:

"I would like to express my deepest appreciation for your efforts and guidance during your term as a member of the Vaccines and Related Biologics Products Advisory Committee. The success of this committee's work reinforces our conviction that responsible regulation of consumer products depends greatly on the participation and advice of the entire health community.

"In recognition of your distinguished service to the FDA, I am pleased to present you with this enclosed certificate. Thank you."

Bob, congratulations, and thank you for being the Chair.

(Applause.)

DR. ZOON: Bob, I think we're supposed to get a picture.

I think this letter says the same. So I'm not going to read it again, but I just want to thank you so much for your participation. It has really been wonderful.

(Applause.)

DR. ZOON: With that, Robert, I turn the duties back over to you as Chair, and thank you for everything.

CHAIRMAN DAUM: I'd actually like to take two or three minutes and just make a couple of comments on a soap box as I won't have an opportunity to do this again any time soon.

Being chair of this committee is perhaps the most honored distinction of my entire professional career, and I thank everybody who has allowed me to help in this regard. It has been a way of giving back to the community from which I've taken many things during my professional career.

I'd like to say a special thanks to the people who make these meetings possible. I had the great pleasure of seeing Nancy Cherry this morning, who is the previous Executive Secretary to this committee. Nancy taught me almost everything I know and I think is the owner of the little bell that it has been my custom to ring when we have the committee called to order.

Jody Sachs has stepped into tall shoes, but has done very admirably, and I have every expectation that she will continue to provide excellent support to Dr. Stephens, the new Chairman of the committee who I'm told it's okay to say that.

Bill Freas has also been wonderful help, as has Denise Royster and her staff, making things work.

We've had meetings at an unnamed Hotel H and an unnamed Hotel M, which you currently sit in. I'm very grateful that we have our last meeting in Hotel M, as it offers many, many, many amenities that are much nicer than they often are in Hotel H.

Dr. Stephens, as the incoming Chair, you'll notice is in the hot seat today where he will comment first at the appropriate time. I can tell you, David, that you trade one disease for another. When you come up here the tension is a little bit less in terms of being the first speaker, but there are other issues that will bear upon you, you'll see.

To the previous Chairs, Harry Greenberg, who is here today, Harry taught me to never make the committee do a working lunch. We've got to get up. We've got to stretch our feet. We've got to go to the bathroom. Thank you, Harry, for that lesson.

(Laughter.)

CHAIRMAN DAUM: Pat Ferrieri taught me where eateries were near Hotel H that would provide relief from the cuisine offered therein.

(Laughter.)

CHAIRMAN DAUM: The FDA folks need to be respected. In my view they're heroes quite literally. It may sound corny to say. The doings of the agency often go on behind the scenes. The culture is an unusual one in that one can't often talk freely in public about things one would like to say.

But I just point to the track record that we Americans are used to in terms of having safe and effective products on the market routinely with almost never, almost never a variance, and it is from their hard work and their determined effort that this is possible.

I take my hat off to all of them. I feel sad that I won't be working with them on a day-to-day basis after this, but I just want them to know that they are in my admiration and esteem and will be forever.

The research that they do is not well known among many sectors of the public. It's extremely important. One of the things I've tried to do under my leadership of this committee is to make people more aware of the research that goes on in the agencies and the crucial importance of it in properly regulating and insuring safe and effective drugs and vaccines.

Our effort was cut short by September 11th. The good news is that a great deal of research funding poured into the agency after that and, of course, partially alleviated the problem.

But there's much more work to do, and, David, I hope that you will continue this effort to provide proper research funding for the agency so that this is not a concern they have to dance to every day.

The Vaccine Shortages Committee provided an opportunity to get some recognition for research effort at the agency in writing, and I was very pleased to help craft a paragraph in that document that called attention to this effort.

Finally, a couple of comments about vaccine safety and public trust. The public, of course, expects the safest vaccine supply possible. They also expect, in my opinion, protection from vaccine associated diseases, and we can't compromise either of these objectives.

There are many people who enter into a dialogue about vaccines, and all of them need to understand in my opinion three rules.

One, the vaccines currently in use for routine childhood diseases are extremely safe, but not always. The vaccines in routine use for childhood disease are extremely effective, but not always. And the risks of today's vaccine are massively outweighed by the benefits of today's vaccines.

And so I would like to close with a plea to all individuals in this room and beyond who are involved in the dialogue and interaction and discourse about vaccines, from manufacturers to regulators, to public health officials, to academics, to vaccine safety activists. Please continue this dialogue. Please come onto the playing field ready to interact and play the game, but please remember these three rules, the rule about safety but not always, the rule about effectiveness but not always, and the rule about which side of the scale the risk and benefits equation comes down on.

Look at the progress we've made in eradicating or nearly eradicating so many childhood diseases.

For everybody's time and effort and putting up with me and my vagrancies and during this chairmanship, I'm very grateful, and thank you very much.

(Applause.)

CHAIRMAN DAUM: And now back to everyday committee business. I'd like to begin by asking the committee members and temporary voting members and others seated at the table to identify themselves, and, David Stephens, we will begin with you.

DR. STEPHENS: Happy holidays. I'm David Stephens from Emory in Atlanta, and I want to thank Bob for his leadership of this committee.

DR. KATZ: I'm Sam Katz from Duke University, and since you gave me the microphone, I'd like as a non-government employee to make a comments. If I understood correctly, Kathy Zoon said this was her last meeting, and I think it's inappropriate that we don't have some sort of accolade, a plaque, a bouquet of flowers, and other ways of saying that she has been an incredible leader.

If what I read in Science magazine a week or two ago is correct, I think she has been the victim of what is very wrong direction by some of the people directing our government. I know that as members under the conflicts of interest we can't lobby; we can't do this or that, but we need to do something to preserve CBER and see that it has people like Kathy Zoon leading.

Thank you.

(Applause.)

CHAIRMAN DAUM: Thank you, Dr. Katz.

Many of us just learned about this a few moments before the meeting began, but your comments are perfectly appropriate, and I appreciate them.

DR. EDWARDS: Sam is always a hard act to follow.

(Laughter.)

DR. EDWARDS: I'm just Kathy Edwards from Vanderbilt.

DR. SNIDER: I'm Dixie Snider. I'm the Associate Director for Science at CDC. Also I suppose of relevance is that I am the Chair of the Advisory Committee on Immunization Practices at CDC and have other engagements with vaccine issues, such as smallpox, and the Technology Transfer Office at CDC comes under my purview.

And like Sam, I would like to take an opportunity just to make a couple of personal comments. One is to Bob and thank him for his excellent leadership of this committee and to all of the people who serve on this committee.

When I was a member of the committee and over, I guess, the seven years or so I've been either a member or a consultant, it's been a wonderful opportunity, and I particularly appreciate all of the relationships I have with the people at FDA. I think CDC and FDA have great relationships around vaccines.

I'm really going to miss Kathy Zoon. She's not only a great public servant, but a personal friend, and I think it's a great loss to FDA.

DR. HAMILTON: Holli Hamilton, NIH.

DR. EICKHOFF: Ted Eickhoff, University of Colorado.

I would like only to say that, Bob, the meetings that I have been privileged to attend have vastly benefitted by your guidance as Chair. So thank you very much.

And to Kathy, again, best wishes.

DR. COX: Nancy Cox from CDC.

DR. GELLIN: I'm Bruce Gellin. As of six weeks ago, I'm the director of the National Vaccine Program Office at HHS.

DR. STEINHOFF: I'm Mark Steinhoff from Johns Hopkins University School of Medicine and School of Public Health.

DR. MYERS: I'm Martin Myers from the University of Texas at Galveston.

DR. OVERTURF: I'm Gary Overturf. I'm Professor of Pediatrics and Pathology at the University of New Mexico.

DR. DIAZ: I'm Pam Diaz, the Director of Infectious Diseases for the Chicago Department of Public Health.

DR. FAGGETT: I'm Walt Faggett, pediatrician here in D.C., Medical Director for Medicaid in D.C., and I, too, want to express my appreciation of Bob Daum's leadership for this past four years. It has really been a pleasure serving on the committee, and the experience here is really going to help us, I know, in D.C. as we look forward to the challenge of smallpox immunization. Some of the decisions that we make will really benefit from the exposure I've had here on this committee.

So, again, I'm very privileged to have served and will look forward to staying in touch.

Thank you.

DR. MARKOVITZ: I'm David Markovitz from University of Michigan.

DR. PARSONNET: Julie Parsonnet from Infectious Diseases at Stanford University.

DR. FISHER: Barbara Loe Fisher with the National Vaccine Information Center.

And I'd just like to say it has been a great privilege to serve on this committee. I believe that we need a strong FDA, and I particularly want to thank Kathryn Zoon for making this possible for me to be on this committee.

DR. DECKER: Michael Decker, Aventis Pasteur and Vanderbilt University, Department of Preventive Medicine.

Bob, you've done a great job chairing, and I particularly have admired the way you've insured full and fair discussion

DR. GOLDBERG: Judith Goldberg, Director of Biostatistics at New York University School of Medicine.

Bob, you've been a great Chair. I've learned a lot watching you.

And, Kathy, you've led this group to allow free and open dialogues, and you've been absolutely responsive to every request that this committee has made, and I've seen the changes in documentation as a result of all of that. Thank you and best of luck.

DR. PRATT: Douglas Pratt, FDA.

DR. MINK: ChrisAnna Mink, FDA.

CHAIRMAN DAUM: While Dr. Sachs is getting ready, I'm Robert Daum. I'm Professor of Pediatrics at the University of Chicago.

DR. SACHS: And I'm Dr. Jody Sachs. I'm the Exec. Sec. of VRBPAC. Welcome.

CHAIRMAN DAUM: Thank you very much, committee members and temporary voting members and FDA colleagues.

We'll now turn to some noncontroversial matters for the rest of the morning, and that is consideration of FluMist influenza virus vaccine, and we'll begin with Dr. Mink, who will begin with an overview of FluMist issues.

Dr. Mink.

DR. MINK: Good morning. I'm ChrisAnna Mink from CBER. I'd like to welcome you to today's VRBPAC.

I'll begin with an overview of the product which is FluMist. FluMist influenza virus vaccine, trivalent A and B contains three strains of live attenuated, cold adapted, temperature sensitive influenza viruses; two Type A, H1N1 and H3N2, and one Type B. Each .05 mL dose contains ten to the seventh TCID50s of the three strains in normal allantoic fluid, which will be abbreviated NAF throughout my presentation.

A review of the regulatory time course is shown on this slide. The original BLA was submitted to CBER on October 30th of 2000. Our first VRBPAC meeting was July 26th and 27th of 2001, after which CBER issued a complete response letter, a CRL, on August 31st, 2001, and we received the sponsor's response to our first letter on January 7th, 2002.

Our second letter was issued to the sponsor on July 10th of 2002, and we received their response on August 26th, 2002.

On November 1st, 2002, the sponsor revised the age indication, which I'll discuss briefly, and that brings us to today's meeting.

Some changes have been requested by the sponsor on the indication side. Originally the proposed indication for the age was from 12 months to 64 years, and as I mentioned, this has now been revised to 60 months through 64 years.

Originally a request for an indication for travelers to areas where influenza viruses were circulating was included and this has been removed. Also in our history we have some unresolved concerns from VRBPAC 2001, and we are now returning to the committee.

To briefly review 2001 VRBPAC, the efficacy vote was divided for children/adolescents and for adults. For the one to 17 years of age group, the committee at that time voted eight yeses and seven noes, with five of the seven members who were voting noes stated they would have likely voted yes if they were requesting an age starting in older children, for example, 15 to 24 months of age.

The expressed concerns for children and adolescents included that there were few subjects under two years of age included in the database. We had no concurrent immunization data. There was no H1N1 field efficacy, and some committees expressed concern that we are extrapolating data for children from seven to 17 years of age.

The efficacy for the adults included 13 yes votes and two no votes for the age group of 18 to 64 years of age. Expressed concerns included that there were few subjects over the age of 50.

There were some concerns about defining a healthy population for receipt of this vaccine. No re-vaccination data was provided at that time. No concurrent immunization data were available, and there were some concerns expressed about the use of clinical endpoints, i.e., effectiveness, not confirmed with influenza cultures' efficacy data.

The safety vote was across the entire age cohort of one to 64 years of age. At that time there were five yes votes and nine no votes; subsequently, on the last day revised by the committee chair, as four yes and ten no votes.

The express concerns at that time included that the final data for some of the critical studies had not yet been submitted to CBER; the possible association of FluMist with adverse respiratory events, including pneumonia and asthma and wheezing.

Other concerns included the occurrence of other adverse events, AEs, occurring post vaccination. There were few subjects under two and over 50 for the safety database. Again, no concomitant immunization, and there was a paucity of transmissibility data.

There was also a discussion about the possibility of reassortment, including with wild type influenza, and the risk of reversion of these attenuate strains.

The current indication being sought and to be discussed today is FluMist is for the active immunization for the prevention of disease caused by Influenza A and B viruses in healthy children adolescents and adults from five years, greater and equal to 60 months, to 64 years of age. A two dose regimen, 60, plus or minus 14 days, for the first use of children five to eight years of age is being requested and one dose for all others and for those over nine years through 64 years of age.

With consideration for the revised age indication and the availability of the final effectiveness and efficacy data, as well as additional safety analysis of FluMist will all be presented today for the committee's deliberation. To help frame the day, I will present the questions.

Question one will be safety for vote. Are the data adequate to support safety of FluMist for individuals five to 17 years of age, 18 to 49 years of age, 50 to 64 years of age? Please consider data related to the respiratory events, such as asthma and upper respiratory infections, shedding and transmission of vaccine strains following receipt of FluMist, and annual revaccination.

If the data are not adequate for specific age groups or there are other safety concerns, please discuss what additional data should be requested.

Question number two is efficacy, also for a vote. Are the data adequate to support efficacy in FluMist in individuals five to 17 years of age, 18 to 49 years of age, and 50 to 64 years of age?

If the data are not adequate for specific age groups, please discuss what additional data should be requested.

Question three is a discussion point. Clinical studies for release of new strains: please comment on the design and endpoints for the clinical study performed in adults which will be presented today for the release of new strains.

And discussion point number four: if the data are adequate to support safety and efficacy, please discuss what additional information, if any, should be requested from post marketing studies.

With this orientation, I turn the meeting back to Dr. Daum.

CHAIRMAN DAUM: Questions from the committee for clarification of Dr. Mink's orienting remarks? Dr. Faggett, please.

DR. FAGGETT: Bob, under the efficacy vote, the first slide, it was stated that a vote for efficacy data for supporting indication, one to 17 years of age, five, seven no; the request was starting at older age, 15 to 24 months. I'm not sure that that's really stated correctly.

DR. MINK: That information is from last year's VRBPAC, and some of the members who voted no for the age group of 12 months through 64 years, five of the seven who voted no expressed that they would likely have voted yes if an older age is being requested, such as 15 months or 24 months of age.

DR. FAGGETT: Okay. Thank you.

CHAIRMAN DAUM: That's how I remember it Dr. Faggett. What's your concern?

DR. FAGGETT: I thought it was, well, like three or four. I thought it was over two. I didn't recall it being less than two. So I stand corrected if that's in the minutes.

CHAIRMAN DAUM: I think this is the way it was.

DR. FAGGETT: Okay. Thanks.

CHAIRMAN DAUM: Good. Let's continue then. We're next going to turn to our sponsor, Medimmune Vaccines, and begin with a presentation from Dr. Young -- this is Dr. Young? Good morning, Dr. Young -- who will give us an overview and a product profile.

DR. YOUNG: Good morning, everyone. Thank you very much.

As Dr. Daum mentioned, I'm Jim Young. I'm actually very excited to be here today. I just hit a very important milestone, and for those of you who know me, think that that probably refers to the fact that I just turned 50 years old yesterday, but it's actually not that. For those of you who are parents can appreciate the fact that I have a three months old son who just slept through the night for the first time last night.

(Laughter.)

DR. YOUNG: And 11 and a half hours. I'm so excited about that. I was beginning to forget what sleeping through the night really felt like.

But actually I'm also very excited to be here today to talk about FluMist, which we believe is an important new product. As I said, I'm Jim Young. I'm President of Research and Development at Medimmune.

We're a biotech company that's located about ten miles north from here in Gaithersburg, Maryland.

You can see from the first slide here the sponsor has changed for the product since the last time the VRBPAC committee met, when it was Aviron, and the reason for that is that in January of this year, Medimmune merged with Aviron and formed a subsidiary called Medimmune Vaccines that now actually has regulatory responsibility for the product.

What I'd like to first do is echo some comments by Dr. Daum and acknowledge the tremendous amount of work and diligence that the review staff has demonstrated during this review process. It has just been a tremendous amount of information that we've submitted to them in the course of this review, and it's really quite a lot of work to sift through all of that information and go through the review.

I'd also like to thank the committee members for finding time in what I'm sure is a very busy schedule this time of the year to actually have this special meeting to review Synagis -- review FluMist. Excuse me.

(Laughter.)

DR. YOUNG: Oops, wrong product. That's our other product.

To review FluMist and give you an update on FluMist.

CHAIRMAN DAUM: You thought I wasn't listening, didn't you?

(Laughter.)

DR. YOUNG: I'm so used to talking about Synagis.

And give you an update on the product and, in particular, address unresolved questions that were raised at the last committee meeting now that a full data package is available for the product.

As Dr. Daum mentioned, I will first begin the sponsor presentation by providing a very brief overview and product description, and then what I'd like to do is describe for you the new proposed indication that Dr. Mink briefly reviewed for you in her presentation. I'll then turn the presentation over to Dr. Ed Connor who is our Senior Vice President of Clinical Development, and Ed will give you an overview of the clinical data that supports the efficacy and safety for the product. It will also give you information about transmission and vaccine virus stability, both phenotypic and genetic stability.

You are, of course, all aware of the significant morbidity and mortality associated with influenza as it is the most common cause of medically attended respiratory illness in the United States among all age groups. It's estimated that there are about 70 million lost work days, 38 million lost school days, and an astonishing 20 to 50,000 deaths annually due to influenza in the United States.

Although about 60 to 90 million doses of the inactivated flu vaccine are administered annually to protect against influenza, most of this goes to the high risk individuals. It turns out that it's estimated that about 150 million healthy Americans are currently not vaccinated against influenza; that less than ten percent of healthy children, less than 30 percent of healthy adults actually get their flu shots, and clearly this is a major concern of the ACIP.

As many of you may know, they recently broadened their recommendation for influenza vaccine to include health infants and toddlers and their contacts. It is in this healthy population that we believe FluMist can fulfill an important public health need.

Given what we believe to be its excellent safety and efficacy profile and ease of administration such that adequate supplies of the inactivated vaccine can be made available for the high risk individuals, as Dr. Mink mentioned, FluMist is a cold adapted, temperature sensitive, live, attenuated influenza virus vaccine that's administered by intranasal mist, hence our very creative name for the product, FluMist.

It is made from vaccine strains that were originally derived by Dr. John Maassab at the University of Michigan back in the mid-'60s wherein he took an Influenza A H2N2 isolate and a Type B clinical isolate, and he independently passaged both of those isolates sequentially in primary chick kidney cells and embryonated hens' eggs at progressively lower and lower and lower temperatures.

And at the end of that process what he derived were some viruses with some very interesting properties. First, they were cold adapted, meaning they would grow in the cooler upper airway.

They were temperature sensitive so that they would have limited replication in the lower airway.

And they were attenuated. They wouldn't cause disease in a ferret model of human influenza, a property that was later also shown in human volunteers.

Now, what we do is we take these master donor viruses that Dr. Maassab generated, and we introduce into them the genes coding for the hemagglutinin and neuraminidase surface glycoproteins that come from the contemporary strains that are circulating in the population that we want to vaccinate against.

We've actually made 14 different versions of these different versions of these vaccine strains and tested them clinically.

Now, today you may also hear the term "CAIV," cold adapted influenza vaccine. These are vaccines that were derived from the same master donor viruses that Dr. Daum had generated, most of that work being done at the NIH. Nineteen of those strains were developed and tested clinically in about 8,000 human volunteers.

Now, at the last VRBPAC presentation, Dr. Brian Murphy gave a very comprehensive overview of the important and unique properties of these master donor viruses and of the vaccine strains that are derived from them, and I'm certainly not going to try to replicate his talk. But suffice it to say Brian presented information that there are at least four mutations in each of four genes for the MDV strains that confer attenuation in the ferret model.

We know that there are at least seven mutations in the A strain and eight in the B strain. We also know that these viruses are extremely phenotypically stable. In any laboratory passage study, animal studies, even clinical studies, we've never seen a revertant of these attenuated phenotype.

And, frankly, that's not very surprising because given that there are at least four mutations in the MDV that confer attenuation, at a mutational frequency of ten to the minus five for any animo acid to change, with those four mutations needing to revert back to the wild type sequence, all four of them required to revert back to the original wild type sequence, that would occur at a calculated frequency of ten to the minus 20.

And in fact, if it were all seven or eight mutations that need to revert back to the wild type sequence to generate a wild type virus, it would be ten to the minus 35 or ten to the minus 40.

Now, to put that in perspective, we know that children shed at most during their peak time of shedding ten to the four of vaccine virus. We know adults actually shed about 100 times less than that. If we assume, however, that everyone sheds ten to the four viruses when they are infected with this vaccine and then we assume that they make 1,000 times more virus in their upper airway, if we were to immunize all 300 million Americans today with this vaccine and they all produced ten to the seven virus not for just the two to nine days that they normally produce it, but forever more, for the rest of their life, it would take 100 years to produce ten to 20 viruses.

Now, this is a schematic representation of the influenza virus structure and the genetic make-up of the virus. You can see it's an envelope virus. It has eight RNA segments contained within the virus. Two of these segments, the HA and the NA, code for the hemagglutinin and neuraminidase surface glycoproteins.

The hemagglutinin is involved in virus cell entry; the neuraminidase involved in budding of the virus from the infected cell, and both of these proteins are the main targets of protective immunity.

The other six gene segments code for a number of different virus proteins that are involved in virus replication and immune evasion.

Now, when we want to make a new vaccine strain, what do we do? We start off with those Dr. Maassab's master donor viruses, and we co-infect cells with the master donor virus chick embryo kidney cells. We co-infect cells with the master donor virus and the new wild type strain that is circulating that we wish to make a vaccine against.

Now, when these viruses go into the same cell, they begin to replicate their RNAs, and they actually end up shuffling them, and the progeny virus that comes out of those cells has various combinations of these different genes from the two viruses, and there are actually 256 different combinations that are possible.

What we do is we then go and fish out a virus that is what we call a 6:2 vaccine strain that derives two genes, the hemagglutinin and neuraminidase genes from the wild type virus because we want to make immune responses to those two proteins, and the other six genes come from the master donor virus which confer the attenuated properties of this master donor virus on this new vaccine strain.

Now, once we've derived that 6:2 vaccine strain, what we then do is propagate it or manufacture it in specific pathogen free eggs. Now, these eggs come from chickens that have been extensively tested to demonstrate the absence of any adventitious agents in those blocks.

We also have an extensive testing program of the vaccine intermediates, the bulk vaccine virus, and the final product.

And as Dr. Mink mentioned, we also do a safety testing of the new vaccine in 300 adult volunteers to insure that it is of the right attenuated phenotype.

As she also mentioned, the vaccine is comprised of a trivalent blend of ten to the seven infectious particles of each of the three currently circulating influenza virus strains that are also recommended for the inactivated vaccine, and it's contained in half an mL dose. This product is presented as a unit dose, which is stored frozen with no thimerosal in this little sprayer device.

And the way this works is you basically take this sprayer and thaw it in the palm of your hand. You then remove the cap off the sprayer, insert the end of the sprayer up into the nostril and depress the plunger, and that delivers half the dose to one side of the nose.

You then take off this little dose divider clip, insert it into the other side of the nose and press the plunger and complete the administration of the dose.

Now, the sprayer actually generates a large particle mist that is deposited in the upper airway where the virus will replicate. As Dr. Mink also mentioned, there is a single annual dose recommended for individuals that are nine years or older, and it is recommended that a child that is less than nine years old receive two doses spread 60 days apart if it's their first time for receiving an influenza virus vaccine.

Now, at the last VRBPAC meeting in July of 2001, it was felt that the data were adequate to establish efficacy in healthy individuals, and as was pointed out -- I think this was right -- the vote that was actually for children one to 17 years was eight to seven in favor of having shown efficacy and in adults 18 to 64, 13 to two.

However, it was noted that the sample size at the lower end of the proposed age spectrum in individuals 12 to 24 months was less robust, and that had the indication for children been two to 17 years, the vote would have been 13 to two, and that was reviewed by the Chairman, Dr. Daum.

When the question of safety was asked in the total population requested, age one to 64, the votes was four to ten with six of the votes qualified as provisional, and that was mainly because the final safety data had not yet been submitted to the agency and, consequently, the CBER review was still ongoing.

It was also noted that there were additional analyses that needed to be completed to resolve some safety questions and particularly the ones with respect to pneumonia and asthma. It was also noted that concurrent immunization data were not available and was needed for children under 24 months of age.

Now, what progress have we made since the last VRBPAC meeting? Well, first of all, we've responded to two complete response letters from CBER, and in so doing established the 20 studies which would serve as the final data set for consideration in the BLA. Fourteen of these were placebo controlled trials and six were open label trials.

In total, the data from these studies represented a total of 20,000 subjects receiving approximately 28,000 doses of FluMist.

We also submitted final study reports for the two very large safety studies involving about 15,000 individuals, 15,000 children, AV019 and AV012, and also the final study report was submitted for a Finnish day care study that provided some information on transmission and virus shedding.

And when the final analyses were conducted on all of this data, it showed quite clearly that there was no signal for pneumonia and that there was a possible signal for asthma or wheezing exacerbation in children less than 60 months of age.

At the last VRBPAC meeting we also indicated that a concurrent immunization study had just been started with FluMist and MMR and VARIVAX, and we've actually recently completed full enrollment in that study and we're now in the final follow-up period with those children.

So for consideration today at today's meeting, here is the proposed indication. For active immunization to prevent Influenza A and B in healthy individual, healthy children and health adults ages five years through 64 years of age.

We are not proposing FluMist be used in individuals with a history of underlying medical conditions which predispose them to bad outcomes with wild type flu, and those are listed here on the slide.

We're also not proposing that the product be used concurrently with any other vaccine, frankly, until we have the data that would support that use.

So with that I will conclude my introductory comments and now turn the presentation over to Dr. Ed Connor, who is going to give you a review of the clinical data supporting the efficacy and safety of the product.

Ed.

CHAIRMAN DAUM: I think maybe before we call on Dr. Connor we might see if there are some clarifying questions from the committee. We'd like to keep questions and comments at this point to issues that require clarification of what you said, and then we'll have opportunity for a more general discussion of sponsor related issues after Dr. Connor is complete.

Dr. Snider, please.

DR. SNIDER: Thank you.

Two relates questions. I appreciate your reminding us about the particle size and 98 percent being larger than ten microns. I presume the other two percent has the potential for reaching the lower respiratory tract, although not all of it will.

The question is: do we know any more than we did in 2001 about the distribution?

And then very much related to that, you spoke about, and I guess we'll be talking more about, shedding. I know that nasal swabs and one would anticipate that most of the shedding would be from the nose where most of the virus is deposited. But is there any information about the virus present in the lower respiratory tract?

DR. YOUNG: I think about all we can say about that is based on the scintigraphy studies that we've done where we've labeled material and then deposited it into the nose, and we've done radiation surveys of various components of the abdomen and have shown that the vast majority of it, as you mentioned, ends up in the upper airway.

And when we do field screens of the lungs, we see what we think is just background radiation coming from the esophagus where some of the vaccine actually hits the back of the mouth and is swallowed into the stomach. We see some radiation in the esophagus that over time quickly moves down into the stomach.

And what we've seen in the scintigraphy studies is that if you use nasal drops, you see the same amount of radiation in the lungs. So I don't think that we believe that there is very much of this that actually gets down into the lung, and in fact, even if there is a small amount, given its temperature sensitive phenotype, we would expect very little replication in the lung itself.

CHAIRMAN DAUM: Ms. Fisher.

MS. FISHER: You made a statement that children shed vaccine virus more than adults. Do you know why?

DR. YOUNG: Probably because of the lack of any preexisting immunity. Adults, of course, will have had numerous encounters with flu virus over the course of their lifetime, and there is some low level immunity that's going to be protective against the virus that would suppress replication of the virus in adults.

I think it turns out that probably young cells are probably a more fertile environment for the growth. We tend to see better replication in younger cells, not that we have old cells, but I think it's a combination mainly driven by the lack of immunity in children where they can shed more virus for a longer period of time.

Adults shed for only about two days. Children will shed for on average about nine days, and as I said, about 100 times more virus than adults.

MS. FISHER: Well, then if everyone uses the -- the children use this vaccine, then they'll grow up to be adults that will continue to shed more virus, right?

DR. YOUNG: No, actually they will get immunity to the virus obviously.

MS. FISHER: they will?

DR. YOUNG: And then, of course, they will shed less virus as well.

MS. FISHER: As long as they keep getting vaccinated.

DR. YOUNG: No, actually, well, they'll need to get vaccinated against the new contemporary strains that are circulating because it changes those two surface glycoprotein and you'd need to reeducate or educate the immune system to those new surface glycoproteins.

But once you've encountered flu, you know, the risk is lower because you have some cross-protective immunity, but you need to get better immunity against the contemporary strains that are changing in the population.

MS. FISHER: Thank you.

CHAIRMAN DAUM: Dr. Markovitz, please.

DR. MARKOVITZ: Yes. What can you tell me about how you pick out the proper reassortment virus? You know, you're looking for the 6:2 mix. How do you know that you have the right six genes and the right two genes?

DR. YOUNG: Yeah. We actually have a pheno-genotyping assay that we use. It's a RFLP type assay where we can actually -- what we first do is once we have the mixture of viruses, they get passaged in eggs in the presence of antibody to the master donor virus to suppress replication of any residual master donor virus, and so what we get out are mainly viruses that have the hemagglutinin and the neuraminidase from the new wild type strain.

We then clone those viruses out, dilute them out and get individual clones from the progeny from that co-infection, and then we genotype each of those clones, and we have a specific assay where we PCR the gene segments and cut them with restriction enzymes that are specific for either the master donor virus version of the gene or the wild type version of the gene.

What we actually do as a prerequisite to that is we sequence the entire genome of the new wild type strain that we're going to make a vaccine for so that we know what restriction enzymes to use and what primers to use to pull out the wild type gene specifically.

CHAIRMAN DAUM: Very good.

DR. YOUNG: Anything else?

CHAIRMAN DAUM: Dr. Stephens, we're looking for comments here to clarify issues raised by Dr. Young. We'll have an opportunity after Dr. Connor speaks to explore some of these issues that committee members are raising in detail per your pleasure.

Dr. Stephens.

DR. STEPHENS: This is a clarification regarding the reversion rate, which you suggest is low, less than minute ten to the minus 20.

What about reassortment, which is more of a concern with wild type and where I think the frequencies would be considerably higher?

DR. YOUNG: Yeah, with respect to reassortment, there's really two different situations you need to consider when you think about reassortment. The first is during an interpandemic period where we're just immunizing against strains that are already circulating in the population. There's actually lots of reassortment going on between the A strains that are circulating now. They've seen H1N2s. So that happens. We know that happens, and that can happen if and only if an immunized individual also has a wild type flu infection at the same time.

And you're right. If that happens, you can get a reassortment between FluMist and that wild type strain. But you need to remember the FluMist comes from a human strain that's already been in the circulation; that we have reassortment that has occurred between strains that are already circulating, and at worst, what you can get back out is a wild type strain because the cold adapted genes that are in the attenuated virus can only make the wild type attenuated.

So at worst if you didn't attenuate the wild type, you'd get back out wild type, and that's going on all the time anyway. You've got that circulating all the time anyway, and so you would have already put that wild type into the person anyway.

So in terms of normal epidemic vaccination against normal epidemic strains, the risk of generating a super virulent strain is virtually impossible because the genes that we have in the FluMist virus are attenuating, and we know that from data that Dr. Murphy presented last time. When you just put individual genes and have done the experiments where they put individual genes from these master donor viruses into wild type strains, you generally get a strain with lower virulence and not higher virulence.

Brian, I don't know if you want to make a comment about that. You've certainly done more of this type of work than anyone else in the world.

DR. MURPHY: Although I don't have the exact data, we probably have made between ten and 15 different mixed gene constellation viruses from the master donor A virus, and none of those showed a virulence that was greater than wild type. Almost every one of them had an attenuated phenotype.

DR. YOUNG: Thanks, Brian.

Now, the other setting where you worry about reassortment is if we wanted to use FluMist as a vaccine for a new pandemic stream before the virus was actually in our population, and clearly you wouldn't want to do that because now you could get reassortment between the new pandemic FluMist with the H1N1 or the H3N2 viruses that are circulating now and prematurely introduce the pandemic virus into the population.

So clearly, you would want to want to wait until the public health authorities deemed it appropriate to now start immunizing with a FluMist pandemic strain because you felt that the risk of spread of the pandemic virus was so significant that you wanted to try to immunize people as quickly as possible and the virus was already in the population anyway.

Okay? You don't look satisfied.

DR. STEPHENS: Well, I think the point was the issue in recombination. I mean the issue is reassortment and recombination. It's probably not reversion. I think you've pretty well demonstrated that that's very, very low, but I think there are issues with reassortment that we need to talk about further.

CHAIRMAN DAUM: Dr. Decker.

DR. DECKER: I've got several questions that follow up on Dr. Stephens' questions. The vaccine virus differs from circulating wild virus in two ways. It's cold adapted and it's attenuated. Are those --

DR. YOUNG: And it's temperature sensitive.

DR. DECKER: Okay. thank you.

Do those three characteristics reside in different or in the same genetic changes? In other words, is one set of changes simultaneously making it cold adapted, temperature sensitive and attenuated? Is it in two of those three and the third is based elsewhere, or are all three genetically distinct?

DR. YOUNG: The answer to that question is that they probably are overlapping. I don't think that given the three different phenotypes that they are al in the same mutations. We know actually that there are five mutations for temperature sensitivity in the A strain, four mutations in the B strain.

We have now generated viruses that have all of the single point mutations distributed singly in viruses and are testing those for attenuation and cold adaptation. So hopefully in the not too distant future we will have sorted that out, but my guess is that there's probably some overlap that the temperature sensitivity is also related to the attenuation.

DR. DECKER: All right. Your response to Dr. Stephens focused on the fact that a reassortment with the circulating strains would just present the same neuraminidase and hemagglutinin antigens that everybody is seeing anyway, which is fine. That's reasonably straightforward.

But the question that I was hearing that I didn't hear an answer to is would it be possible for a reassortment or similar genetic combinations to, for example, produce a virus that is cold adapted, no longer temperature sensitive, and no longer attenuated and which, therefore, could exploit the human ecologic niche more effectively an the current virus and create something that would pose a new medical issue.

DR. YOUNG: No actually what you would then get is the wild type virus back. That's already ‑-

DR. DECKER: No, the wild type virus isn't called adapted. So you'd produce a virus that could replicate both in the nose an din the lung.

DR. YOUNG: Actually a lot of the wild type viruses are cold adapted, and actually a number of them are also temperature sensitive.

But I think the point is that most of the mutations -- there are four different genes for attenuation. So automatically if those four genes, any one of them, ended up in the wild type virus, it would attenuate the virus, period. It would just attenuate it.

The other two, we're not sure yet if any of the mutations in those are related to cold adaptation or temperature sensitivity, but it would be certainly no worse to get one of those gene than if a reassortment occurred between the genes for an H1N1 virus and the internal genes for an H3N2 virus.

CHAIRMAN DAUM: Thank you.

I'd like to move on now to Dr. Young's presentation, and then we can have a general discussion where committee members can explore these issues further should they wish.

I would like to call on Dr. Connor, who is the next Medimmune speaker, and hearing cell phones and beepers going off in the room, I would like to ask once again that everyone in this room be respectful of the committee deliberation and either turn them off or turn them to vibrate mode.

I thank you for your cooperation.

Dr. Connor.

DR. CONNOR: Thank you, Dr. Daum, and good morning.

First of all, this committee has heard in some considerable detail the data about the safety and effectiveness and efficacy of FluMist at the prior VRBPAC presentation. You heard from the principal investigators of each of the pivotal studies, Dr. Bob Belshe, Dr. Steve Black, and Dr. Kristin Nichol, the data on adults and children, safety and efficacy. And those individuals, as well as representative folks from the former sponsor presentation at the last VRBPAC, are here with us today. Steve actually couldn't make it today because he was ill, but the other folks are here, as well as Tony Piedra and Manju Gaglani, from Baylor who conducted the ABO 12 study, the large safety trial of FluMist in the field.

That body of data is also supported by what is literally decades of research on the predecessors of FluMist, the monovalent and divalent cold adapted vaccines.

My purpose this morning is really to do three things. First of all, what I'd like to do is to review the efficacy-effectiveness data briefly with you, as well as the safety of FluMist in children and adults, and to provide the final data set with regard to the specific open safety questions that were included in the last meeting. Those specifically include asthma and wheezing and pneumonia.

And lastly, I'll provide to you some data on the vaccine virus shedding and transmission issue.

The principal studies that were conducted to support the efficacy-effectiveness of FluMist in children include the pivotal trial, which was AV006, the study conducted by Dr. Belshe, and was a field trial, a two year field trial, demonstrating efficacy for H3N2 and for Type B.

In addition to that, because H1N1 was simply not circulating during the years in which AV006 was conducted, a challenge study was done with H1N1 using the vaccine strain and demonstrated efficacy of 83 percent.

In adults, the initial trial of efficacy were done by John Traynor, and that study, 003, was a wild type challenge study demonstrating efficacy of 85 percent.

AV009 was the pivotal field trial demonstrating effectiveness of FluMist in adults, and those studies conducted by Kristin Nichol looked at effectiveness measures, that is, disease and illness measures as primary outcomes.

I'm going to focus my review time on the pivotal efficacy trials.

AV006 was the pivotal efficacy trial in children, as I mentioned, conducted by Bob Belshe. It was a randomized, double blind, placebo controlled trial of 1,602 healthy children between the ages of 15 and 71 months of age at entry.

These children received either one or two doses of FluMist in year one, and an annual revaccination dose in year two as a single dose.

There was active surveillance for illness and illness cultures, and the primary endpoint of the trial was culture proven influenza.

During the time that the trial was conducted in year one there was an A/Wuhan H3N2 and a B virus that were well matched to the vaccine, and during the second year an A/Sydney H3N2 circulated, which was mismatched.

The primary efficacy results are shown on this slide and clearly demonstrate that FluMist efficacy of approximately 93 percent against culture confirmed influenza, and what you can see here is 92.6 percent efficacy in year one and comparable efficacy for other outcomes that were measured in the trial.

As I mentioned, in addition, this study gave us the opportunity of looking at annual revaccination, and here in year two when there was a mismatched strain circulating, one can see efficacy point estimates of 87 percent comparable to in year one; in addition, comparable efficacy in the other outcomes that were measured.

One of the things that's important to note about these studies also was that there were very tight confidence intervals around the point estimates of the effect in both of the years.

In addition, we then looked at the AV006 trial to garner data regarding the efficacy of children in the various age groups of patients that were studied, and here what you can see is the efficacy by age group through each of these various age cuts.

And first of all, the point estimates of the efficacy are quite good compared to the population as a whole. The confidence intervals are quite tight, even within these age subgroups, and for children above 60 months of age, there was point estimates of efficacy of 90.6 in year one and 86.9 in year two.

It was also true that children who entered the second year of the trial actually had aged to 83 months by the time that they entered, and I have provided those estimates of efficacy here also.

You can see, again, that the confidence intervals are quite tight around the point estimates of efficacy, and there's no trend toward any changes in the efficacy, and there's no trend toward any changes in the efficacy according to the age group of the patients that were studied.

The primary trial, the field trial for adults was AV009, as I mentioned. This was a trial that was randomized, double blind, and placebo controlled. It enrolled 4,561 healthy working adults. That included ages 18 to 64 years.

Patients received a single dose of vaccine, and the primary endpoint for these trials was effectiveness measures. The primary endpoint was actually any febrile illness, and there were a number of secondary measures of effectiveness, that is, disease/illness outcomes in contrast to culture proven outcomes.

The secondary endpoints are listed here, and there were a number of other illness definitions that were also measured: febrile URI, severe febrile illness, which were pre-specified, as well as post hoc analyses that were conducted with the CDC definitions for influenza like illness and the Department of Defense definitions.

During this year of AV009, the circulating strain was A/Sydney, which was an H3N2 and was mismatched to the vaccine strain.

This is a list of the various illness definitions that were used during the conduct of the trial, and as Dr. Nichol mentioned in her presentation to the VRBPAC last year, one can see that any febrile illness, while it's actually a fairly sensitive measure of detecting disease, it's not a very specific measure for influenza.

And each of these other definitions which include various combinations of more fever or consecutive days of symptoms are actually much more representative of influenza like illness, and so that distinction is listed here. You can see each one of the combinations for each illness that were specified.

Now, these are the primary outcome results for the AV009 trial, and what you see here are a percent reduction, FluMist compared to the placebo group, for each of the illness definitions, and here what we're showing is occurrence of those illnesses.

And while there was not a statistically significant difference in the groups for any febrile illness, there were statistically significant in all of the other outcomes that were more specific for influenza.

These activity levels, that is, a 25 percent or so reduction in effectiveness measures are actually very impressive because remember that effectiveness is measuring the total disease burden, and influenza represents some portion of that.

So in order to be able to effect effectiveness changes one needs to have very high levels of vaccine efficacy. These actually have been studied for TIV in which CDC investigators have demonstrated that effectiveness measures or reductions of about 34 percent are associated with vaccine efficacy, that is, culture confirmed efficacy, of approximately 89 percent, and Bob Belshe showed previously in the AV006 data that in children effectiveness measures of a reduction of approximately 20 percent were associated with culture confirmed influenza reductions of approximately 95 percent.

So these are very highly impressive results in terms of measures of effectiveness.

These data are the percent reduction for the other measures that were looked at in the study: days of illness, days of missed work, health care provider visits, and days of antibiotic use, and you see here the same percent reduction with the various illness definitions shown at the bottom of the slide.

What you can see here are significant reductions in all of these parameters for virtually all of the disease designations and even for any febrile illness the less sensitive or less specific diagnosis, there were significant reductions in days of illness and days of antibiotic use.

Now, in addition, one of the issues that we've been asked to address through interactions over the last months has been the question of the 50 to 64 year old population within the AV009 study.

And what we show here, first of all, obviously in doing this we're looking at subset analyses in a study in which the study was obviously not designed or powered to look at those subsets. But importantly, if one takes the 50 to 64 year old population, which is shown here in blue, compared to the total adult population in the AV009 study, what we're looking here for is evidence that the 50 to 64 year old population is somehow different than the population as a whole.

And what you can see is, while in occurrence we didn't see differences in the 50 to 64 year old population, for each of the illness definitions actually even including occurrence, when the DOD-ILI designation was used, we saw differences in each of the groups comparable at least to the population as a whole.

And the other aspect of this analysis is that when you look at disease severity measures, particularly days of missed work and health care provider visits, there are highly statistically different differences observed both in the 50 to 64 year old population as well as in the population as a whole.

So the general trend and pattern within that age group is consistent with us not seeing evidence that that population was substantially different than the population as a whole, the randomized population.

That's actually illustrated here again with data that Kristin Nichol provided, which shows the analyses for febrile URI, one of the more specific influenza diagnoses, and here what is shown is the point estimates and the confidence intervals for the all patient population, 50 to 64 year olds, and 18 to 49 year olds.

And what you see, first of all, is for this illness as a nation, and there are others that are very comparable to this, for each of the measures of effectiveness you see that, first of all, the point estimates of each of these are very tight. They're very close to each other, but the confidence intervals are essentially all overlapping, and when they weren't overlapping or when the point estimate was outside of the confidence interval, it was actually higher for health care visits for the 50 to 64 year old.

So I think in looking at that kind of assessment, one can glean evidence to suggest that or, rather, we don't have evidence to suggest that there was any significant difference between the 50 to 64 year old population and the population as a whole.

So from an efficacy-effectiveness perspective, we believe that we've demonstrated that FluMist was highly effective in the prevention of influenza in both healthy adults and in healthy children.

We've also demonstrated that efficacy and effectiveness was observed across all of the age subgroups that were studied in these trials. Efficacy in children greater than 60 months was similar to the population as a whole and to the younger children on AV006.

We've also demonstrated that effectiveness in 50 to 64 year olds was similar to the effectiveness in the randomized group, albeit that population was a smaller group. The confidence intervals were somewhat greater. They, by and large, represent in that study about ten to 11 percent of the population as a whole.

And lastly, field efficacy was established for H3N2 and for B strains. H1N1 just didn't circulate during the hears that the study was conducted. Challenge studies in children with the vaccine strain and adults with wild type H1N1 support that activity.

In addition to that, previous field studies with the predecessor cold adapted vaccines that were done in various settings, including a Kathy Edwards site and by Kathy Edwards, demonstrated the H1N1 efficacy in that population.

And lastly, we've also been able to demonstrate comparable efficacy after annual revaccination in the AV006 study.

I'm going to turn now briefly to a discussion of the safety data and specifically initially one should remember that the safety population that's included in this analysis include approximately 16,000 healthy adults or healthy children, rather, and there are about 3,000 who were revaccinated.

The population includes approximately 4,000 healthy adults in the safety database.

What I'm going to review for you is the general safety of FluMist, including the SAEs and mortality, reactogenicity, and medically attended events, and then specifically turn to the issue that were of interest to the VRBPAC last time, particularly asthma and wheezing and the other issues that are listed on this slide.

First of all, from the big picture perspective, we're looking here at mortality and SAEs. There were two unrelated deaths in the 20 studies that were submitted for consideration in this BLA in the FluMist group. Both of those deaths were unrelated to the vaccine. One was a drowning in an adult, and the other one was a posterior fossa tumor in a child.

The SAE rates that you see here were low and similar in the FluMist and the placebo group. These are the SAE rates for children, percentage of patients with SAEs for one to four years, five to 17 years, and the entire pediatric population, as well as in open label trials. These are placebo controlled trials. These are open label trials for reference.

For the adults the SAEs are broken down by 18 to 49, 50 to 64, and 18 to 64, with the open label studies over here.

There was a numerical increase or numerically higher percentage of SAEs and placebo controlled trials in 50 to 64 year olds. Those are completely explained by accidental injury, hospitalization for previous illness or surgical hospitalizations, and so we saw no difference in any of the age groups for SAEs.

The next series of slides are going to consider reactogenicity events, both in children and in adults, and what we've plotted here are the percent of patients with these solicited AEs or reactogenicity events for children within ten days of a vaccination.

And here what is presented in dose one. In the placebo controlled trials, that included children between 12 and 71 months of age.

Here what we see are across the bottom of the slide various reactogenicity parameters: cough, runny nose, sore throat, et cetera, various measures of temperature, and a constellation of symptoms included in the CDC-ILI definition is listed at the end of the slide.

What we see here are the reactogenicity events collected on diary cards, and you can see that there was a small, but statistically significant increase in mild URI symptoms after dose one. When you look at the constellation of these symptoms and think about them in terms of CDC-ILI definition, for example, there was no suggestion of an increase or statistically significant difference in those parameters.

And if you think about temperature as one of the measures of severity of reactogenicity events, there were no differences between the two groups in terms of temperature greater than 101.

If you look at the adult pattern for the same reactogenicity events here for seven days, which is how they were collected on diary cards through the studies that collected reactogenicity, one sees that there's a slightly different pattern of the type of events. Some were events in which adults report rather than children, like sore throat or those sorts of things, but basically there was the same sort of pattern of mild URI symptomatology that was increased following FluMist administration.

But when you look at fever, there was no difference between the two groups whatsoever, and when you look at constellations of symptoms that are a surrogate for sort of looking at flu-like illness, those rates were extremely low and comparable between the two groups.

Now, another measure of looking at severity of reactogenicity events is to look at the medication use, and so we looked at medication use within the ten days of the vaccination period or the reactogenicity period for children here. What you see is a small increase. The delta here is 5.3 percent of anti-pyretic analgesic use within the ten days post vaccination in children, and you see no other differences in any of the other rates.

These differences were not seen after dose two. I should mention that after dose two the reactogenicity events were actually quite a bit lower, and there were no statistically significant differences in that population.

This is the parallel medication use slide for adults within seven days of the reactogenicity period, and here you see no differences between the two groups, and FluMist was not associated with an increase in medication use in adults.

Now, of interest here are age groups of children, and so we looked at the reactogenicity events that are included in children between 12 and 59 months of age, and between 60 and 71 months of age in the various studies, and what one sees is that, in fact, in the older children there were no statistically significant differences between any of the reactogenicity events, and overall the reactogencity events were lower in the children that were 60 to 71 months.

This is the comparable slide for adult breakdowns by age, and here you see 18 to 49 years of age and 50 to 64 years. You can see here that the pattern of adverse events or reactogenicity events in the population of 50 to 64 year olds was either similar or lower to those of the younger adult population.

Another topic of interest for deliberation is the safety of annual revaccination, and the primary data for that comes from the AV006 trial, and what we show here are children who receive -- this is a cohort of 642 children who received annual revaccination for two years within AV006 and then ultimately for a third year in an extension trial of AV006. These are the same children who received the vaccine in each of three consecutive years.

And what you see is a pattern of reactogenicity events that were lower on annual revaccination. The reactogenicity profile was similar during the second and third year of annual dosing, and overall the rates were somewhat lower compared to the primary immunization.

So from a reactogenicity perspective, I think we can say that FluMist is associated with mild URI symptoms in both children and adults; that there was no significant increase in acute influenza-like illness associated with the administration of the vaccine; and there was no increase in fever greater than 101 following FluMist administration in either children or adults.

Reactogenicity events were lower following annual revaccination.

I'm going to turn for a moment to medically attended events, and these events are seen primarily in the large safety study, AV019, which was conducted by Steve Black at Northern Kaiser in California.

In this study, this was a randomized, double blind, placebo controlled trial. It looked at the safety of FluMist in approximately 9,700 children. There were two doses given between one and eight years and one dose in nine to 17 years of age. The outcomes were ascertained from the diagnoses in the HMO database, and the primary outcomes were medically attended events and SAEs within 42 days.

This study made many comparisons. The comparisons were all of the diagnoses identified in the HMO database by setting dose, age group, and diagnosis, and there were more than 1,500 comparisons that were made in this analysis without statistical adjustment.

What you see here are the settings in which the evaluation was conducted: emergency room visits, clinic, hospital, and combined settings; dose one, dose two, and combined doses.

And then for pre-specified age groups in the protocol one to 17, one to eight, nine to 17 years, and then 18 to 35 months and 12 to 17 months.

What you see here are some of the prespecified group diagnosis results, and what you see are acute respiratory events, systemic bacterial infections, acute gastrointestinal events, and rare events potentially related to wild type influenza and what we show in many of these and the subsequent slides are the actual occurrence in the FluMist and placebo group, the calculated rate per 1,000 patient months, and the binomial relative risk with 90 percent confidence intervals, and in these studies the lower bounds of the confidence interval of one or above one is considered statistically significant.

So in these group diagnoses we saw no significant increases in any of those events associated as a medically attended event in the Kaiser trial.

Now, as I said, there were a number of analyses that were conducted during these studies, and when you look through all of the MAE analyses, here were essentially 14 MAEs, or medically attended events, diagnoses that were statistically significantly increased in the FluMist group and 21 that were decrease in the FluMist group.

So there are multiple comparisons. It's expected that you're going to see some of those comparisons by chance alone.

We then went through all of the event rates that were higher in the FluMist group, and when you did that and looked through the analysis of both temporal distribution as well as what the events were and other analyses, one identified three events for which the rate was significantly increased in the FluMist group. There was biological plausibility associated with the event, and we could not exclusively exclude a cause and effect relationship for those events, and those are primarily your eye events, musculoskeletal pain events, and asthma.

These are the actual results of the MAE analyses for the prespecified diagnoses within the Kaiser trial, and what we see is here are the upper respiratory tract infection events between one to 17 years, one to eight years, 18 to 35 months. These are prespecified age cuts in various settings, as I mentioned earlier; doses, either one or combined; and the rates and binomial relative risk.

So we saw statistically significant increase, but low rate events, low difference events between the two groups for upper respiratory tract infection and musculoskeletal pain. These are both events that we saw as reactogenicity events in the reactogenicity analyses.

When you look, for example, for URI among these various age groups, and particularly for the group of interest, which is five to 17 years in the combined settings and combined dose, one sees a rate of 25.4 percent in the FluMist group and 29.9 percent in the control group.

So clearly, as a medically attended event there a measure of, an additional measure of severity of reactogenicity. We didn't see any differences in the age group of interest for that particular event.

And the story was a little bit different in terms of asthma and wheezing, where here what we saw was for the designated term "asthma" in the database, we saw between 18 and 35 months of age in the settings of the clinic or dose, and this is mostly driven by the dose one results. One sees in terms of the rates and binomial relative risk statistically significant differences in this population.

And when you look more carefully at those populations, there still is an issue that we'll speak about in terms of a potential signal.

So we've identified a potential signal in this kind of study. The study was not obviously designed specifically to look at asthma and wheezing or those types of outcomes, and so we looked further, particularly in this population, for other evidence of an issue.

What we did were several analyses. First of all, we searched the database and the medical records for a variety of things. First of all, to identify the fidelity of the use of the term "asthma" and "wheezing" in the population. And what I mean by that is that obviously what we designated as being significant in the Kaiser trial was the specific term "asthma." And as we all recognize, particularly in young children where there are intercurrent illnesses that are associated with wheezing and where the diagnosis of asthma is actually usually not made until a little bit older, in the older age group, we look to see what the fidelity of the use in the database was.

And as expected, we saw that there was overlap between the term "asthma" and wheezing, particularly in the youth children. So whether the analysis is asthma plus wheezing or asthma alone, which was where the signal was initially identified, is the issue here.

We also assess the circumstances of medically attended events, and we look to identify in the younger children particularly whether the signal that we were seeing was related to whether or not they had a prior history of asthma and wheezing.

In this trial, the children with asthma and wheezing or particularly with asthma were excluded based on the parents' recognition of a diagnosis of asthma specifically in their child. So clearly, because of the overlap and because of the use of the terms, it is not unexpected that a number of children were in the trial who were recognized in the database as having had some previous episode of either wheezing or asthma, but the parents may not have recalled that asthma typically was the diagnosis that was given to the child.

So we looked in these additional analyses using the asthma term alone or asthma wheezing and outcomes. We looked by cumulative six months age groups. So we began at the younger age group and looked cumulatively across the age to see whether we could identify a place where the signal was detected.

We looked at dose one and dose two and across the younger age groups with history positive and history negative children. These were another 800 analyses that were done in the Kaiser study, and what we identified was only one cumulative age analysis from AV019 which was statistically significant. That was the 12 to 59 age group, depending on how you look at it. Sometimes it was 12 to 53, the conservative 12 to 59 group.

For the whole population using asthma as the diagnosis, it had a relative risk if 3.53 and a lower bound of the confidence interval of 1.1.

When you combine terms and use more inclusive terms, such as asthma and wheezing combined, the relative risk declined to 1.58, and that difference was not statistically significant.

If this signal in the overall population is correct, the absolute increase in the FluMist group between 12 and 59 months is approximately .4 or .5 percent increase if the signal is correct.

And we also looked carefully at the group above 60 months to be sure that there wasn't a signal that was there. We looked across 60 to 107 months, which is the eight to nine threshold when one or two doses are given and all the way up to 17 years in which we found no increases in the relative risk. All of these relative risks are less than one, and none of them were obviously significant.

So that we found no signal in the children who were equal to or greater than 60 months.

When we went back to explore in the younger children the question of whether or not this was an event that was isolated to the history positive children, what we found was that as we looked at history positive patients we saw increased relative risks. None of these were statistically significant compared to the history negative children, but in both groups there were increases in the relative risks in the population.

As expected, the delta rate, the change in the rate was higher for somebody who had a prior history of wheezing compared to somebody who did not, and so fundamentally we saw no significant difference, but increased relative risk in both the history positive and history negative children. The absolute rate was higher in the history positive children.

We also then looked at the other studies from which we can gather data related to asthma and wheezing. In AV006, the pivotal trial, we looked for cumulative age analyses and identified no significant increase for asthma and wheezing in these type of analyses in that trial.

AV012 was a large field trial of FluMist safety. It is a non-placebo controlled trial in which the comparative analysis is the pre-vaccination period with the post vaccination period, and I think that while the interpretation, because of some issues related to the methodology comparing the pre and post vaccination period are complex, I think what we can say is that the rates from AV012 in the population of interest was not inconsistent with the rates that we saw with AV019 in the younger population. The rates were similar to the FluMist group in the AV019 patient population.

So I think what these trials tell us is that primarily AV019, which is the largest in placebo controlled trials looking at these issues, is sort of the primary place in which we can explore the issue of asthma and wheezing.

From the perspective of hospitalizations, there were two children that were hospitalized for asthma and wheezing. One of them was in the FluMist group in AV006, who was hospitalized for a day. The other was a placebo child in AV002, who was hospitalized twice for a day. Both of these children had a history of asthma.

There were no hospitalizations in the AV019 population for asthma. The events that were recorded were events that were out-patient visits or emergency department visits. Those visits were associated with medication use at the time of the visit.

So from an asthma and wheezing perspective, I think we can say that among all of the analyses of large placebo controlled trials in children, a statistically significant difference was observed in AV019 only, and that the children for children 12 to 59 months, the relative risk was 3.53 for asthma. The rate was higher for history positive children compared to history negative children, and we didn't see a signal in the children that were greater than 60 months of age.

I'm going to turn very briefly to the issues of some of the other open issues that were discussed during the last VRBPAC meeting. From the perspective of conjunctivitis, this was an event that was evaluated in AV019. Conjunctivitis, we identified a temporal association with vaccination within the first 14 days.

What you can see here is a statistically difference for the 25 to 48 month category, but not any other age groups, and for children that were above 60 months, there was an increased relative risk, but not statistically significant.

If you take this relative risk and look for what the impact of that is, it's about a .1 percent increase in the FluMist group in greater than 60 months. These were all mild, self-limited episodes of conjunctivitis.

Pneumonia was a topic of considerable discussion at the last VRBPAC meeting. I think that what we can say now is that FluMist was not associated with an increased risk of pneumonia. These are the rates of pneumonia in the all pivotal trials for both the FluMist group and the control group or the placebo group for all pivotal studies, for children and for adults, and you can see that there are no differences between the groups across the studies for pneumonia.

So that across the final analysis data set there was no increased risk of pneumonia in those children or adults.

We also assessed the risk of CNS events. We saw no. In all of the studies we saw no cases of encephalitis, Guillain-Barre, Reye's Syndrome, or other rare disorders that are associated with wild type flu. Other CNS events occurred at a low rate. There was no significant difference in CNS events in AV019 within the 42 day period. The events that were recognized were essentially seizures or seizures associated in a child with epilepsy, and the rates were comparable in the FluMist and in the placebo group.

Concurrent immunization was also a topic of discussion at the last VRBPAC meeting. As Jim alluded to, they are currently excluded from the proposed label, and with the indication requested being five years of age and older, the logistics and management of concurrent immunization is deemed not to be a significant problem in that age group.

We are very much committed to doing the additional trials of concurrent immunization in children. We have an ongoing trial of MMR and VARIVAX. That trial, which was reported at the last VRBPAC, is now fully enrolled with 1,251 children, and we have a number of other trials that are planned and are in discussion with CBER for other childhood vaccines in the younger age group.

Lastly, I just want to turn to the issues of vaccine virus shed, of the shed of vaccine virus, as well as vaccine virus transmission.

As Jim mentioned, there's considerable amount of evidence from the predecessor cold adapted vaccine to suggest the following challenge. The percent of patients that shed virus, the mean peak titer, and the mean duration of shedding are all higher for young children than in adults, and that's the data that's shown here.

In addition to that, the human infective dose 50 is lower for young children compared to adults.

The data that we have regarding transmission and virus shedding, the characterization of shed viruses come primarily from the Finnish day care trial. This is a study that was conducted by our partners in FluMist, Wyeth, and was a double blind, randomized, placebo controlled trial that was conducted in 197 children, age eight months to 36 months.

These children, there were 98 FluMist patients and 99 placebo patients. Out of these 99 placebo patients there were 93 of those, that is, there were six children who were not in a play group with a vaccine or with a vaccinee. So those children were obviously not available for FluMist transmission, and the analyses are conducted with 93 patients in the placebo group.

This trial was conducted in 51 play groups in two cities in Finland. Forty-five had both vaccine and placebo children in the play groups were physically in separate buildings in two cities in Finland. So they were geographically distributed separately from each other.

The other two that were in the same building were physically separated play groups, and there was little chance of commingling of the participants of the staff.

There was an average of 4.1 study children per play group, and the children attended day care for at least three days a week for more than four hours a day.

Each placebo trial was exposed to an average of 1.9 vaccinees.

Now, I know that several of you have been involved in doing day care studies in the past and recognize the difficulty and complexity of doing these kinds of studies. Obviously these children and families had to agree not only to either be part of the trial, but had to agree to be in the day care while the trial was being done.

This trial is actually one of the largest and most comprehensive studies of its type. It is actually in size secondary to only Gelfand's study of oral polio transmission.

The trial involved nasal culturing the first two days after dosing and at least three times a week for three weeks. So the trial itself did about 2,000 cultures, which represented approximately ten cultures per child.

There in addition to that was extensive phenotyping and gene typing of the isolates that were identified.

It's also important to recognize that this setting was designed to maximize the chance of vaccine virus transmission. One wanted to be able to detect transmission so that we identify and calculate rates, and this is a setting in which we maximize the transmission using in this setting young children who are seronegative generally and children who had extensive exposure to each other.

The statistical methodology that was used in the final analyses of the Finnish day care study was the estimation of the probability of transmission using the Reed-Frost model, and it's important to note that this model takes into account the number of vaccine-placebo interactions.

So when you're dealing with transmission rates, it's important to approximate the number of people in each of those groups, not simply the attack rate in the population.

What we found were that there were 80 percent of vaccine recipients that shed virus. Thirty-two percent shed H1N1; 12, H3N2; and 74, Type E virus.

The mean duration of shedding in days was 7.6 days. There was one placebo child who shed Type B vaccine virus on the day 15 visit, and using the Reed-Frost model for that documented case of vaccine virus transmission, the probability of transmission was .0058 with the upper bound of .017.

Now, in addition to that, there was wild type A H3N2 circulating the community in Finland during the time the trial was conducted, and there were six additional placebo children who shed Type A influenza virus during the study. Two of them shed wild type A strains and did not shed vaccine virus. So those are clearly not vaccine virus transmission.

There were four additional patients, placebo patients, who shed a Type A virus that was isolated and identified in Finland, but could not be reisolated and identified and thus could not be confirmed to be either wild type or vaccine virus.

Let me walk you through those four cases. The first one we consider to be a possible case of vaccine virus transmission. That child shed on one day in a play group where one vaccinee shed vaccine virus seven and ten days earlier.

So although it couldn't be identified as vaccine or wild type virus, it's possible that that was a vaccine virus transmission.

There are two cases which we would consider to be highly unlikely to be vaccine virus. The first one was a patient, a child who shed on two occasion. The first was before any vaccination occurred in the play group. That's most likely wild type virus, but we couldn't totally exclude that it was vaccine virus.

And the second time they shed was a few days latter, which was one day after the vaccine was introduced into that play group. So clearly, again, this is unlikely to be vaccine virus.

The other unlikely case was a child who shed on one occasion. There was no other participant in that play group who shed Type A virus, but there was one who shed B virus five days earlier. So other A shedders in the population, and the isolate from that patient was a Type A virus.

And we consider the fourth patient really not possible because that child shed two days before vaccine was introduced at all into that play group.

If one calculates the probability of transmission based on the Reed-Frost calculations, you see that using the confirmed case I've showed you that transmission probability. If you add the one possible case, the probability of transmission is .01 and the possible plus the two unlikely cases is .02.

And we also did extensive virus characterization of the shed viruses. The temperature sensitive and cold adapted phenotype was confirmed in all of the isolates that were tested. There were 124 isolates in this trial that were tested to confirm this.

In addition to that, we took the last isolate from each of the children that we had isolates at that time available for those, and we chose the last isolate so that it was farthest away from the introduction of the vaccine, and we completely sequenced 55 of the shed isolates.

There were no reversions of the master donor virus attenuating mutations. As expected, there were a minor number of mutations that were observed in influenza. That's not unexpected.

And if we took viruses t hat contained the mutations that were observed in more than one of the isolates and put them back into ferrets, they all retained the attenuated phenotype.

When we look at the transmitted virus, that retained the 6:2 genotype. It was identical in genetic sequence to the virus that was shed by the vaccine recipient in the play group. It retained the cold adapted and temperature sensitive and attenuated phenotype. It was not associated with increased reactogenicity, and was not observed in any of the other placebo members in the play group.

So I think we can conclude from the Finnish day care study that in this day care setting the probability of transmission was estimated to be .006 or .02. There was no phenotypic or genotypic reversion observed in the shed or transmitted viruses.

So, in conclusion, we believe that we have demonstrated the efficacy of FluMist in the prevention of culture confirmed influenza in children. Efficacy was consistent across all the age groups that were studied, and the efficacy in children greater than or equal to 60 months was similar to the group as a whole.

Effectiveness of FluMist was demonstrated in trials in adults. The effectiveness was consistent also across age subgroups, and we found no evidence of the 50 to 64 year olds being significantly different from the population as a whole.

Efficacy was comparable on annual revaccination.

We believe that FluMist is safe in children greater than or equal to 60 months through 17 years of age and healthy adults in 18 years through 64 years of age.

There was a mild increase in self-limited URI symptoms, but no increase in fever greater than 101 or composition of symptoms that would constitute influenza-like illness.

We saw in adults the safety profile consistent across age groups, including in the 50 to 64 year olds.

We believe that additional information is needed to assess the risk-benefit in children under 60 months. We identified a possible signal in that age group and consequently need to collect additional information in that pediatric age group.

We believe that the safety profile was similar and the events were lower on annual revaccination, but the risk of vaccine virus transmission is low. The probability is estimated in the day care center, and those rates are likely obviously to be lower in older children and in adults.

And we've demonstrated genetic and phenotypic reversion has not been observed in the studies that we've done so far.

So we present that data to you as well as the data from the previous VRBPAC presentation as a portfolio of information to support the proposed indication, which is for active immunization and the prevention of disease caused by Influenza A and B in healthy individuals age five years through 64 years of age.

We believe the FluMist represents a potentially important addition to the portfolio of the public health armamentarium, as well as the medical armamentarium for both increasing the rates of influenza immunization, as well as for prevention of flu in the population in the United States.

Thank you.

CHAIRMAN DAUM: Thank you, Dr. Connor.

What I'd like to do now is to get input from the committee with regard to clarification of Dr. Connor or Dr. Young's presentation, things that went by too quickly perhaps or issues that you'd like to know whether there are other data about, and then we'll take a break and hear from Dr. Mink and our FDA colleagues.

So Dr. Snider first and then Dr. Edwards.

DR. SNIDER: Thank you.

I actually had four questions. Hopefully relatively short and sweet in terms of an answer.

With regard to the statement about efficacy in children greater than 60 months, in looking at the materials, I actually see up to 83 months, but I didn't see data from 83 months through 17 years of age, and I was wondering about if there are efficacy data that weren't shown on that point.

DR. CONNOR: No, the original pivotal studies were done in children who were 12 to 15 and 17 months of age, and as I said, the oldest children that were in those trials had aged up to 83 months by the time they were in the second year, but there's not additional efficacy-effectiveness data within the population.

DR. SNIDER: So we have safety data, but no efficacy data?

DR. CONNOR: That's correct. The efficacy data is in the younger children, and what we presented were data to show that the efficacy in all of the age groups that we demonstrated were consistent.

There is a fair amount of evidence, however, obviously with the cold adapted predecessor vaccines in that age group demonstrating that efficacy exists.

DR. SNIDER: Right, right. Thank you.

The second question very quickly. In adults you mentioned that they were healthy adults, but I don't recall your mentioning whether they were smokers or not and whether there was a difference in efficacy or safety as it related to smoking status, if that's known.

DR. CONNOR: Yeah, I don't. They were healthy adults from the perspective that they were healthy working adults working at least 30 hours a week. To my knowledge, there's no specific information about whether we can separate the population by whether or not they were smokers or not, and I don't think that information was collected as part of the trials.

DR. SNIDER: Okay, and then finally, I'll run these two questions together because it has to do with adverse events, runny nose, for example, asthma, and wheezing. Is there any association between those types of respiratory events and shedding, the amount of shedding, more or less in those people or, you know, the type of strain they might be shedding? Any information that would give us a clue around the etiology?

Obviously this is not necessarily relevant to the questions FDA is posing, but interesting in terms of what the etiologies might be for these adverse events.

DR. CONNOR: Obviously because the adverse event was a post hoc sort of identified issue and studies were driven by primarily influenza diagnoses, there's not specific information that correlates shedding with those specific outcomes.

CHAIRMAN DAUM: Thank you, Dixie.

Dr. Edwards, Katz, Eickhoff, Myers, and Markovitz.

DR. EDWARDS: I have two questions for Ed. The first is that for those of us who are pediatricians, we know how extraordinarily difficult it can be to decide whether a patient has bronchiolitis or asthma in those first few years of life, and which I would think probably complicated extensively your analysis of the wheezing episodes.

Could you tell us a little bit how you have dealt with bronchiolitis, whether the diagnosis of asthma was consistent in the Kaiser population, whether there are guidelines upon which that diagnosis is made in that population?

DR. CONNOR: Yeah, the best we could do, Kathy, is that the analyses that I included as asthma and wheezing actually include other synonyms for wheezing. So reactive airways disease, shortness of breath, bronchiolitis, bronchitis are included in that analysis.

As you mentioned, it's complicated if you just look at these are just simply illness visits. So what you're recording is what's in the record at the time that the visit occurred, and various people are seeing the child. They could have recorded that it was rule out asthma. They could have recorded a variety of different things.

So what we've tried to show you is the original signal and then the most inclusive diagnosis that we could get out of the Kaiser database.

DR. EDWARDS: Did you extract out the episodes during times when RSV was circulating to eliminate that or was that analysis done?

DR. CONNOR: Yeah. Well, not particularly because we just basically looked at the two randomized groups. So we were assuming that the same things were happening in both of the groups and looked at the comparison between the two treatment populations.

DR. EDWARDS: You may want to subsequently look at that.

DR. CONNOR: Yeah, yeah.

DR. EDWARDS: The other question that I had relates to the challenge study, and obviously it was beautifully designed, and as I mentioned to Dr. Katz, could only have been done in Europe, but I think that the fact that it was conducted during the time of co-circulation with wild type virus may, indeed, have decreased your transmission with the vaccine strain because of interferon generation or interference of a perhaps more potent fibrous than the attenuated vaccine.

So do you have any comments or ideas about that?

DR. CONNOR: Yeah, I think the actual number of cases that were wild type flu that were identified during the transmission study were pretty low. So the expectations that while it certainly was circulating and could interfere with the interpretation of at least those cases that we were talking about, there wasn't a whole lot circulating in those day care.

CHAIRMAN DAUM: Thank you.

We move on to Dr. Katz, please.

DR. KATZ: As I should have anticipated, Dr. Edwards has asked most of the questions I was going to, but I'd add one comment. It may seem gratuitous, but the whole question of transmission, the other vaccine that we used for many years where transmission was an issue was oral polios, and it was considered advantageous that there was transmission from immunized children to those who didn't have the benefit of the immunization, and yet this was a preparation where reversion to neurovirulence was much more likely to occur than as you've been able to show with your nasal flu vaccine.

And my question was: in any of the transmitted children did anyone look at antibodies to see if they developed an immune response as the result of transmission?

DR. CONNOR: That wasn't looked at, Sam, in the one kid who we know transmitted the virus.

CHAIRMAN DAUM: I have Drs. Eickhoff, Myers, Markovitz and Diaz, and then I think we'll take a break and hear from the FDA, and there will be ample time this afternoon to return to many of these issues as per the committee's pleasure.

Dr. Eickhoff, please.

DR. EICKHOFF: Thank you.

The concurrent vaccines trial, the MMR and varicella vaccine trial, what age groups are those children? I understand the trial is now fully enrolled.

DR. CONNOR: We're now halfway enrolled. The age group is between 12 and 15 months. So they're the younger kids.

DR. EICKHOFF: That leads me to the second question, which is if you could share with us to the extent that you're able your long term plans about that 12 to 49 month age group.

DR. CONNOR: I think obviously we're very interested in understanding the issues that pertain to that population. As I mentioned, we have fully enrolled the trial that was the MMRV trial.

In addition to that, we have trials designed and in discussions with CBER to begin to look at the other vaccine components so that for all of the other childhood vaccines -- and obviously we'll have to go back and look at the issue of the signal of the asthma-wheezing issue in that population.

We have ongoing trial a study with Steve Black and Kaiser trying to sort out are there better tools to be able to distinguish history positive and history negative children. This trial obviously just used parental history of asthma or the parental report of a history of asthma in the child as the discriminator.

We're also going to be looking at other studies in that group of children.

DR. EICKHOFF: Kindly could you again share with us your thoughts about the other end of the age spectrum? I remember, oh, ten years ago, I think, or more, I think, John Traynor's study from Rochester and the apparent perhaps not synergistic, but both vaccines were better than either one alone.

This may not ever fly as public health policy with regard to influenza vaccine, but if you could, please share with us your thoughts about the over 65 group.

DR. CONNOR: Yeah. I think, you know, obviously we focused our attention on the healthy individuals that are within the age group that could, if the vaccine was provided, could enhance the vaccine uptake in the healthy population, and once we've studied that population, then in addition we'd like to be able to study the vaccine in other higher risk populations, including the elderly population, as well as other patients.

So there are clearly plans and desires to look at that population. Obviously out focus has been on the healthy population principally.

DR. EICKHOFF: Thank you.

CHAIRMAN DAUM: Dr. Myers, please.

DR. MYERS: In follow-up on that and Dixie's question, this is a vaccine that at least on the current indication is not indicated for any of the high morbidity risk groups, and in a way, that makes the safety bar go higher, it seems to me.

And so I have two issues. The first one is that although it's intended for healthy individuals, it's likely to be administered to those who have unrecognized immune deficiency, unrecognized pregnancy in smokers. It's likely to transmit to the elderly in family settings and to younger siblings.

And so I guess I have some questions about safety in those groups and would ask if you could address some of those.

And then a second thing is I'm struck by the fact that in all of your studies about 20 percent of recipients of normal allantoic fluid intranasally require anti-pyretics. Forty percent of adults have headaches, basically have febrile URI after getting normal allantoic fluid.

And I remember at the last meeting raising the question has normal allantoic fluid been compared to a placebo for reactogenicity, and again, is this a vaccine intended for the normal host, normal healthy adult?

I think we need to ask why there are so many febrile URIs in the placebo group.

DR. CONNOR: I think to take the second question first, normal allantoic fluid has been used as the control group for most of these studies and other CAIV studies mostly obviously for the placebo, for the masking effect and other effects.

We have looked in several settings, and I can either show slides or show slides later of the distribution of the placebo activity. If you look at the distribution of the URI symptoms, for example, after a dose of placebo, you see no temporal distribution associated with giving the normal allantoic fluid.

The other fact is that there are some studies that Wyeth has conducted using saline placebo in the same setting, and what you actually see is the healthy patient effect. What you see is that you're screening out patients before they get vaccinated, and then both in the FluMist and in the placebo groups or in the other cases in the FluMist and the saline group, you see regression basically back to what is the normal rate of respiratory symptoms in adults in that population.

So, for example, you see return back to about 20 percent of children having URI respiratory symptoms. So the evidence actually suggests that what's happening is that the base population rate is what you're actually seeing with normal allantoic fluid, and the reason why there may be some trend is that the trend is basically screening out the healthiest patients right before you're doing the vaccination.

And I can either show some of that information or not, but the other issue is that we obviously have focused, as I mentioned, our attention primarily on the healthy individuals. There have been some studies that have been done on some immunocompromised host populations, a relatively small study that was done in HIV infected adults in which there was nothing of interest in terms of significant disease in that population, and that's the primary extent of the population that we studied thus far.

CHAIRMAN DAUM: Okay. We have Dr. Markovitz and Diaz, and again, we're looking for input on clarification of what the presentation consisted of at this point, with time to return to many of these important issues a little later.

MR. MARKOVITZ: Yeah, I have two questions. One, it seems as though -- and this sort of follows up on some other questions asked -- it appears that COPD patients, for example, would be an obvious eventual target here, and I understand you're not asking for an indication here, but have they been studied and you've seen similar things that you've did as seen with kids vis-a-vis asthma or is this something that has just not been studied at all thus far?

DR. CONNOR: Yeah, the only studies that have been conducted at studying COPD and the VA studies that are not part of this application in which COPD patients -- I think part of that data was mentioned the last time at the last VRBPAC -- in which there are some analyses of that trial, including some adverse event outcome. It wasn't anything specific that was seen in terms of serious adverse events or other outcomes in the COPD patients that were different, although obviously the rates of illness in events were higher than in the healthy population.

MR. MARKOVITZ: So there's some data, but not conclusive at this point.

DR. CONNOR: Right.

MR. MARKOVITZ: The second question is it appears that a lot of the studies are done in a rather homogeneous population. Do you have any information in terms of like inner city populations or anything like that?

DR. CONNOR: Yeah, i think we can show you data later about the geographic distribution and the demographic data distribution of the populations. We see no difference based on background, ethnicity, racial distributions, age distributions, et cetera, and we can show you some of that data later if you are interested.

CHAIRMAN DAUM: Dr. Diaz.

DR. DIAZ: I had just a few quick questions. I, too, was struck and concerned about the difficulties of diagnosing and/or labeling younger children with asthma as a diagnosis and was wondering in the studies on asthma and wheezing if additionally you looked at the use of any pharmaceuticals in the population, i.e., bronchodilators or other indicators of concern in wanting to treat true wheezing episodes.

DR. CONNOR: Yeah, I think the data that we have from the first study which looks obviously at a generally health population, but populations that still got into the study that had something in their medical records that were associated with asthma and wheezing; we did look at those individual events for what happened to them.

And while one of the kids were hospitalized, as most kids who have asthma and wheezing events, they were treated with various kinds of medications, including brochodilators, and about 20 percent of them some steroid use.

That was generally distributed primarily in patients that had true diagnosis of asthma. In the general population we didn't specifically go back and look independent of anything else whether there was --

DR. DIAZ: That's what I was wondering.

DR. CONNOR: -- yeah, whether there was asthma diagnoses.

My sense is from looking at that data, however, is that there's pretty good correlation between anything in the record that says anything about wheezing, which is basically what we're culling out, and the use of bronchodilators in general.

DR. DIAZ: Additionally, I just wanted to clarify in your efficacy studies. I seem to recall looking at efficacy out to 42 days. Am I incorrect in that or was that the endpoint?

DR. CONNOR: Yeah, the follow-up period was generally through 42 days, and it was all culture confirmed influenza in the outcome.

DR. DIAZ: So you don't have --

DR. CONNOR: I'm sorry. You're referring to the efficacy outcomes rather than the safety outcomes.

DR. DIAZ: Correct.

DR. CONNOR: Which are through the whole season, yeah.

DR. DIAZ: Which is through the whole season?

DR. CONNOR: Efficacy.

DR. DIAZ: Efficacy. That's what I wanted to clarify.

DR. CONNOR: Yes.

DR. DIAZ: Because that's a very important point.

DR. CONNOR: Yes.

DR. DIAZ: Additionally, the issue about transmission, I think, is an important issue. Are there any markers other than the attenuation markers as such that you can follow resorted strains if, in fact, you did revert to that worst case scenario that somebody described of reverting or resorting to the wild type strain? Would there be any genetic markers that could still predict that that particular strain had resorted from the attenuated strain that you know of?

DR. YOUNG: Yeah, basically what you have to do is there are ways tat we can either sequence or do RFLP analysis of the strains to follow that.

In fact, within that Finnish day care study looking at the viruses shed from the vaccinated individuals, we could actually, because there were a single point mutation differences, you know, that are silent mutations between the cold adapted strains so that we could actually show that there were four reassortant events that occurred between the H1N1 cold adapted virus and the H3N2 adapted virus. So you can actually follow whether the genes do get switched, but again, we've never seen any change in phenotype when that happens.

DR. DIAZ: Right, and lastly I just wanted to ask, and I presume you don't have any studies that look at transmission in a household setting?

DR. YOUNG: We have not done that with CAIV that I'm aware of. I guess there were some -- Paul, there were historical studies done with CAIV?

Paul Mendelman.

DR. MENDELMAN: In the briefing document that we provided, the ten publications for transmission studies were done historically with predecessors of FluMist, and those included various settings, including households and college roommates as well as children and adults and married couples.

CHAIRMAN DAUM: Okay. Thank you very much, committee members. Thank you sponsors for your presentation.

It is 10:45.

Dr. Mendelman wishes to inform that in those studies no transmission occurred.

(Laughter.)

CHAIRMAN DAUM: It's 10:45 here in the Eastern time zone, an we will take a ten minute break and reconvene at 10:55.

(Whereupon, the foregoing matter went off the record at 10:48 a.m. and went back on the record at 11:00 a.m.)

CHAIRMAN DAUM: Will everybody please take their seats so that we can begin?

I'd like to continue our meeting by calling on Dr. ChrisAnna Mink of the FDA to give us the FDA presentation.

If everybody could take their seats and cease conversations as quickly as possible so we can get moving on the committee's agenda.

Dr. Mink.

DR. MINK: I'd like to now begin the FDA clinical summary for the FluMist application.

You've heard much of this this morning. So will try to abbreviate where I can.

A total of 20 studies have been submitted to the biologics license application. Fourteen were randomized, double blind, and placebo controlled, and three of these were considered to be pivotal safety trials. They were randomized two to one, FluMist to placebo. There include 06, the efficacy trial; 09, the adult trial; and 019, the large safety trial for Northern California Kaiser.

There are also six non-placebo controlled trials.

Final reports for three of the 20 trials were submitted with the sponsors' responses in January 7, 2002, and have not previously been presented in full to the committee. These include 019, the Kaiser trial; 012, Texas community trial performed largely in a health maintenance organization; and the Wyeth-Lederle sponsored transmissibility study in day care.

A total of 20,228 first doses of FluMist have been given, and in the age range requested of five years through 64 years, 14,154 first doses have been given.

I would now like to review the trials recently submitted. We'll start with the 019 Kaiser trial.

This enrolled healthy children from one to 17 years of age, randomized two to one, FluMist to placebo; two doses, 28 to 42 days later for the one to eight year olds groups and one days for nine to 17 years.

It started in October of 2000, and they used the 1999-2000 vaccine strain for this safety profile. Both of the A strains differed from 1999 on 2000 season, though there was some similarity, and the B strain was the same.

The exclusion criteria listed was the parental reported history of asthma or possible asthma.

In this trial, there was no active monitoring for solicited reactogenicity events plus vaccination, events such as cough, fever, or runny nose.

The Northern California Kaiser database was searched after each dose for their primary safety outcomes listed here. Serious adverse events were defined consistent with 21 CFR, including death, life threatening, hospitalizations, significant disability or congenital anomaly or birth defect.

Medically attended events were defined as an encounter with the health care provider. There were 170 individual events searched and four prespecified groups categories.

You've heard this earlier. The four prespecified categories were acute respiratory events, which included approximately 25 diagnostic codes; systemic bacterial infections; acute gastrointestinal events; rare events, historically associated or you'll see coined potentially related with wild type influenza.

The utilization settings include hospital, out-patient clinic, emergency department, and all settings combined.

Prespecified age groups were 12 to 17 months, 18 to 35 months, one to eight, nine to 17, and then all age groups one to 17 years.

Statistically the plan was to estimate incidence of MAEs and SAEs in the FluMist recipients relative to the placebo subjects in the post vaccination period. This is expressed as relative risk, RR, in 90 percent confidence intervals constructed using the binomial method.

Multiple encounters for the same MAE were counted only one. Ninety percent confidence intervals were used for these safety assessments.

Interim analysis was performed with unblinding to provide safety data at VRBPAC 2001. Over 1,500 analyses were performed, and there were no adjustments for multiple comparisons.

A total of 9,689 evaluable subjects were available. This included a 6,473 FluMist and 3,216 placebo subjects. Approximately 8.5 percent of participants had a history of asthma upon review of the database and medical records. Approximately 75 percent of those subjects were from one to eight years of age.

As you can see, the age distributions for those one to eight years were about 58 percent of enrollees and about 42 percent were nine to 17 years. There were similar demographic characteristics for age, gender, and ethnicity in the two treatment groups.

Eighty seven percent of the one to eight year olds received the dose two of their vaccine regimen. The reasons provided for not receiving dose two included unable to contact, which was about 55 percent of the FluMist subjects, or noncompliance, which was about 35 percent in FluMist subjects.

Adverse events following dose one were reported as the reason for not returning for dose two for 39 FluMist subject and 27 placebo subject. In the line listing of these adverse events, most appear related to the respiratory tract.

The results for the SAEs are shown on this slide. No deaths were reported. A total of 20 SAEs were reported with the rate of .02 percent for both groups, FluMist 13, placebo seven subjects.

Of the 13 SAEs in the FluMist subjects, 12 were reported after dose one. These events included 11 hospitalizations, six psychiatric hospitalizations, and three others, such as an ED visit, one clinic visit, and an out-patient surgery. There were no patterns for the diagnosis seen for SAEs.

The next slide shows the number of subjects with medically attendant events by setting. As you can see, clinic visits were the most common, occurring in about 36 percent of both groups.

Hospitalizations were infrequent, occurring in less than .6 percent of both the FluMist and placebo subjects.

For the prespecified group diagnoses, none of the groups had occurred at significantly increased rates in the FluMist recipients. There were no reports of systemic bacterial infections and no relative risk could be calculated.

There was a decrease in the relative risk, significant for acute respiratory events in the FluMist recipients for all ages, settings, and doses combined with the upper bound of the relative risk confidence interval being less than one.

The results for asthma events, overall the asthma events occurred in approximately .4 to 1.3 percent of FluMist recipients, depending upon the age groups evaluated. In the interim analysis presented at VRBPAC last year, there was an increased relative risk for asthma events observed in FluMist subjects ages 18 to 35 months, one of the prespecified age groups.

In the final analysis, this increased statistically significant increased relative risk was still observed in the 18 to 35 month age group.

As shown on the next slide, for all settings following dose one in 18 to 35 month old children there were ten events, ten subjects in the FluMist group and zero in the placebo with a lower bound of 1.95. For all settings and all doses combined, the relative risk was 4.06 with these confidence intervals.

The NA -- I hope you can read the footnote -- means not available, not able to calculate due to zero events occurring in the placebo subjects.

This slide describes the asthma events in the 18 to 35 month old children. There were a total of 18 subjects, 16 FluMist and two placebo, that had 20 asthma events. Seventeen of these subjects were seen in the clinic, and one subject was seen in the emergency department, and this was following dose one.

All subjects required treatment with 94 percent of them receiving beta agonists.

To help further evaluate the age related risks for asthma a post hoc analysis was performed by the sponsor looking at subjects in increasing six month increments, starting at 12 months of age through 107 months of age. Selected age groups for this analysis is shown on this slide. Here's the 12 to 35 month old group, but we stop at the 77 month.

As you can see, there's an increase relative risk of 3.53 identified for the 12 to 59 month age group, with the lower bound of 1.1 in these upper bound confidence interval.

The statistically significant risk was not identified for age groups over 12 to 59 months of age. This slide shows this analysis for the 12 to 59 months after dose one with the increased relative risk. This is a CBER generated table. So the confidence intervals are a little different because a different calculation method was used.

After dose two the relative risk was not significantly increased and subjects 60 to 107 months of age after dose one or dose two, there were no increased risks identified, nor was there an increased risk in nine to 17 years of age.

To summarize, for asthma events the increased risk was 3.53 for the 12 to 59 months after dose one, not seen after dose two, or in the 60 to 107 month or children nine to 17 years of age.

Upper respiratory infections were also evaluated. There was no increase in the relative risk for URI in combined analysis of all ages, settings, and doses. There was one SAE and a placebo recipient for hospitalizations for URI and croup observed on day four post vaccination.

There was an increase in the relative risk for URI in three of the 41 separate analyses as shown on the next slide. The sponsor also showed these data to you earlier.

You can see there was an increase in the entire cohort in the emergency department after dose one with the lower bound of 2.14. No events occurred in the placebo group.

The events in the one to 17 years were primarily accounted for by subjects in the one to eight, with nine events here and 11 here. Again, the lower bound was 1.7 with no events in the placebo subject.

There was also an increase identified in 18 to 35 month old subjects for all doses and all settings combined with a relative risk of 1.3.

Pneumonia events were also evaluated, and in that search there was no increase in the relative risk observed for pneumonia, bronchitis or bronchiolitis in any age group, setting, or dose observed. I will discuss this in a bit more detail later.

Additionally, a search for abdominal pain was performed. This occurred in about .7 percent of FluMist and .8 percent of placebo subjects. There was an increased relative risk in two analyses, in the nine to 17 year olds in the emergency department and one to 17 years in the ED, and there was a decreased relative risk in two analyses, one to eight year olds in the clinic and one to eight year olds in all settings following dose one.

There were no specific abdominal disorders identified and no cases of intussusception, mesenteric adenitis, or intestinal obstruction were reported.

A search for rare events potentially related to influenza show that there were no cases of encephalitis, encephalopathy, Guillain-Barre, Reye's Syndrome, or other influenza associated disorders.

There were ten subjects, seven FluMist and three placebo, who reported 11 seizure events. Five of the seven FluMist subjects and one of the placebo subjects was less than five years of age. As you can see, the relative risk was not increased for the events of seizures.

So in conclusion, for this trial AV019 was the major contributor to the safety database, especially in the pediatric age group. SAEs occurred at a rate of .2 percent, and there were many children with asthma enrolled, despite exclusion criteria.

There was an increased relative risk for asthma in the 18 to 35 month olds after dose one in the prespecified age groups and in 12 to 59 month olds in a post hoc analysis. The children with asthma events did require medical therapy.

No significant increase in relative risk for asthma was observed for children over 60 months of age. There was no increase in the relative risk for pneumonia identified, nor were there rare events possibly related to influenza reported.

An increased relative risk for URI events and musculoskeletal pain, which was presented earlier by the sponsor, was observed in children 18 to 35 months of age, suggesting that possibly these children upper respiratory infections or mild influenza-like illnesses.

There were 1,500 analyses performed in this trial, and an increase for some of the relative risk for some of the MAEs may be due to chance alone. Additional studies are needed to assess the association of asthma events following the receipt of FluMist.

The next trial that will be discussed is Study AV012, the Texas community trial, and I will discuss years one and years two.

The design was open label, nonrandomized, and the children were 18 months to 18 years of age who were eligible for participation, and children with a history of mild asthma were permitted in this trial.

The vaccine, each year that the kids were enrolled, they received a single dose of FluMist. The study was planned to assess herd immunity following FluMist and safety monitoring was a secondary objective.

There were several limitations for the design of this trial, and effectiveness data will not be discussed. We are presenting only the safety data from years one and years two. These data contribute the safety experience following 9,549 doses in 7,448 children.

For the revised plan for safety assessments, which was implemented after the trial was initiated, SAEs for 42 days post vaccination became the primary measure of safety. These were captured by postcard reporting of database searches.

Medically attended acute respiratory illness, MAARI, became the secondary measure of safety. These were captured by database searches for overall MAARI events and for selected respiratory events, including asthma.

Data analysis of MAARI events used the post marketing method described by Griffin, et al. Each participant served as his own control. Event rates within a specified vaccination period were compared to event rates within a reference period. There were two vaccination periods: day zero to 14 and day zero to 42.

A single reference period was constructed by combining pre-vaccination period and a post vaccination period. The results show that in year one the study period was from August 17th, 1998, through January 30th, 1999, and 4,298 subjects were enrolled. Approximately 13 percent of these had a positive history for asthma.

In year two, the study period was September 13th, 1999, through February 10th, 2000. Fifty-two hundred and fifty-one subjects were enrolled, and the asthma history was positive for about 17 percent of subjects.

A total of 2,101, or about 40 percent of the subjects, were repeat vaccinees.

SAE results. There were no deaths reported. In year one there were eight SAEs and the 4,131 evaluable subjects were a rate of .2 percent. Six of the eight SAEs occurred beyond day 21. The two that occurred within 21 days included hospitalization for depression in a 16 year old and aseptic meningitis in a seven year old.

In year two there were 16 SAEs, all hospitalizations, reported in 15 of the 5,033 evaluable subjects for a rate of .3 percent. Nine of the 16 occurred on or after day 21. Two SAEs were related to the respiratory tract, and this included on day 11 in a 23 month old, who was admitted for RSV documented pneumonitis and febrile seizure, and on day 44 post vaccination an event of pneumonia in an almost four year old.

There were generally multiple different diagnoses included in the SAEs for year one and year two, and no pattern was observed.

MAARI results, in brief. There was no increased relative risk for overall MAARI for either vaccination period observed in year one. In year two, there was an increase in MAARI events for both vaccination periods, day zero to 14 and day zero to 42, as shown on this slide.

Asthma events overall were reported in approximately one percent of all participants in each year one and year two. In year one there was no increase in the relative risk for asthma events for either vaccination period. In year two, no increase was observed from the day zero to 14 period, but there was an increase in asthma events for day zero to 42 with the relative risk of 1.83 and those confidence intervals.

There are several limitations to this trial, including that it was not randomized or blinded. No control group was available for statistical comparisons, and the post marketing method used for the data analysis may be problematic for evaluation of pre-licensure safety.

For example, the method has most often been used for acute events, not chronic events, such as asthma, and has not been previously used for events associated with seasonality, such as influenza.

The MAARI evaluations for safety were not prospectively defined, and as you saw in the data, there were differences in MAARI rates observed between years one and years two.

Conclusions from this trial were that 7,448 subjects received 9,549 doses of FluMist, which is approximately 37 percent of the total first doses observed in our safety data base.

The rates of SAE were .2 to .3, which was very similar to the rates observed in AV019, the randomized, double blind trial.

There were no reported events potentially related to wild type influenza. There was an increased relative risk from MAARI observed in some analyses, but not in others.

Briefly to compare AV012 and AV019, the rates for asthma events were similar in the two trials, between .5 and 1.5 percent for FluMist recipients, and this occurred despite the marked differences in the design of the trials, most notably one being open label versus randomized double blind.

To further evaluate the asthma and wheezing events, the sponsor performed a review across all 20 trials in the BLA. As they discussed earlier, there was significant overlap in the use of the diagnostic codes for asthma, reactive airway disease, RAD, wheezing, and shortness of breath, SOB.

So across these 20 studies, approximately 82 to 80 percent of asthma RAD, wheezing, and SOB events, and 74 percent of asthma RAD occurred in children from one to nine years of age. This age group accounts for approximately 50 percent of the total study population.

Asthma events occurred throughout the 42 day post vaccination period with no clear evidence of temporal clustering.

A search of SAEs related to asthma, there were four hospitalizations identified for the grouped category. In AV006, year two, a 23 month old with a history of asthma was admitted for status asthmaticus on day eight after dose three of FluMist. In 008, an adult, a 69 year old with heart disease, wheezing, and congestive heart failure, was admitted on day 16 post FluMist. And in 02, a child had two hospitalizations for RAD on day four and day 33 in a placebo subject also with a known history of asthma.

This slide shows the occurrence of events for the protocol defined age groups. In 06, that was 15 to 71 months after dose one. There was no increase in relative risk. In 019, one to eight after dose one, no increase, statistically significant. In 019, in nine to 17 years, there was no increased risk, nor was there an increased risk seen in adults.

These are for, as I noted, the predefined age groups in the protocols. However, in the analysis by age subgroups, as discussed above in AV019, there was an increase observed for 12 to 59 months of age and for 18 to 35 months of age. AV019 was the major contributor of children to the database, and thus the review of the 20 studies is very similar to these results.

Study 010 was also performed, which enrolled nine to 17 year old subjects with moderate to severe wheezing. Three of 24 FluMist recipients compared to zero of 24 placebo recipients had exacerbations in the 32 days post vaccination in that trial.

Thus, the concern is ongoing for the risk of asthma and wheezing in young children and possibly for subjects with a known history of asthma.

The third trial that we will discuss is the Wyeth-Lederle vaccine sponsored trial D145, which is assessment of the transmissibility of CAIV-T, FluMist, in day care attendees in Finland. This trial was randomized one to one, double blind to assess shedding and transmission of cold adapted influenza vaccine on strains in day care attendees who are eight to 36 months of age. They were eligible to participate in the study if they attended day care at least three times a week for at least four hours per day. There had to be at least four contacts in the play group with at least one vaccinee within that play group.

The subjects were given one does of cold adapted vaccine or placebo. Then nasal swabs were obtained in day zero, day one, and three alternating days, such as Monday, Wednesday, and Friday, per week through day 21.

Isolated vaccine viruses were typed, A or B, subtyped, H1N1 or H3N2, phenotyped for the cold adapted and temperature sensitive, and a subset was genotyped.

The original primary objective was to assess the percent of placebo subjects who were shedding virus identified as vaccine strains.

And the sponsor's post hoc analysis, using the Reed-Frost model, was to assess the probability that a vaccinee would infect a placebo subject.

A total of 197 subjects, 98 FluMist and 99 placebo, were enrolled from 51 day cares as the sponsor described for us earlier. A typical play group included two CAIV and two placebo recipients, though there were a lot of variations.

The all available shedding population, i.e., the assessment of shedding, could be performed from all 197 subjects. I present here for you the FluMist results. Eighty percent of the FluMist recipients shed at least one vaccine strain. Forty-three shed Type A; 72 shed Type B; six shed Types A and B; six shed all three types; and there were 13 subjects who shed viruses that were not able to be subtyped or genotyped. This included six Type A and seven Type B.

The duration for shedding is shown on this slide. For the H1N1 there were 33 isolates. All were vaccine strain. There was no H1N1 circulating at that time is my understanding in Finland.

The median day for shedding for H1N1 was three, but the range was one to 21, and as you'll recall, 21 was the last day cultures were obtained.

For H3N2 there were 12 isolates. Median shedding was eight days with a range of three to 17, and for Type B, there were 72 isolates, 65 confirmed as vaccine strain, and the range of shedding was one to 15 days. There were no wild type isolates recovered from FluMist recipients during this trial.

For the transmission assessment, the placebo subjects included 93 available. This excludes six subjects who had no contact with the FluMist vaccinee. The all evaluable population defined by no protocol violations included 57 subjects. No protocol violations mean that they were not outside of the eight to 36 months of age; that they had nasopharyngeal cultures obtained less than four days. They had at least contact with one cold adapted recipient, and that there were at least four subject in their play group.

From these placebo subjects, viruses were recovered from seven subjects. One subject shed Type B Ann Arbor vaccine strain identified on day 15. Six placebo subjects shed a total of nine Type A isolates. Two subjects shed wild type on two different days, a total of four isolates, and four subjects shed a total of five isolates that could not be identified.

That's really hard to say.

For the subjects that had transmitted Type B vaccine virus, this subject was identified to be shedding on Day 15. He had at least ten other negative cultures obtained on other days.

The subject did have contact with two cold adapted recipients who were known to shed Type B. The B Ann Arbor isolate detected from our placebo recipient was detected five days after the last identified shedding by the vaccinee. This placebo subject had coryza on days eight or 18, cough on days eight or nine, and irritability on day zero, 14, and 17.

Essentially this reactogenicity profile was similar to direct recipients of cold adapted vaccine.

The probability of transmission is provided on this slide by the original analyses, including isolates identified as vaccine virus equals one, the Type B, among 57 eligible placebo subjects. The rate of transmission is 1.75 percent with those confidence intervals of .1 to 8.75 percent.

For the all available transmission of 93 subjects, assuming that all subjects with unidentified virus had vaccine strain, that would be an N equals five. The rate of transmission would be 5.38 percent with those confidence intervals.

To try and help correlate this with the Reed-Frost, this placebo subject had contact with two vaccinees. The Reed-Frost data presented by the sponsor, the rate of transmission was less than one percent per vaccinee. So a rate of 1.75 for two vaccinees, it's fairly comparable.

Genotype was performed on selected shed isolates. They were chosen for genotyping if the culture grew a single virus subtype, if it had a viral content of 100 to 1,000 infectious foci, and the sponsor also chose later isolates to try and increase the chance of identifying nucleotide changes.

The consensus genomic sequence of the isolates from all placebo subjects and the subset of FluMist subjects were compared to the sequences of the relevant monovalent viral harvest, i.e., H1 was compared to H1, et cetera.

The B Ann Arbor transmitted isolate had three nucleotide changes identified. Fifty-five of a possible 237 isolates from FluMist subjects were tested, and nucleotide changes were identified in 76 percent of H1N1, 92 percent of H3N2, and 77 percent of B.

Of note, the B isolate with these nucleotide changes was found to be identical to the B isolate recovered from the vaccinee.

Genotype and phenotype stability showed that the nucleotide changes were not random, but look at H1 and H3N2 together because they have the same internal donor gene sequence. So for Type A, the changes were most commonly identified in PB1, 2, NP and the M genes, and for Type B they recurred in the M gene.

All of the tested isolates, N equals 55, maintain their attenuated, cold adapted and temperature sensitive phenotype. Evaluation of the retention of attenuation of the shed viruses in the animal model is ongoing.

In conclusion, there are several limitations for this trial. Despite its labor intensive nature, it is a small sample size and there were many protocol violations. Shedding of vaccine virus was frequent, occurring in about 80 percent of subjects and lasted through day 21, which was the last day cultures were obtained.

Transmission did occur. However, we can estimate only accrued rate. The recovered vaccine viruses had a high frequency of nucleotide changes. However, the cold adapted and temperature sensitive phenotype markers were retained. These changes were not random, but the clinical significance is not known and evaluation of the attenuation in animal models is still ongoing.

Additional assessments of shedding were also performed by the sponsor. In four trials routine cultures were obtained from day zero to ten post vaccination. Of those 569 cultures, about 34 and a half percent of the subjects were identified to shed vaccine virus, and illness cultures were obtained from about 13.8 percent of those. Forty of 290 cultures obtained from FluMist subjects with illness post vaccination were identified to shed any influenza virus, i.e., its type not identified.

Twenty of those 40 isolates were tested by genotyping and phenotyping, and of those tested, they proved to be vaccine strain.

And the genotype and phenotype stability of the strains recovered from the ill subjects, of the 20 vaccine strains isolated from these ill subjects, none were reported to have altered their cold adapted or temperature sensitive phenotype. As most of these trials were performed when wild type influenza was circulating, primarily H3N2 and B, there's no resortment of vaccine strains and wild type were not identified.

I briefly will review the selected data for post vaccine solicited reactogenicity events. The sponsor gave you much of this data this morning, and I will do a very brief summary from 06 and 09.

The selected REs included runny nose, nasal congestion, sore throat, cough, irritability, headache, chills, myalgia, decreased activity, and fever.

The events were solicited for ten days in 06 and seven days in 09, the adult subjects.

I have selected only a few events, and I chose the ones that were statistically different between FluMist and placebo in dose one for children 15 to 71 months of age in AV006.

As you can see, any RE, that is, the subjects had one or more reactogenicity event, occurred more than 60 percent of the time for FluMist subjects and placebo subjects after dose one and dose two.

Runny nose was significantly different as were vomiting myalgias, and low grade fever. After dose two, generally speaking the rate of reactogenicity events were a bit less common.

In year two no statistically significant differences in REs between the FluMist and placebo group were identified. The rates of REs I said before tend to be less frequent with runny nose and congestion occurring in about 42 percent of subjects and coughs in about 24 percent, and the rates of REs were similar in subjects who had received one dose or two doses in year one.

This shows the results for the AV009, the adult trial. This is the entire study cohort of 18 to 64 years of age. Again, you can see the rate of at least one RE event was 70.9 percent for FluMist and 61.9 percent for placebo subjects. There were significant differences noted for runny nose and congestion and sore throat.

The rate of fever over 101 was very uncommon, occurring in less than .6 percent of both groups.

To summarize, between FluMist and placebo groups, significant differences were observed for runny nose and low grade fever for children and runny nose and sore throat for adults. REs occurred commonly in both groups, more than 60 percent in children and adults. We have no solicited RE data for seven to 17 year old subjects. There are no apparent differences in RE rates presented by the sponsor this morning by age group for subjects less than 50 or those 50 to 64 years of age.

Most of the safety data available were generated in healthy subjects.

I'd now like to review additional concerns from VRBPAC of 2001. First we'll start with pneumonia. There was much concern at VRBPAC for possible increase in pneumonia after receipt of FluMist in AV006, year one. However, at that time not all studies were finalized and submitted for review.

Upon assessment across all 20 trials now finalized in the database, no increase in pneumonia, bronchitis or bronchiolitis events have been identified.

This gives those to you as data. In AV006, the relative risk at that time you saw in VRBPAC last year was 3.48 with wide confidence intervals. In 09, in 12 to 59 months, no significant increase nor for the 60 to 107 months, no significant increase, nor for the older age group of children.

We also discussed abdominal pain last year. There had been an increase in relative risk for abdominal pain observed in 06 in year one after dose one. This was a solicited event post vaccination in that study, and the relative risk was 2.69, with a confidence interval or lower bound above one.

In the updated assessment across 20 trials we still observed the increase in 06, but there's a decrease in the relative risk in subjects less than nine years in 019, but remember these are medically attended events, not solicited events.

There's also a decrease in subjects 18 to 64 years observed in 09. Again, no intussusception, intestinal obstruction or mesenteric adenitis were reported.

For rare events potentially related to influenza in the review across all 20 studies, there were no reported cases of encephalitis, encephalopathy, Guillain-Barre, or Reye's Syndrome. There was no significant increase for central nervous system events, including seizures following the receipt of FluMist identified.

Additional concerns included concurrent immunization data. No data for efficacy or safety with concomitant immunizations in any age group are available. For the use of FluMist for five to 64 years of age, the possible concurrent vaccinations for the four to six year old age group include DTAP, MMR, and inactivated polio vaccine for its possible use with pneumococcal vaccine.

Additional concerns included the need for annual revaccination data. At this time we have no data for revaccination of adults. For older children and adolescents from study AV012, there are 1,054 subjects who were immunized in both years one and two from whom data are available, and 459 of these were from the ten to 18 year old age group.

Those 459 had similar MAARI results and SAE results as presented for the entire study cohort. For young children under nine years of age, AV006 has year two and year three data -- at that time it was study 015 -- for safety and for efficacy.

As I reported earlier, there was no increase in reactogenicity events identified in repeat vaccinees, and there was demonstrable efficacy in year two. Year two included 1,358 subjects, and of those, 375 were over 60 months of age.

In previous VRBPACs, not just 2001, there have been discussions about annual release of influenza vaccine. Clinical testing for the new strains for code adapted influence of vaccine has been performed by the sponsor. As you recall, the master seed virus is a 6:2 reassortant, which contains six genes from the attenuated master donor virus and two genes, the hemagglutinin and NA, of the wild type strain, and these are formulated annually.

The goal of this study was to test attenuation of newer reassortant strains in humans before incorporation to the trivalent preparation. The primary objective was to assess the safety of the new strain as demonstrated by similar rates of fever, defined as oral temperature above 101, in adults, CAIV, and placebo recipients from day zero to seven. This was based upon the fever rates observed in adults in 009.

For 2002 master virus seed vaccine contained ten to the seventh TCID50s for B Hong Kong in normal allantoic fluid. Plans, healthy adults randomized four to one with safety monitoring from day zero to seven, day zero to 14, and SAEs from day zero to 28 with a six month follow-up included.

Reactogenicity events were predefined as listed earlier. The study provided approximately 98 percent power to rule out a five percent absolute increase in fever, assuming a rate of fever of less than one percent in the control group and that the true difference between the groups was zero.

A total of 330 adults were enrolled. Two hundred and sixty-four received the col adapted vaccine, and 66 received placebo. We have the initial safety phase data from days zero to seven. For the rate of fever above 101, there was one cold adapted recipient and no placebo subjects. The confidence interval around this was minus 4.9 to 2.3.

Thus, the study met its primary endpoint of less than five percent different. Also, there was less than or equal to five percent difference for runny nose and sore throat, which had been statistically significant different in AV009.

Additionally, this study demonstrated the feasibility of performing annual testing.

I will briefly review studies submitted in support of efficacy. This will be abbreviated as you heard much of it this morning.

First, 06 was the pediatric efficacy trial, U.S., multi-centered two year, prospective, double blind, randomized FluMist to placebo in a two to one ratio in healthy 15 to 71 month old children. It was initiated in the 1996-97 flu season with one does in the two dose regimen, with two doses given 60 plus or minus 14 days apart, both being evaluated.

The primary endpoint was the first episode of culture confirmed influenza any time, on the day of, or after receipt of the second dose of vaccine.

There were several additional secondary endpoints which will not be discussed this morning.

At the time there was no H1N1 circulating in year one or year two, and thus, we do not have field efficacy data for this strain. You saw these data this morning, but just as a reminder, there was H3N2 and B analysis performed for efficacy against any strain, the number of culture positives. You can see for those who received two doses it was 93.4 percent with tight confidence intervals. Those enrolled in one dose was 88, eight percent with a little wider confidence intervals reflecting the small size, and for all randomized participants, the efficacy was 92.6 percent.

Efficacy by age group was also performed, and you can see for each group of interest over 60 months of age against any strain, the efficacy estimate was 90.6 with those confidence intervals.

In year two, as I mentioned, 1,358 subjects returned. They received one dose of the same study vaccine that they had received in year one, i.e., they were not rerandomized.

The circulating strain that year, H3N2 was A/Sydney, and it was a variant from the vaccine strain of A/Wuhan. So against all community acquired influenza, there was 87.1 percent efficacy. Against A/Sydney, a variant, there was 85.9 percent efficacy with confidence intervals of 78 and 91.9.

An evaluation of efficacy by age group, again, for subjects over 60 months of age -- and remember they're now a year older -- efficacy against any influenza was 86.9.

Because there was no circulating H1N1, the sponsor performed a challenge study with a subset of subjects from 006. The goal was to compare viral shedding of vaccine strain, cold adapted, monovalent, A/H1N1 in previous FluMist recipients compared to previous placebo recipients.

A total of about 222 subjects, or 20 per site, were challenged with A/Shenzhen, H1N1, five to eight months after their year two vaccination. The challenge vaccine was the same lives as the H1N1 used in the year two vaccine.

After challenge, viral shedding was assessed on days one through four. The results showed 82.9 percent protection against shedding of the vaccine strain.

In the adult effectiveness trial, healthy working adults 18 to 64 years of age were randomized two to one to receive one dose of FluMist or placebo. The primary effectiveness objective was to show a smaller proportion of FluMist compared to placebo recipients had any febrile illness, AFI, during influenza outbreaks.

There were several additional secondary endpoints and illness evaluations, including severe febrile illness, SFI, febrile URI, FURI, CDC influenza-like illness, which is defined as fever plus cough or sore throat on the same or on consecutive days.

Additionally, comparisons of effectiveness for subjects under 40 and over 40 was defined prospectively, and a comparison of effectiveness for subjects under 50 and over 50 was performed in a post hoc analysis.

This slide shows the effectiveness against illness during influenza outbreaks for the entire study cohort, looking at percent reduction, FluMist compared to placebo. For AFI it was not statistically significantly reduced. For SFI, FURI, and CDC ILI, all three illness categories did have significant reductions as you can see for the FluMist recipients.

The rates of AFI associated events in the total study cohort are shown on this slide. The days of over-the-counter medication use, days of antibiotic use had significant reductions. Days of health care provider visits and missed work days for AFI were reduced, though not statistically significantly.

The number of subjects enrolled by age group are shown on this slide. As you can see, for the 50 to 59 and 60 to 65 year olds, they were a smaller percentage of the overall study enrollment. In the pre-specified analysis in over 40 and under 40, percent reduction for illness categories is shown in this slide. It's 9.3 and for under 11.2 for over 40, and the differences with statistical significance are shown by asterisk for FluMist compared to placebo. There's no pattern. There was a decrease under 40 for SFI and for CDC ILI, but for those over 40, it's seen for FURI and CDC ILI.

In looking at subjects under 50 and over 50, again, asterisk shows significant decrease for FluMist compared to placebo subjects, and for those under 50 there was significant decrease for SFI, FURI and CDC ILI. No significant decrease in any of the illness categories were observed for subjects 50 to 64 years of age in this post hoc analysis in a small number of subjects.

For AFI, SFI, FURI, and CDC ILI, though there were no decreases in the occurrence of those illnesses, subjects 50 to 64 years of age compared to subjects under 50 did have greater reductions in illness associated events of missed work days, antibiotic use, and health care provider visits.

Additionally, a study of wild type influenza challenge was performed in adults. The objective was to assess the efficacy post challenge with wild type influenza against laboratory documented influenza illness in adult subjects. this is FluMist compared to placebo and FluMist compared to TIV, and TIV compared to placebo.

Laboratory documented illness included symptoms of influenza with identified shedding of wild type influenza and/or serologic response defined as greater than or equal to a fourfold rise in hemagglutinating antibodies to the challenge virus.

Because there was a small number of subjects evaluated, with about 30 in each group, the laboratory documented illness for all strains combined showed efficacy of FluMist against the challenge of 85 percent and for TIV of 71 percent. These were statistically significant.

However, the study was not powered to assess efficacy against individual strains.

Efficacy conclusions. Efficacy against culture confirmed influenza illness was demonstrated after one or two doses in healthy children 15 to 72 months of age in year one and after revaccination in year two.

Efficacy was demonstrated for children in the subgroup of 60 months to 72 months in AV006. These are the only efficacy data available for children 16 months to 17 years of age. No field efficacy data for H1N1 are available.

Effectiveness conclusions. Effectiveness was not demonstrated in healthy working adults against the primary endpoint of AFI. It was observed against SFI, FURI, and CDC ILI in a post hoc analysis of the CDC criteria.

For the post hoc analysis of subjects greater than 50 years of age, there were no significant decreases in AFI, SFI, FURI or CDC ILI. Efficacy against culture confirmed influenza was not assessed in adults.

For safety conclusions, a total of 14,154 individuals, 60 months to 64 years of age have been vaccinated with FluMist. There are few subjects in the database over the age of 50 years. An increased risk of asthma events was observed in study AV019 in 12 to 59 month old subjects, but was not seen in subjects over 60 months of age.

No significant increased risk for pneumonia events was identified. Overall SAEs were recorded in less than one percent of the subjects.

With this review I'd like to acknowledge the great help of the clinical review team, Dr. Douglas Pratt, Dr. Antonio Geber, and Dr. Wasima Rida, our statistical reviewers. Our BLA committee chairperson is Dr. Roland Levandowski.

I now present for the committee the questions or would you like to ask me questions first?

CHAIRMAN DAUM: I think we'd like to ask you questions first and maybe begin --

DR. MINK: It was worth a try.

CHAIRMAN DAUM: We're awash in data here and probably need some clarification, and we can review these questions before we start the afternoon's discussion.

DR. MINK: Okay.

CHAIRMAN DAUM: So let's go to the committee for clarification questions at this point. Dr. Edwards, Decker, and Stephens.

DR. EDWARDS: I have two questions. The first is in assessing the reactive airway asthma data, it's interesting that there does not appear to be a temporal clustering of those, and if you look at the time between zero and 14 days, there did not appear to be more disease than if you extend it, which may be a function of a healthy patient effect.

But I wondered what your thoughts were about lack of temporal clustering. It seems that the biologic basis of reactive airway disease would have some temporal relationship to the administration of vaccine. So that was the first question I had.

DR. MINK: I don't have data to give you an easy answer. The temporal clustering was performed really primarily as post hoc, although there was some plans for it the subject in AV019, to look at temporal clustering for events. So most of the data for temporal clustering are from AV019.

It's my understanding though with some respiratory viruses, such as RSV, that some of the reactive airway effects are long seen after natural infection and that they aren't necessarily temporally clustered.

DR. EDWARDS: But RSV and flu are different viruses. So okay.

The second question that I'm a bit confused about is the age 50 years to 64 year data. The ACIP has made recommendations in years with plenty of vaccine that that is a group at high risk or higher risk for influenza, and you just showed us some data from the effectiveness study that would suggest that perhaps the age 50 to 64 age group does not have as much effectiveness as in the younger group.

But then in a subsequent slide, you said the effectiveness looks good in that. So I'm a little confused in that. Could you please clarify that?

DR. MINK: Sure. The analysis in the subjects over 50 and under 50 was post hoc. The study was designed, and I believe implemented before the CDC recommendations for the over 50 age group was in place. So in a post hoc analysis, we looked at the over 50/under 50.

The confusion is for the occurrence of the illness categories, AFI, SFI, FURI. There was no decrease in the over 50 group for any of those illness categories.

The next slide shows that for the illness associated events, such as health care provider visits, antibiotics, those events, there seemed to be a decrease in the 50 to 64 years even though there was no decrease in the occurrence of illnesses.

Does that clarify for you?

CHAIRMAN DAUM: Thank you.

I have Drs. Decker, Stephens, Faggett, Diaz, and Myers.

Dr. Decker, please.

DR. DECKER: Kathy asked the questions I had about efficacy. I've still got some questions about the safety data.

The occurrence of asthma in the younger children, which although there's a question about the cluster, it seemed as I saw the data you presented and in the pre-read, it looked as though this was popping up in multiple looks at the data, making it more convincing it might be real.

But it also looked to me as though it was only really being seen after the first dose and wasn't an issue in the second dose. Did I see that correctly?

DR. MINK: According to the data from AV019, the statistically significant increase was only after dose one. It was not significantly increased after dose two in the 12 to 59 months or the older children either.

DR. DECKER: Which led to a question in my mind. It made me wonder if prior experience with influenza virus perhaps through inactivated vaccine would convert these first exposed children to second exposed children. Are there any data on that?

Has there been any look at what happens if you give the FluMist product to people who previously received, and particularly to these young children who have previously received the inactivated product?

DR. MINK: I don't have data for a subset of subjects who receive flu inactivated vaccine and then subsequently received FluMist. However, I know some subjects in the database, you know, going back in their histories had received flu. I don't have a large enough or even a comprehensive evaluation to tell you what the reactogenicity profiles of those subjects are.

CHAIRMAN DAUM: Dr. Stephens, please.

DR. STEPHENS: My questions are somewhat similar and deal largely with the over 50 group, which I think is problematic for at least some of us. I had a similar question to Dr. Decker, and I think you answered it in terms of prior influenza inactivated vaccine in that particular cohort. You really don't have that information; is that right?

DR. MINK: I don't have the data separated out by who received flu vaccine previously.

DR. STEPHENS: As I understand this group, we're talking about an n number of 511; is that correct or close to?

DR. MINK: For FluMist.

DR. STEPHENS: For FluMist in the over 50 age group.

DR. MINK: In AV009, right.

DR. STEPHENS: In 09, which is the large part of the data set for this particular -- so we're talking about relatively small numbers in this particular category.

DR. MINK: Right.

DR. STEPHENS: There was a VA study mentioned earlier by the sponsor. It looked like that study was withdrawn from the license application as well as two other studies. Can you comment on that?

I mean, that would be more data that I think some of us would be looking for in this particular group.

DR. MINK: Those subjects weren't in the age cohort being requested, and they were not healthy subjects. They had at least one risk factor for needing inactivated vaccine. So they're not included in the database in healthy individuals five years to 64 years of age.

CHAIRMAN DAUM: Thank you.

Dr. Faggett, please.

DR. FAGGETT: Concerning the Texas HMO trial, I'd like a couple of questions. Number one, what population was studied there, and could you let me know if the Medicaid population was included in that study? That's the first part of the question.

DR. MINK: Yes. It was predominantly subjects from Temple Belt in Texas, with most of them being seen at Scott & White Clinic HMO. About 80 percent of the subjects had Scott & White coverage, but many of the other 20 percent were Medicaid.

The ethnic diversity of the study was representative of Texas.

DR. FAGGETT: And the purpose of the study was to look at herd immunity, and the safety information was secondary to that?

DR. MINK: In the original design.

DR. FAGGETT: Okay. So I'm just not clear on what conclusions. The study raises more questions than it answered. It would appear as if the implication there is increased risk of asthma and wheezing in young children from that study.

Is there any other conclusion I could draw from that?

DR. MINK: I have the data presented that there was no increase in MAARI events in year one, but there was an increase in year two. The year one started in August, and everybody finished before flu season started. Year two started about a month later, and not all of the subjects completed their follow-up before flu started. So there's a lot of complicating factors in that study.

DR. FAGGETT: So it's really difficult to draw a good conclusion from it.

DR. MINK: The conclusion that I provided for you is that 9,500 or so doses were given, and the rate of SAEs was comparable to the randomized double blind trial of about .02 percent, and that -- there was one more -- and that there was no rare events associated with influenza identified in that trial.

DR. FAGGETT: Yes.

CHAIRMAN DAUM: Thank you.

Dr. Diaz.

DR. DIAZ: Just two questions. In reviewing the duration of shedding for the different influenza strains, you made a comment about the H1N1 data, the range being out to 21 days, and yet that was the last date that culture shedding data was obtained.

DR. MINK: Right.

DR. DIAZ: Are we to assume that with the H3N2 and the B strains since the range is three to 17 and one and 15 that there's actually endpoint data at 21 days to show that there was no shedding at that time?

DR. MINK: Yes. Twenty-one days was the last day cultures were obtained.

DR. DIAZ: Okay. Thank you.

And then perhaps my colleagues that have raised questions about the concern about the reactogenicity associated with the first dose disappearing with the second dose and questions about perhaps data about prior immunization with the current influenza vaccines, the killed vaccines, I just wanted them perhaps to clarify for me the interest in having that data.

These seem to be very different vaccines obviously, and presumably individuals especially in the older age groups, if they hadn't had experience with prior immunization, would certainly have had experience with prior influenza. So is it the egg product or what is the interest in the prior immunization with the killed vaccine?

DR. DECKER: Well, I'll answer my question. My thought was that if it's, in fact, the infection with the vaccine virus in the FluMist that's triggering in some small proportion of children these reactive airway events, that in these kids who were getting a two stage immunization, two doses a month apart, if you first immunize them with the killed vaccine, which should not trigger the airway event, you may induce enough immunity that they're protected from the reactive airway disease when they get the dose of the live flu. So you may get the best of both worlds and avoid this possible reactive airway disease pathway.

So it's a speculative question, but there are some other vaccine situations where we've seen similar things happen.

CHAIRMAN DAUM: I really want to interject at this point and return to the presentation from the FDA about this FluMist vaccine. This certainly is some interesting speculation that if time permits we could explore this afternoon, but we're not really talking about a sequential schedule at this point. I'd like to focus on this presentation right now.

Pam, were you done or did you have a follow-up?

DR. DIAZ: I had a follow-up, but we can do it over lunch and then see if we need to bring it up.

CHAIRMAN DAUM: Perfect. Thank you.

Dr. Myers and then Dr. Eickhoff, Dr. Edwards.

DR. MYERS: I think Dr. Faggett addressed part of the issue. There are a lot of issues with the 012 study, but to get at the issue of timing and frequency of asthma and URI, whether this is a normal rate of illness in the normal population and so on, as I understand the analyses that were done, we looked at comparison periods before and remote to the immunization, and the only data we've seen is relative risk and confidence intervals. Could we actually see illness rates in the comparison periods? Would that be of value, or are you seeing so many problems with the study that it wouldn't be?

DR. MINK: There are a lot of comparisons, and there are several limitations to the data. So I think it would be with caution for interpreting the actual event rates.

CHAIRMAN DAUM: Thank you very much.

Dr. Eickhoff, please.

DR. EICKHOFF: A comment and a question. My institutional memory of this committee is probably as good as anybody's and is really better than most, but it is difficult to remember any occasion in which the sponsor's analysis of the data and the FDA analysis of the data had such a high level of concurrence. So that tells us something.

My question is, Mr. Chairman, not a question of the FDA data, but a general question which I'll be happy to save for this afternoon.

CHAIRMAN DAUM: I would prefer that, Dr. Eickhoff, if that's okay with you. I think we really want to focus on debriefing all of the aspects of this that we need to hear about, but please don't throw your comment away and please return to it.

I have Dr. Edwards, Gellin and Goldberg.

DR. EDWARDS: I want to just make a comment that I think some studies that we did almost two decades ago now that were funded by NIH that looked at John Maassab's product, not the FluMist product, and subsequent studies have shown very clearly that the more experience that you have had with a flu strain, either with inactivated vaccine or with subsequent infections, it appears that the titer of cold adapted virus that would grow in people that have had less immunity is different than in people who had more immunity.

So I think it is relevant and makes wonderful biologic sense that the first dose, if it was going to be more reactive, would be because the titer of virus would be greater than the second where you would have some preexisting immunity and the titer would be less.

And I think that's what's being shown also in the reactogenicity data in the young children as opposed to the older people. So I think it is biologically relevant, and I think it sort of addresses Pam's question, but would also make it, I think, more comforting that if you had been previously primed with an activated vaccine and then followed cold adapted; that I would anticipate there would be less of an issue with reactogenicity than with more.

CHAIRMAN DAUM: Thank you very much.

Dr. Gellin.

DR. GELLIN: Let me just follow on to Kathy's question and ask a different one.

Was there a correlation between reactogenicity and shedding?

DR. MINK: Actually that question was asked this morning, and as I recall Dr. Connor said he didn't have data directly assessing that, with the primary reason being that most of the concerning adverse events were identified after the studies were completed.

Shedding was only routinely performed in four of the trials. AV002 was one of the very early trials, and AV011, which was the challenge study, and then the Wyeth transmission study.

I guess it's still possible to look at reactogenicity data compared to the shedding for those subjects, but that has not yet been done.

DR. GELLIN: And then just a point of clarification, in the many slides you showed us on the SAEs your conclusion was no pattern seen among the various events. Would you elaborate another half sentence on that?

DR. MINK: Yes. Actually in your briefing document, I provided a line listing of all the SAEs for year one and year two in AV019, a total of what is it, 24 events? And there is really no pattern.

Some of them were accidents. Some of them were, you know, surgery for a toe injury. There's a collection.

CHAIRMAN DAUM: Dr. Goldberg, please, and then Dr. Snider.

DR. GOLDBERG: Can you just clarify for me? You said that 20 vaccine strains were isolated from ill FluMist subjects and that there was no reassortment in those 20 strains. Does that make you feel comfortable that we don't have to worry about this?

DR. MINK: Dr. Levandowski, would you like to comment about the reassortment from the 20 ill? Is he still here?

DR. GOLDBERG: I mean, I can clarify my question a little more.

DR. LEVANDOWSKI: I'm Roland Levandowski. I'm with the Center for Biologics also.

I don't know that there's much of a comment that we can make on that. There's a small number of strains. It's the information that's available to us, and to the extent that that says that reassorting doesn't occur, I think that's the only conclusion we can draw from that small sample that we see. It's reassuring, but we can't necessarily predict that some event couldn't occur in the future.

DR. GOLDBERG: My concern is that observing nothing in 20 could still be compatible with a fairly substantial underlying rate, and I was wondering if you had any further data or from other models or animal work or the sponsor perhaps.

DR. LEVANDOWSKI: I think that was addressed by the sponsor also this morning in talking about seeing reassorting occurring between the cold adapted strains within the vaccine recipients. There is some evidence to suggest that.

And we know from natural occurring influenza, one example this past year, the appearance of H1N2 influenza viruses is a natural reassorting event between the HAN1 strains that have been circulating and the H3N2 strains that have been circulating.

So reassorting events would be expected to occur, but there's no evidence that we have that that's occurred with the cold adapted strain.

DR. GOLDBERG: Thank you.

CHAIRMAN DAUM: Dixie.

DR. SNIDER: Yes. Two comments. One, my recollection, and Nancy can correct me, Roland, if I'm wrong, but the H1N1 issue, of course -- I mean, the statement is true that there's no field efficacy data for an H1N1, but the fact is there hasn't been much H1N1 circulating.

So to try to get that data we don't know how long we'd have to wait, and so I think, you know, this year maybe, maybe two years, who know when? So, you know, the challenge data may be the best surrogate.

The question I had is, or maybe it's a reflection, is that, you know, I certainly think the comments that have been made about the use of inactivated vaccine in advance of using FluMist may have some merits, but there's also the issue that is not addressed in the database we have unfortunately, at least not that I'm aware of, but I have a question about it, and that is natural infection causes reactogenicity, runny noses, and I would presume precipitates asthma attacks, maybe even more severe.

So one of the potential things that could be happening here with the vaccine strain is that you're eliciting mild flu symptoms with an attenuated strain, and hoping that you're trading off those milder episodes of reactogenicity and milder episodes of asthma for more severe episodes.

I realize that may be hypothetical, but the question for the sponsor and FDA is is there any data in any of these databases that might support the hypothetical statement I just made.

CHAIRMAN DAUM: Do you want to answer that briefly? It's sort of an afternoon question.

DR. MINK: Okay. Then I'll see --

(Laughter.)

DR. SNIDER: Okay. I yield, Mr. Chair.

CHAIRMAN DAUM: Well, perhaps we can save it.

DR. SNIDER: I can eat, you know, now.

CHAIRMAN DAUM: No, we've got to address this, and there's at least three committee members that would like to hear a discussion about it, but would you be upset if we deferred it until after we came back?

DR. MINK: I wouldn't.

(Laughter.)

CHAIRMAN DAUM: My goal here is to try to get us to lunch at 12:15, and Dr. Katz is waiting to speak. So we'll hear from him, and then I think we'll hear from Ms. Fisher, and then we'll take a lunch break.

DR. KATZ: I guess mine is more a generic question. We've heard about from both the FDA and the sponsor about reactogenicity, effectiveness, efficacy. I've heard nothing about antibody data, and I'm wondering have those disappeared from our radar screen. What is their relevance?

Basically, you know, if we're looking at some of the old experience, you know, and initial exposure to given antigens or epitopes, and then second, with new strains of influenza coming along presumably in the cold adapted vaccine or any others, you may reinforce the antibody response of your primary, but your new ones may be less robust, which is the current adjective people seem to use.

Do we have any antibody data or do we consider them important anymore?

DR. MINK: We do have antibody data. Those were presented last year, and we elected not to revisit the serologic profiles this year. The primary reason we elected not to revisit serology this year is that no correlate has been identified following FluMist. Actually probably, depending on who you ask in the room after an activated vaccine with without a clear correlate for serology and even more of a quagmire is perhaps the local immunity, nasal immunity, that we chose not to revisit the issue this year.

CHAIRMAN DAUM: Ms. Fisher.

MS. FISHER: The subject with transmitted vaccine virus had flu illness symptoms. How confident are you from this data that the transmission rate would only be about five percent if the vaccine was used on a widespread basis.

DR. MINK: The data I presented are the only data that I have estimating the rate of transmission or the probability of transmission, and in our conclusion we felt most comfortable saying that transmission does occur, but at this time we can only estimate a crude rate.

CHAIRMAN DAUM: I feel confident we will return to the issue of transmission this afternoon.

We're going to take a lunch break of 60 minutes duration. There are two choices within the hotel: Allie's American Grille or the Lobby Lounge. Good luck, everybody, getting served.

We will convene at 1:15 sharp with the open public hearing and go right on to the afternoon's discussion.

(Whereupon, at 12:18 p.m., the meeting was recessed for lunch, to reconvene at 1:23 p.m., the same day.)

AFTERNOON SESSION

(1:23 p.m.)

CHAIRMAN DAUM: Can everybody please take the seats and call the meeting to order for this afternoon?

Is Dr. Sachs in the room?

While we're waiting for Dr. Sachs to come I'd like to correct for the record an issue that came up this morning. During Dr. Young's presentation, he kindly attributed to me the benefits of having done some studies on influenza virus as it relates to FluMist. However, wrong guy. I'm not the person who did those studies, have never done studies with influenza virus or FluMist or else I would not be sitting here today.

Those studies were done by Dr. Maassab, and the record should show M-a-a-s-s-a-b is the person who performed those studies.

Thank you very much for lengthening my CV, but I can't accept that at this moment. We now -- or any other moment. It's my last meeting.

We're now going to convene the afternoon's business and begin with the open public hearing. I'll turn the floor over to Dr. Sachs for that portion of the meeting.

DR. SACHS: This is the point where anyone from the public is welcome to speak briefly. No one has requested time from me earlier, but you're more than welcome now, you know, to come up and announce yourself.

If there's no one who wishes to speak publicly, I'm going to close the open public hearing and resume our meeting.

Thank you.

CHAIRMAN DAUM: Okay. What I'd like to do is our style in the committee is to allow the committee to explore whatever issues initially interest them that arose from this morning's presentations.

I think in order to understand what field we're playing on it might be good to ask Dr. Mink if she's here to put the questions up briefly so that we can just see them, and then we'll have committee members as they wish raising issues of their pleasure.

DR. MINK: Question one for the committee is for safety and for a vote.

One (a), are the data adequate to support safety of FluMist for individuals five to 17 years of age, 18 to 49 years of age, 50 to 64 years of age?

Please consider data related to respiratory events, such as asthma and URI, shedding and transmission of the vaccine strains following receipt of FluMist, and annual revaccination data. If the data are not adequate for specific age groups or there are other safety concerns, please discuss what additional should be requested.

Question two, efficacy for a vote. Are the data adequate to support efficacy of FluMist in individuals five to 17 years of age, 18 to 49 years of age, 50 to 64 years of age?

If the data are not adequate for specific age groups, please discuss what additional data should be requested.

Discussion point number three. Clinical studies for release of new strains. Please comment on the design and endpoints for the clinical studies performed in adults for the release of new strains.

Discussion point four. For additional studies, if the data are adequate to support safety and efficacy, please discuss what additional information, if any, should be requested from post marketing studies.

That's all.

CHAIRMAN DAUM: Thank you very much, Chris.

So the first two questions are going to need a committee vote. The third and fourth questions are discussion questions, and so we can leave the questions now and digress really to issues that committee members would like to explore based on this morning.

So, Dr. Stephens, would you start us off, please? Then Dr. Cox.

DR. STEPHENS: Could we clarify that the questions relate to healthy individuals? Is that correct? So the word "healthy" is a part of it?

DR. MINK: The word "healthy" should be in the slides, yes.

CHAIRMAN DAUM: That's an important clarification, David. Thank you.

Dr. Cox.

DR. COX: Yes. With regard to the question about transmissibility, I'm wondering why investigators were not able to obtain the original clinical specimens from those individuals who were believed to shed influenza viruses and then just PCR and sequence the viral nucleic acid to find out if those were actually wild type or cold adapted viruses.

CHAIRMAN DAUM: Okay. For this question we will turn to the sponsor. Do you care to respond to that?

Please identify yourself. Thank you.

DR. COELINGH: Dr. Coelingh from Medimmune.

The nasal swabs, Nancy, that were taken, I'll just go through how the study was done. The nasal swab specimens were taken in Finland, and they were inoculated onto MDCK cell cultures in duplicate. One of those cultures was used for immunostaining with monoclonal antibodies to identify whether there was an influenza virus present in that specimen.

The other specimen was frozen and sent to Medimmune where we did phenotypic analysis, genotype, et cetera, and subtype analysis to identify what type of strain was in the specimen.

Okay. Those specimens were frozen and sent to Medimmune for further analysis. Some of those viruses we could not identify a virus in that isolate.

And to go on to the rest of your question, in the ones that we couldn't identify any virus to subtype, we went back and inoculated those into eggs, and we blind passaged in both MDCK cells and Rhesus monkey kidney cells and had HAs on those and also tried to RTPCR virus using primers specific for the M and the MP genes, and we were unsuccessful at doing that even though all of the controls worked appropriately.

CHAIRMAN DAUM: Thank you very much.

Dr. Eickhoff, please.

DR. EICKHOFF: Let me readdress the question that was on my mind earlier this morning and I think on Dixie Snider's mind as well, and that relates to the general area of wild type influenza and asthma in healthy children in the one to five year old age group.

I know that Ken McIntosh 25 years ago did some studies of wild type influenza in children at National Jewish Hospital, and so these were children with preexisting airways, disease, and I don't remember the exact nature of it, but not surprisingly influenza exacerbated their reactive airways disease at least as measured by pulmonary function tests.

But what do we know about wild type influenza and asthma related events in healthy children in that age range?

Any pediatrician on the panel may respond to that.

CHAIRMAN DAUM: Kathy, would you like to start that response?

DR. EDWARDS: I think the data that I can recall most quickly are data that were generated from a database, the TennCare database, which although it doesn't fund doctors well at all, it does create a database that you can query in terms of events.

And in some data that was published in the Journal of Pediatrics or Pediatrics the year before last, Kathy Neuzil looked very carefully at the TennCare database and showed that the numbers of visits to doctors' offices and individuals that had previously been given either bronchodilators or had an ICD-9 code of asthma was significantly increased during the flu season when it was compared with other individuals.

So obviously that's retrospective data, but I think showed quite clearly that children with asthma during flu season will have more encounters, more wheezing episodes, more illnesses than those who do not.

CHAIRMAN DAUM: Thank you.

Other committee members? Dr. Snider.

DR. SNIDER: Well, I just would add as an allergist-immunologist who tries to keep up with the literature, I think there is, just to build on what Kathy said, an increasing appreciation of the role of the number of respiratory viruses in triggering episodes of asthma. So I think the idea that natural influenza infection would be responsible for triggering episodes of asthma in children who are so predisposed is not only hypothetically and biologically plausible, but has, indeed, started to gain scientific evidence base.

CHAIRMAN DAUM: I'd like to throw out a couple of questions just for comment. One of them was inspired by a comment that Dr. Katz made this morning, and that is that supposing there is transmission, the first question is is that a good thing or a thing that we should be concerned about.

The model of OPV was mentioned this morning as one kind of vaccine that did have transmission in one part of it's life in this country. That was a helpful thing in terms of preventing polio. Is that a good thing?

And then the second thing is transmission to whom exactly, and what do we need to know, if anything, more than we do now about what kinds of recipients, hosts the viruses might be transmitted to? And where are we in terms of our understanding about that?

Those are two questions that I would love -- I don't know the answers. If people would comment about them. Does anyone want to speak directly to them? Dr. Overturf?

Dr. Markovitz, I've got you in line when we change topics, b ut I'd like to explore this just a little bit first.

DR. OVERTURF: Actually I think it's a major issue, and I agree with Dr. Myers' comment this morning that it seems to me that when you move the shifting of the vaccine program, and we actually have shifted with this vaccine a concern using large amounts of vaccine in otherwise normal, healthy individuals, and before the issue had been primarily giving to children who were at higher risk and elderly adults.

But when you do it in this population, I would like to have more information about how the vaccine virus -- what it will do in a large, open population in terms of transmissibility to others and whether that transmissibility actually increases the hazard to populations that we really haven't addressed, like asthmatics, patients with COPD, patients who are smokers, and a lot of other issues.

And I don't think that the data that we have now really addresses those issues because it's too small. But I agree that moving it from a population of high risk, moving it from a situation where you're protecting high risk populations to healthy populations, the issue is are we going to actually incur more problems than we actually solve in that setting.

I think that needs to be addressed, the safety of the vaccine. When I was going back over the historical data that was provided by the sponsors in terms of persistence of virus, there was very little data particularly in the population over 50, again, about how long the virus is there and if it's transmitted in that population, and I think that's an important question.

CHAIRMAN DAUM: Thank you.

Could we keep the discussion on this issue for a few minutes?

Dixie and then Pam.

DR. SNIDER: Well, just to build on what Gary just said, I think one other thing to throw in that mix is whether those other populations have received inactivated vaccine and how that might impact on whether they have any adverse outcomes as a result of this transmission is, again, something we don't know, but is important to know because there may be a way if there is significant transmission, and there were these adverse events that would occur in individuals, it's conceivable that they could be attenuated if the high risk individuals received inactivated vaccine before they were exposed to the attenuated vaccine.

CHAIRMAN DAUM: Dr. Diaz.

DR. DIAZ: Likewise that was the major reason for my question earlier about in household transmission, was the concern about potential high risk individuals within a household setting where a healthy individual in that household may actually be immunized with this vaccine virus.

And likewise I was not necessarily reassured, and perhaps the manufacturer can comment, if there were household transmission studies done with that particular interest in mind, i.e., transmission to high risk individuals.

And finally, the issue about shedding in children out to 21 days, likewise if there's any data or should there be data about potential shedding in higher risk individuals who may shed a much longer period of time.

CHAIRMAN DAUM: Does the manufacturer wish to respond to the question of Dr. Diaz?

You can use either this mic or that mic, totally your choice. Easily solved.

DR. CONNOR: Let me for a moment try to just put into perspective some of our thinking about the transmission issues.

First of all, I think as we mentioned earlier this morning there are some data that look at transmission in the historical studies. Those were referred to earlier. There are a number of studies that have looked in various settings, including household contacts and husband-wife pairs, that have not been able to demonstrate transmission in that setting, although obviously those were smaller studies than these are.

DR. DIAZ: And not necessarily in a setting where that household partner is a high risk.

DR. CONNOR: Is a high risk patient, right, although there have been high risk patients studied in that population, including patients that in normal populations of families, but not specifically targeted at the high risk populations.

The other sort of factors to think about are that there are a number of factors that are going to influence vaccine virus transmission obviously. They're going to be either the frequency of virus shedding; how many people in the population are actually shedding virus; the level and the duration of shedding; the susceptibility of the contacts as referenced to the prevalence of TIV in vaccination in which you've got seropositive contacts shedding for shorter periods of time and the intensity of the contact.

And if you then just think about the next issue, which is that the inoculum of the shed virus is obviously much less than the inoculum of the vaccine administration so that you're actually getting less virus in a transmission situation than you're getting in active vaccination, but there's a low probability of transmission. How you calculated the probability is quite low. Our estimates based on the Reed-Frost model would be that it would be approximately .006 or so.

We've also done some work with Ira Longini who has looked at the population modeling of the FluMist transmission models, and using the transmission rates that are the 006 rates, if you calculate the reproductive number associated with this virus, it turns out to be well less than one.

So, in effect, mathematically it's implausible that the transmission of the vaccine virus would actually represent any more than some secondary transmission, but it wouldn't persist in the population is what those calculations tell us.

And there obviously then is a low rate of transmission in the older kids and in adults compared to the children in the day care settings that were studied.

I think for most of us -- George, the next slide for a second -- this just illustrates something that I've shown before, but maybe we can go to the next one first, which I think what we're talking about are what are the consequences of transmission. I mean, what are the problems if transmission actually occurs, and I think I've already addressed the first one, which is there's a risk to the population, and we should understand what that is in terms of the risk of transmission.

Ira Longini suggests to us that that is not possible to perpetuate the CAIV virus in the population.

The other individual questions are what is the risk to individuals, and the individuals of interest are either an immunocompetent host so that the transmission occurs in the household or someplace else, to another person who is immunocompetent. A couple of things are going to happen, not the least of which is that that person is going to respond to the vaccine, and they're going to shut off virus shedding.

So at worst transmissions with immunocompromised hosts would be equivalent, although it would actually be less because the dose would be less, than giving that person FluMist.

The immunocompromised individuals, a setting in which what one thinks about, are first of all the virus is temperature sensitive. So presumably transmission to an immunocompromised host going to produce an upper respiratory tract infection, and that it would remain limited to the upper respiratory tract by and large, but prolonged shedding would probably happen. The immune response would not allow the virus to be shut off quite as readily in an immunocompromised host than in a immunocompetent host, and that would allow -- and obviously there are antivirals that could be used to treat such situations.

And the other possibility, as folks were mentioning, is so what's the probability in an asthmatic. Again, if you think about the risk, it is at least equivalent or no worse than the data that we provided to you about FluMist.

So you may see some low frequency of wheezing and exacerbation and that is balanced against the benefit of the vaccine which prevention of wild type influenza in the first place.

And if you're interested, we have thought a bit about so what is the -- I don't know. For me medically, I sort of have to put that into some context. What's the actual risk that something would happen in the population?

If you think about that and just go on, this is the kind of way in which we've been thinking about it. The first is that you have to make certain assumptions in a simplistic kind of modeling approach, and the assumptions would be that the inoculum from the transmitted virus was equivalent to giving the virus. I mean, that's what we know is what would happen under those circumstances.

If you assume that the transmission rate in the household setting or in the school setting, here we did it for thinking about an immunocompromised person in school; that the transmission rate would be what was in day care. That's the worst case scenario. It clearly would be better than that in other circumstances.

And if you estimate the number of immunocompromised hosts that are out there in the school age population in a six year or seven year school situation, the prevalence if you add up the cancer transplant, chronic steroids, HIV, congenital immunodeficiencies that would happen in that population would be about .0015 percent. The rate is about that.

If you assume that everybody was fully immunocompromised so they continued to shed virus and they were all in school, which is obviously an -- these are all very conservative kind of estimates -- the calculated risk of transmission based on the day care transmission rate would be something like .0000009. So the rate is low in those kind of settings that an individual would transmit to an immunocompromised host in the school system, and it's likely to be less than that because, in fact, the inoculum is not equivalent to the vaccine dose, and there are other characteristics.

And you can go through those same kinds of exercises for asthmatics and look at excess risk associated with the population also, and just the next slide, George.

You go through the same sort of analyses, making certain assumptions, and if you assumed that the asthma wheezing risk was as bad or what it was in the patients that were less than 60 months and the history positive patients, you'd calculate a risk of excess asthma and wheezing that also would be quite low.

So I don't know if that helps. There obviously is a lot of assumptions built into that kind of thinking, but it gives some kind of context in which to think about sort of the medical circumstances in which those things would happen.

DR. YOUNG: Dr. Daum, can I make one more comment?

CHAIRMAN DAUM: Please. A succinct one.

DR. YOUNG: Okay. George, can you just put up that table?

I want to just clarify something for everybody because we talk about transmission is transmission and it's the same in everyone. We need to look at some of the fundamental shedding infectivity data for different populations that we know about with CAIV.

Ed showed this slide early on. In children you see much higher levels of shedding when you infect them with CAIV. Their peak titers are two to three logs of virus, and their duration of shedding is four and a half to nine days.

Adults shed less. They shed far less virus. You can see this one to two to three logs lower, and the duration of shedding is only one or two days.

So when we're talking about adults who get immunized with this vaccine, they're making a lot less virus for a shorter period of time than day care kids are shedding where they're seronegative.

And then look at the infectivity when you do challenge studies with different strains of CAIV. You need two and a half to four and a half logs of CAIV to infect the child. You need five to six and a half logs to infect a human.

So if you just think about --

(Laughter.)

DR. YOUNG: Sorry. To infect an adult.

CHAIRMAN DAUM: There you go again.

DR. YOUNG: There I go again, Dr. Daum.

(Laughter.)

DR. YOUNG: These were studies Dr. Daum did -- only kidding, only kidding.

If you look at how much a child sheds, it's virtually impossible for them to infect an adult. If you look at how much an adult sheds, they can't infect a child or an adult with the amount of virus that's being shed. So the likelihood of transmission from adult to adult, adult to child, or child to adult is very low, very much lower, for sure than child to child where they're seronegative. They're making more virus. They're more infectable.

So don't think about transmission as the same from person to person, whoever get the vaccine. I think you need to remember that that's the case with this virus. It's different than wild type where you're getting more levels of shedding than you do with CAIV.

CHAIRMAN DAUM: Dr. Young, I think you had too much sleep last night.

(Laughter.)

CHAIRMAN DAUM: We're going to take one question from Dr. Steinhoff about -- oh, could you leave that on, whoever took it off? -- about this very slide.

Dr. Steinhoff, this very slide.

DR. STEINHOFF: It's about the previous slide because I just wanted a clarification, and others may all understand this, but I don't quite understand how you use that probability number from the Reed-Frost calculations.

You quoted a number somewhere between .006 and .04 as a probability, and I'm gathering, but I don't know; perhaps you could explain. Is that the probability derived from the study in Finland of a single excreter infecting at least one other person in that setting? Is that what that is a probability of?

I'm not sure. I mean, what's the numerator and what's the denominator?

DR. KOHBERGER: Bob Kohberger from Wyeth.

That transmission probability is one to one, one infected, one susceptible. Yes, that's susceptible, becoming infected.

DR. STEINHOFF: That means then you'd have to multiply this by a variety of factors, how many shedders you have and how many susceptible you have, which then brings in all of the other factors of how many people are immunized and how many are susceptible in actual contact; is that correct?

DR. KOHBERGER: That is true. If you have 100 infected and one susceptible, they're more likely to get infected than one to one.

DR. RIDA: Might I offer a suggestion?

Wasima Rida, Biostatistics at CBER.

The basic reproduction number is a product of that transmission probability times the expected number of contacts in an entirely susceptible population. That's the formal definition.

So to calculate the basic reproduction number, you would have to know what is the expected number of contacts a FluMist vaccinee would come into contact during their infectious period. So that's basically what a basic reproduction number is.

CHAIRMAN DAUM: Thank you.

I think that was a clarifying discussion.

Dr. Hamilton.

DR. MINK: Dr. Daum.

CHAIRMAN DAUM: Dr. Mink.

DR. MINK: Actually on the table, can you show that again, please? That one.

Please can you tell us are those subject FluMist recipients or does this include all of the historical experience, NIH, et cetera, for -- okay. Thank you.

CHAIRMAN DAUM: All right. So we'll try to move on. This has been very helpful though. I think it has clarified some issues and some uncertainties perhaps.

Dr. Hamilton, you've been patient. Then Dr. Steinhoff and Katz.

DR. HAMILTON: Just one more question. When child is used, does it refer to everyone under 17? Is any data on this broken out between a five and five to 17?

CHAIRMAN DAUM: Would the sponsors care to respond to that?

DR. HAMILTON: When you use the word "child," what age groups are you referring to? Because we're asked to opine on one to five years and then five to 17 years.

CHAIRMAN DAUM: I don't think we're going to be asked to opine about one to five. Maybe we could see the table again that Dr. Hamilton is trying to explore here.

DR. HAMILTON: So children is one to 17, and we don't have data broken out for the five to 17 year olds, and that's what we're asked to opine on.

CHAIRMAN DAUM: Can you come to the microphone? Tell us who you are.

DR. MURPHY: My name is Brian Murphy.

CHAIRMAN DAUM: Any connection to FluMist, please?

DR. MURPHY: Excuse me. My name is Brian Murphy. I'm from NIAID.

We did these studies. The children here are young children generally, approximately six months to three years of age, who were chosen to be seronegative to the virus that was given. So they represent basically immunological versions.

So those HID50s are basically what you would see, the infectivity titrations of viruses in individuals who have no immunity to the virus. The adults are people who are age 16 to probably 40 or 50 at the time these studies were done.

There's one other point I just wanted to make. The question of household transmission that you had talked about. This virus has been given. The Influenza A virus has been given to two month old children, to very young infants, and was shown to be safe in that age group.

Conversely, it has also been given to very old individuals, definitely greater than 65 years of age at ten to the seven in both cases, and it has been shown to be safe in those populations.

So even if it does transmit, if it transmits to individuals at the very ends of the spectrum of age, the virus should be safe in those rare cases of transmission if it occurs.

CHAIRMAN DAUM: Dr. Katz, we're staying on this transmission issue for now. Is your --

DR. KATZ: Yeah, I'd like to set this into an appropriate context. We're spending all of this time talking about transmission of an attenuated virus. We're not contrasting it with the transmission of wild virus. It's a matter of education, not of science, of how we use influenza vaccine. All of these household contacts, all of these high risk people by our current recommendation should have been immunized with trivalent inactivated vaccine.

So the issue becomes moot if we have program that work, if we educate health care people, if we educate the public.

So I think we're losing perspective in focusing so on transmission of the attenuated virus. We're not even contrasting it with what happens with the wild virus if a child brings it into the home or if a day care person.

All of the routes of transmission we're talking about, we're trying to interrupt the transmission of wild virus. I don't think we're appropriate in placing such emphasis on the transmission of attenuated virus.

CHAIRMAN DAUM: Dr. Stephens and then Dr. Snider. We're staying on this subject till we're finished with committee interest, and then we'll move on.

DR. STEPHENS: It would at least seem to me on the surface that the 019 trial where you have this Kaiser database of families could be useful in assessing some of the questions at hand in that would it be -- and maybe it has not been done -- but it would be helpful to have looked at the family issues of asthma or other respiratory illnesses or other events occurring in families of the children with this database.

And my question is really to the sponsor, whether that was looked at at all to help address this particular transmission issue.

CHAIRMAN DAUM: Dr. Mendelman.

DR. MENDELMAN: Paul Mendelman, Medimmune Vaccines.

Specifically in the 019 trial it was not addressed. However, there was an intriguing analysis in the pivotal efficacy trial, AV006, Dr. Belshe, and in that trial, it provided indirect information about the lack of transmission, and what was done in that trial is that -- and Bob is here to either correct me or comment if needed -- because more than half of those children were in families that were vaccinated in that 1,602 child trial and allowed the opportunity to look at a sibling who got vaccine and a sibling who got placebo compared to singleton placebos.

And it turned out that the children in those families that had both a vaccinee and a placebo the attack rate in those placebo children for wild type flu during the flu season was the same. They had the same attack rate essentially. So they were not protected by the child, a sibling of theirs giving them transmitted vaccine virus and protection downstream.

It's indirect. It's intriguing, but it is information.

And, Dr. Daum, you did ask about what data we do have. So, George, you had put up the HIV adult slide. This is a study that was -- because you're talking about transmission and say, well, what's the end result if it gets transmitted to somebody, and the DMID study 98005 done in collaboration with the NIH was done and is presented in the BLA, and in that study of 50-some adults by placebo or vaccine, shedding was looked at in the first three days, seven to ten days, and a third time subsequent to that.

It's in the short set. Okay, great.

So shedding was done on day three to five, seven to ten, and day 28 to 35, and the results are shown, and there was shedding of Influenza B virus in one HIV infected subject on day five, and nobody in the further days when culturing was done.

Now, in the published literature not presented in the BLA, a very similar study design in pediatric HIV done by the NIH in collaboration with us, and in that study there was no prolonged shedding either in the vaccine to children. There were 24 HIV infected children, and they were one to eight years of age with a mean of 4.2 years.

CHAIRMAN DAUM: Thank you, Dr. Mendelman.

Dr. Snider and then Dr. Steinhoff on this issue, and maybe we can move on to something besides shedding at that point unless there are issues the committee has not addressed as yet.

Dr. Snider.

DR. SNIDER: I think maybe my question has been answered, but let me just phrase it a different way. I seem to recall prior to lunch there was a quick statement made about other studies that were not part of the BLA having been done that look at transmission and demonstrated no transmission.

So I guess the question is: is there any other data that has not yet been presented about transmission or lack thereof that needs to be brought before the committee?

CHAIRMAN DAUM: Let's ask Dr. Mink first for a comment and then the sponsor.

DR. MINK: The data that we have for evaluating transmissibility is the Wyeth-Lederle transmissibility study in Finland, which we presented to you after a total review.

The other data, I actually think you have already seen data from the sponsor. They are historical. The table that they presented they said was based on historical data.

So I'm not aware of any other data that has been submitted for review.

CHAIRMAN DAUM: Thank you.

DR. MINK: And I should have seen it probably.

CHAIRMAN DAUM: Are there other data?

Thank you. I think that answers that question.

Dr. Steinhoff, you wanted to comment on this very issue.

DR. STEINHOFF: The only comment I wanted to make was to follow up on something that Dr. Katz said. If you look at the literature about household and family transmission of wild type influenza virus, numerous studies show that the transmission rate is between 20 and 30 percent in a family setting of wild type virus contrasted to what I'm hearing is somewhere between one and five percent from the Finland study depending on what you pick.

So there is a relative ratio that one can consider.

CHAIRMAN DAUM: Thank you, Mark.

The historical data that are being referred to can be found in Table 36, Dr. Overturf reminds us, of which presentation?

DR. OVERTURF: It's Table 36 of the sponsor's presentation.

CHAIRMAN DAUM: Thank you very much.

Ms. Fisher, germane to this issue?

MS. FISHER: Well, I just have to make a comment about Sam Katz's comment about comparing the life virus flu vaccine to the live polio vaccine, and that is that the flu is not polio, and we have moved away from the live virus polio vaccine to a killed vaccine precisely to prevent vaccine induced polio.

So I don't think the comparison is correct.

CHAIRMAN DAUM: Thank you very much.

Can we perhaps move on? Are there other points about transmissibility that haven't been addressed? I think we've had a pretty thorough discussion about it.

Dr. Snider, did you wish --

DR. SNIDER: Not about transmissibility.

CHAIRMAN DAUM: Well, Dr. Markovitz is ahead of you if we're going to move on. So you're second.

But any other issues with transmissibility? Fair warning, although we can return to it if someone needs to.

Dr. Markovitz.

DR. MARKOVITZ: Yes. I wanted to follow up on a question Dr. Myers posed earlier, which I thought was very appropriate, and I think we need to hear a little bit more about, and that has to do with just how reactogenic the allantoic flu and placebo was, and I'm wondering whether there are some data comparing this to a true placebo because obviously the vehicle for the vaccine does count in terms of how patients perceive what's going on.

And I'm wondering. From the tables it's kind of hard to tell when it says, for example, that a patient had a runny nose or whatever. I mean, was this the runny nose to end all runny noses or was this a little runny nose?

In other words, is this something that would have patients coming back and saying, "I'll never get that vaccine again," or is it just one of those little things?"

And I don't know that I've heard any data to speak to this issue.

CHAIRMAN DAUM: Let's reframe that or frame that by are there data comparing reactogenicity with other placebos? I think that's fair.

DR. MARKOVITZ: I think Ed Connor started to answer that before, in fact, but I'm not sure I fully followed it. So --

CHAIRMAN DAUM: Let's hear from Bob Belshe and then Ed Connor and if there's other data bearing on this issue, let's hear them next.

DR. BELSHE: Hi. I'm Bob Belshe from St. Louis University, and I've been involved in a number of these clinical trials.

The data we're displaying here is a clinical trial of a different intranasal vaccine. These are the saline placebo recipients in young children age six to 18 months. Now, they selected at time zero on day zero for the absence of illness. Specifically, they do not have a cough. They do not have a runny nose.

And you can see at the end of day one 12 percent have a runny nose. They're receiving saline intranasally.

And so what we believe this phenomenon represents is selecting a healthy population who normally have about 20 percent of the group having rhinorrhea or runny nose, and so this is a return to the mean by day five to seven. We've returned to the mean where about 20 to 25 percent of children here have a runny nose.

And so this is, I think, a very good example, which is a true saline placebo and not normal allantoic fluid and reflects what we've seen in the FluMist trials.

CHAIRMAN DAUM: Bob, that's very helpful data.

Dr. Connor, did you want to expand on that?

I have Dr. Snider next.

DR. SNIDER: Well, I was going to change the topic unless you wanted to stay on it.

CHAIRMAN DAUM: No, I think we've addressed the question, and we can move on to the topic of choice.

DR. SNIDER: Okay. The topic of my choice at the moment has to do with the efficacy data and particularly the fact that we're being asked to make some comments about the adequacy of data to support the efficacy of FluMist in individuals five to 17 years of age and 50 to 64 years of age.

And as has been mentioned earlier, although there is quite a bit of safety data available for the five to 17 year age group, there is at least not in the BLA data with regard to efficacy.

And then as has been pointed out by FDA and others, the number of people in the 50 to 64 age group is considerably less than in some of the younger adult groups.

And I was wondering if the sponsor could make some comments about why that was the case, what the problems were in trying to get numbers of people there in those age groups or if there is historical information they feel, you know, has an important bearing on this issue, if they could tell us what that is.

CHAIRMAN DAUM: Thank you very much, Dixie.

Is there a sponsor response?

DR. CONNOR: I think there weren't any particular difficulties in getting patients into the study. The studies were simply done as childhood studies that went up to 71 months of age and as adult studies that included patients all the way up through 64 years of age.

Our view, I think, is that we've clearly been able to demonstrate efficacy in children, and we've been able to demonstrate effectiveness in adults, and that there doesn't seem to be a biologically plausible reason why the middle group would have any other different effect.

The studies were simply designed and executed in the way that they were looking at specific issues in children and adults.

The issue of the 50 to 64 year olds is a post hoc fact, that is, that's how many patients were in the trial as the trial was recruited, and I think that our view of looking at that data is that, first of all, when you look at all of the specific or the more specific influenza measures within the AV009 trial, you see reductions in measures of effectiveness.

In addition to that, when you actually look at the days of illness, clearly there weren't differences when you look at the group as a subset among the occurrence of illnesses, except in the DOD ILI definition, but across all of the other measures of effectiveness for severity of illness, days of illness, day of missed work and antibiotics, you see significant effects in all of those measures.

So I think the perspective is that we've demonstrated efficacy in children. We've demonstrated effectiveness in adults. The issues about any of the age groups go to the question of whether there's any evidence that there were differences in the populations either of children and adults in that population, and we actually have seen no substantive evidence that there's any difference.

Certainly in the pediatric population the data that we showed suggests that there is not even a trend to anything happening as you get to the edges of the population base, and the data that we've shown in the 50 and 64 year olds, while smaller in that population, doesn't have significant evidence that that population is substantively different than the population as a whole.

There weren't any other issues or difficulties related to actual inclusion of those other populations.

DR. SNIDER: So if I understand correctly, you're saying from a biological standpoint you feel like we should be able to extrapolate to the five to 17 year old group based on the other data.

And then as far as the 50 year olds and older, I guess I was wondering. I mean it may have just come out that way depending on who was served by the particular caregivers who participated in the trial, but I was also wondering if maybe people were excluded because there were at that age a lot more people who wind up with contraindications, and that may have been a reason why they were smaller.

DR. CONNOR: Maybe I can ask Kristin Nichol, who conducted the effectiveness trials to make some comments.

DR. NICHOL: Kristin Nichol from the VA Medical Center, Minneapolis, University of Minnesota, one of the investigators for AV009.

Dixie, with regard to the specific question about enrollment, there were no specific difficulties of which I'm aware, and of course, we conducted AV009. This predated the ACIP recommendations putting people 50 to 64 in a high priority group because about 25 to 30 percent of them are high risk.

By the way, the ACIP is not suggesting that healthy people 50 to 64 are high risk. They are high priority because about a quarter to a third of them may have a high risk condition.

So there were no specific issues. We certainly did exclude participants who had any of the ACIP indications for vaccination at the time that we conducted the trial.

With regard to the question about evidence of benefit or lack thereof, perhaps I could make a comment as well. Again, the 50 to 64 year old high priority designation from the ACIP came after this trial. So this is truly a post hoc analysis.

We did pre-specify an analysis by age 40, under and over 40, and did not find any evidence of a differential effect. With regard to the 50 and over, there are fewer subjects, only about 640. So we do have limited power to have precision in our estimates or to find significant P values.

However, as summarized in this slide, which looked at the febrile upper respiratory illness category, you will see that the confidence intervals for people 50 to 64 around the point estimates for effectiveness across the different outcomes categories include the point estimate for the entire group, as well as the point estimates for people 18 to 49.

There is imprecision in those estimates, particularly in occurrence in days of illness, somewhat amazingly actually given the sample size. We do find statistically significant benefit when we look at days of work lost, health care provider visits, and antibiotic use.

If we looked at other health care illness definitions, you would see a virtually identical pattern.

But in any event, we are not able to see any evidence of a differential effect by age.

CHAIRMAN DAUM: Thank you very kindly for those comments.

Dr. Myers, please, and then Dr. Gellin.

DR. MYERS: Well, I had a comment and then a question.

The comment, Ms. Fisher, is that between 20 and 25,000 Americans die every year from influenza, and so it is a significant disease, and although we don't see the damage from polio as a consequence of influenza, it is disease of significant morbidity and mortality, and I just wanted to correct the record on that.

My question for the sponsors, I was really glad to see the HIV data because that helped a lot as we're struggling with these questions. With that said, the absence of data on high risk patients is really striking from your application, influenza being a disease with specific high risk groups.

And the reason, and I was wonder if there's other data. That's my question. The concern, of course, is that as a licensed vaccine is utilized, we've already seen the asthmatic in the children, but they're going to be high risk people immunized inadvertently either accidentally by transmission or more commonly because they just don't know their high risk because they have underlying diseases.

And so I was wondering if I could ask the question about the absence of the data, your plans to collect that data or what your thoughts are about inadvertent administration of vaccine to people who have high risk conditions.

CHAIRMAN DAUM: So that's a very worthwhile question. We'll ask Dr. Mendelman to respond, and I'd also like to hear from Dr. Mink on this question as well.

DR. MENDELMAN: The question is broad, and prior to formerly Aviron pursuing the studies or initiating the studies with FluMist, there were 99 peer reviewed journal articles in the literature and 32 review articles, and that went back about 25 years, and in those studies, they were published, and I'm sure there were file drawered studies that were not that we're aware of also, children with cystic fibrosis were studied; children with bronchopulmonary dysplasia were studied; children with asthma were studied.

Now, these are small numbers, Dr. Myers, going back, but when you add them all up, they come up to that number of 8,091 that was shown on one of the initial slides by Dr. Young across those 25 years of studies.

Now, studies in chronic obstructive pulmonary disease were also, you know, conducted as well so that there's a supportive data base that goes back 25 years, but the data that the FDA has to review on our file are there that we've submitted to the agency.

And so there's a background of supportive information that, you know, one could give credit to because they are the Maassab master donor viruses that were made in those monovalent and bivalent.

And if I could have that first slide, George.

The largest study, of course, is Dr. Edwards' study, and the question has been brought up about repetitive -- it's in the new slide set, George -- and it was conducted over five years. It's one of the largest or largest studies between 1985 and 1990, 5,210 participants, one to 64, and most of them were over 15 years of age.

And in that study, and the design is shown here, there were -- of the four seasons when flu circulated, two were H1N1 seasons and two were H3N2 seasons, and the next slide will show the data for all participants, and you can see that this was compared to inactivated vaccine; that in 1986, 1987, that the cold adapted bivalent, 78 percent with confidence intervals you see, and the inactivated vaccine, 79 percent. That's with the H1N1, and 1988, 1989, 90 percent, point estimate for cold adapted and 74 percent for the inactivated.

And then in the two H3N2 years, 59 percent and 56 percent for cold adapted and 71 percent and 79 percent for the inactivated.

And in the publication Dr. Edwards notes there was no serious reaction to the vaccine across these five years of study.

Subsequent to that, in the next slide, Dr. Neusel (phonetic) pulled out the pediatric data with Dr. Edwards from that trial. There were about 800 children in that trial who received a bivalent cold adapted, and you see here the efficacy similar for both the inactivated vaccine and the bivalent vaccine in that trial by Dr. Edwards of Vanderbilt.

CHAIRMAN DAUM: On this very subject and very brief.

DR. CONNOR: I just wanted to as Brian Murphy again if he wanted to comment at all on the high risk populations that have been studied previously.

DR. MURPHY: No.

DR. CONNOR: Okay. That's fine.

CHAIRMAN DAUM: Thank you very much.

Dr. Mink, could you speak for the agency on this issue?

DR. MINK: What we consider as part of the BLA is the product is manufactured by the sponsor with the clinical safety data submitted in support of the labeling indication requested. So what we consider in reviewing this product are the 20 studies that have been reviewed and presented to you.

CHAIRMAN DAUM: I guess if I could press you just a little bit, if there were a licensure of such a product is there a position yet or is it too early to ask as to what would be required afterwards in terms of assessing these issues?

That's what I really --

DR. MINK: That's discussion point number four.

CHAIRMAN DAUM: You will hear from us.

(Laughter.)

Dr. Edwards, this very issue?

DR. EDWARDS: Yes.

CHAIRMAN DAUM: Okay.

DR. EDWARDS: I have a question, particularly in the 50 to 64 age group. Why the approach was taken to not do an efficacy trial or given the fact that there is a licensed product that now is suggested to be given in that age group, why there might not have been studies that compared the licensed and the unlicensed product.

I think effectiveness measures are generally not what we see for licensure.

CHAIRMAN DAUM: Sponsor want to speak or FDA want to speak to that question? It's a good one.

DR. NICHOL: I guess I'm the elected official, or unofficial person.

CHAIRMAN DAUM: Thank you.

DR. NICHOL: With regard to effectiveness versus efficacy, it's my understanding that this was discussed at some length before the onset of the trial with various people. I'm looking to the sponsor here, but I'm quite sure there were some discussions with FDA and others with regard to whether or not effectiveness might be an outcome as opposed to culture confirmed efficacy that would be acceptable.

Of course, this is a randomized, double blind, placebo controlled trial. When we looked at effectiveness outcomes rather than efficacy, what this meant for us was several things.

One was that we were interested in looking at a real world outcome, and of course, in the real world most often we do not have culture confirmed influenza that we're looking at. We are looking at people coming into the medical care community with influenza-like illness.

By choosing a less specific outcome than culture confirmed influenza, which is very specific, we realize that we inflated our sample size need substantially, but we really wanted to have a real world look at what would happen in a population if you immunized them with live attenuated influenza virus vaccine.

Recall, as was discussed earlier today, that when one sees a reduction of, for example, 34 percent in influenza-like illness, if one backtracks to what that might have corresponded to if one had culture confirmed influenza, in a clinical trial of the inactivated vaccine conducted over two years and the second year or actually in the second year of the trial, which was the year after our study was done, the efficacy against culture confirmed influenza, the specific outcome that we're most used to seeing perhaps was 86 percent.

But when they translated that into a reduction in influenza-like illnesses, the kind of clinical effectiveness that we saw, they saw a reduction of 34 percent.

So, yes, we did choose the effectiveness outcome. We were very interested in a real world outcome as opposed to the culture confirmed outcome which doesn't replicate what happens in the health care provider's office.

DR. EDWARDS: Was there a thought to including in that an inactivated arm?

DR. MINK: At the time that we did AV009, we already had a sample size requirement that was substantial because we were looking at a clinical effectiveness outcome, and we chose not to really inflate the sample size requirement by going to a three arm study.

CHAIRMAN DAUM: Thank you.

I think Dr. Gellin has been patient.

DR. GELLIN: Perhaps a related question, but I'll ask the same question that Kathy had, but maybe inverted and maybe as has been set up as a real world example, maybe this is an un-real world question. So it's a question of the institutional memory that I don't have that other people in this room do.

How often do new products come to a committee like this when the question is efficacy and the data is effectiveness?

CHAIRMAN DAUM: I think we're going to ask for some agency input on that question.

Dr. Midthune.

DR. MIDTHUNE: I can't think of any. Obviously we have brought products here where the efficacy parameter is the immunogenicity comparison, but those are for products where there have been previous clinical disease endpoint studies.

CHAIRMAN DAUM: Thank you.

Dr. Decker, some industry perspective on this question?

DR. DECKER: Yeah, just a reminder that I think it was the last meeting of this committee we voted or the committee voted to approve a license extension based on effectiveness data, if I remember correctly, Prevnar and otitis media.

That wasn't a totally new product. It was already licensed, but it was an extension of the indication.

DR. MIDTHUNE: That was an extension of that indication, but there were data that actually showed tympanocentesis results where there were actual serotyping of the pneumococcal isolates derived from that.

That was the Finnish study. There was also the Kaiser study, which just looked at acute otitis media, but there was both in that particular application.

CHAIRMAN DAUM: In fact, the issue might have been viewed quite differently were there only effectiveness data in that instance.

Yes? Did you want to speak to this again? Would you please? Obviously an important issue for us.

DR. NICHOL: Forgive me for coming to the microphone again. Kristin Nichol from Minneapolis, one of the AV009 investigators.

I forgot to mention perhaps, after Kathy's question about effectiveness, just to remind the committee that there is a challenge trial demonstrating efficacy against culture confirmed illness among adults. It's a relatively small trial, but it is an efficacy trial looking at wild type challenge, and efficacy was 85 percent against all three wild type strains combined. The study was not sized to be able to look at efficacy for each type specifically.

CHAIRMAN DAUM: These subjects were 18 to 41 years of age?

DR. NICHOL: That's correct, and they were randomized either to placebo or to receiving the FluMist or trivalent inactivated vaccine.

CHAIRMAN DAUM: Other committee comments?

I guess I would take the initiative then and ask in studies like these or others where we've had efficacy data presented against culture confirmed influenza, do we know anything about the people who failed. Has there been any attempt to study the folks against whom efficacy did not occur? And do we know anything? Is there anything special about them or unique about them that we should hear about?

Anyone want to take that question on? Dr. Mendelman?

DR. MENDELMAN: Again, the proof of principle that we got was the pediatric study, which was culture confirmed and large. The efficacy trial in adults that did involve a TIV arm gave us proof of principle that we could go on and do the large effectiveness trial in adults.

And the supportive data and multiple efficacy trials was submitted in the license application in the historical review section, 7.8, which is certainly available, and that reviews those efficacy trials that were done.

In those, for the committee, and we could present them on screen if you would like, but the range is, you know, wide, but the overall is in that 70 to 90 percent efficacy for the cold adapted vaccine across multiple efficacy trials, but the largest is Dr. Edwards' that we've noted to the committee, which was comparing it with TIV.

DR. CONNOR: Bob, just to answer your question, I think that the --

CHAIRMAN DAUM: Please.

DR. CONNOR: I think that in most of those trials obviously the efficacy in the pediatric trials were quite high. I mean well above 90 percent, and there obviously were very few of the patients who failed, and we haven't actually characterized those patients any further, but the numbers are really very small also.

CHAIRMAN DAUM: Dr. Katz.

DR. KATZ: This is perhaps to beat a wounded if not a dead horse. That is that antibody studies sometimes help.

(Laughter.)

DR. KATZ: And I was going to ask Nancy Cox because each year when we review new influenza virus vaccines at CDC, she presents us data on antibody response, HAI antibodies usually to new strains, the cross-reactivity with other strains, and you must have some feeling, if not data, not relating to these studies, but what are the necessary antibody studies and the antibody levels which will guarantee you protection against influenza.

DR. COX: I thought you knew better than to ask that question.

(Laughter.)

DR. COX: There is no specific antibody level that guarantees protection. The level of an HA titer of 40 or greater is often used as an index of protection in the studies that we do and the studies in various vaccine trials. That level is expected to protect about 50 percent of the vaccinated population.

What we can say generally speaking for an activated vaccine is that the greater the antibody level, the better. For live attenuated vaccine, there hasn't been as good a correlation of antibody levels with infection. So there probably are other factors, including local antibody, that are contributing to protection.

But there is definitely some correlation with antibody, even with the live attenuated.

CHAIRMAN DAUM: Thank you, Dr. Cox.

Dr. Belshe, you wished to speak to this very issue?

DR. BELSHE: Yes. I'd like to just contribute a little bit. Regarding the breakthrough infections in AV006, there were a small number of vaccinated children who did develop natural influenza. Those illnesses were significantly shorter in duration in terms of days of fever, less than two days of fever compared to placebo recipients, which had an average of approximately five days of fever.

So there was a more mild illness in those breakthrough infections.

Regarding the correlates of immune protection, we did extensive studies on behalf of the NIH in the AV011 trial, which was a challenge study using vaccine virus as a challenge. We were able to demonstrate that secretory IgA and serum HAI antibody were independent of correlates of immune protection, and that there was very weak correlation between those two correlates.

So if you had either antibody or secretory IgA, you were significantly protected against a vaccine virus challenge. They were very powerful correlates. Approximately 85 percent were secretory IgA, and more than 90 percent for serum HAI antibody.

CHAIRMAN DAUM: Thank you, Bob. Those are very helpful comments.

Let me call on Dr. Overturf. I apologize.

DR. OVERTURF: Just as a follow up. In the document that we've all been given, it looks like the historical data on antibody levels shows that the cold adapted vaccine in terms of fourfold immune responses is only about half as good as TIV, maybe 60, 70 percent in some studies.

So that brings up the other issue, and I thought maybe Dr. Cox would address this, is what is used currently standardized each year's lot of vaccine if it's not antibody. Obviously since systemic antibody doesn't seem to be as good with cold adapted inactivated vaccine, what will be used or is that going to be necessary?

DR. COX: I think that that's a question that could best be answered by my FDA colleagues. We're involved in regulatory issues.

CHAIRMAN DAUM: Dr. Levandowski, you want to catch the pass here?

(Laughter.)

DR. LEVANDOWSKI: I'll do my best. The questions about potency of inactivated vaccines, that's not done by antibody tests using antibodies from animals or from people. It's done by a method that's called single radial immunodiffusion, which is an immunologic type of test, but it's an in vitro test, which is done with a standard antigen, comparison between the standard antigen and the test antigen, which would be the vaccine, and from that the quantity of antigen that's present can be quantitated.

That has been correlated with immunogenicity in studies that were done way back during swine flu and the return to the H1N1 viruses in the late 1970s. So we use the antibody testing in terms of looking at whether current vaccines are likely to make antibody responses that are reasonable to newly circulating strains, but we don't use that so much as a tool to determine whether the vaccine -- let's see -- what the potency of the vaccine is.

I'm not sure I've answered your question.

DR. OVERTURF: Well, the question is: has somebody thought about what will be used as the standard measures for this vaccine, which may not be the same as used in previous influenza vaccines.

DR. LEVANDOWSKI: Well, now you're asking about potency of this vaccine, I guess is the question.

DR. OVERTURF: Yes.

DR. LEVANDOWSKI: The potency of this vaccine is based on infectious units. So the number of infectious units in the product is what that will be based on.

CHAIRMAN DAUM: Thank you very much.

What I'd like to do is to change gears a little bit now, and we'll take a very short break, but before we do that, we'll put the first question back up on the screen, and when I come back, when we come back, I'd like to have the committee focus on additional issues that need to be explored to deal with the first question.

I think you'll find that most of them have been explored, perhaps not all of them, and then we will begin the process of actually voting and being heard on these questions.

Before we take this break, I'd like to make a brief presentation of my own.

(Laughter.)

CHAIRMAN DAUM: Through the miracle of Jody Sachs, the committee has been able to procure a present, and the present, of course, is for Dr. Kathy Zoon.

We just learned this morning of your decision to move over to NCI and leave FDA, and it's obvious from Dr. Katz's comments and others' that there are some circumstances here, but I think the most important point is this is an enormous loss for FDA, and I can only hope from what I know of interacting with you all of these years that it's a good move for you and that it will be a wonderful benefit for NCI to get someone of your caliber, but this agency will sorely miss your work.

This is a small token.

(Applause.)

DR. ZOON: I don't want to hold up your break because I know how important breaks are, but I just want to say how much I appreciate your gift. It means a lot to me, and the recognition of both my colleagues around the table, the audience, for your recognition, and in my new job I will try to do my very best to make a significant impact on the public health at NCI.

Thank you.

(Applause.)

CHAIRMAN DAUM: Tax dollars, of course, were not used to fund that gift.

(Laughter.)

CHAIRMAN DAUM: It is 2:35 here in the central time zone -- Eastern time zone. I'm sorry. At exactly ten to three we will reconvene, and with the first question up, we will have question focused discussion.

Thank you.

(Whereupon, the foregoing matter went off the record at 2:40 p.m. and went back on the record at 2:54 p.m.)

CHAIRMAN DAUM: Could the final conversations sort of cease and we move back to our business of the day?

I'd first like to say a special thank you to Dr. Sachs for upgrading the quality of the Musak. I don't know if any of you have noticed, but we've been treated to her tapes and music at lunchtime and during breaks. I must say it's a lot better.

The second thing is as usually happens during these kinds of meetings, there are actually three people who have asked for some time before we turn to the question itself, and so we're going to hear three separate comments.

First, I'd like to call on Dr. Midthune of FDA to clarify an issue that came up last hour.

DR. MIDTHUNE: I just want to clarify that the sponsor, at that time Aviron, did discuss their plans for the efficacy evaluation with us, and that we were in agreement with their approach to evaluate efficacy for the pediatric population and effectiveness for the adult population.

CHAIRMAN DAUM: Thank you. That certainly helps shed some light on some things people were concerned about.

Dr. Parsonnet wanted to raise a global issue not focused on one question or another, and this is the time to do that.

DR. PARSONNET: I guess my point comes out really from the hat I used to wear, which is on the Anti-infective Advisory Committee of the FDA and where we always had comparators. We always looked at a new antimicrobial agent and compared it to one that was already in existence, and that, I guess, sort of brings up this issue of the elephant in the room here, which is that there already is a flu vaccine that exists, and we've seen very little data comparing the proposed vaccine to a currently existing product.

And that gets to issues both of safety, where we really haven't seen head-to-head comparisons of safety, head-to-head comparisons of recurrent use of the vaccine, annual revaccination, and really very little on comparative efficacy.

And I guess it just really raises a question for the FDA about how we consider a new vaccine in the setting when there is one that is already approved and also even a more broader question about how if it is approved, a clinician would then go about making a decision about the use of these two competing products.

And so I guess, I think partly because of my previous experience on another committee, I'm just a little bit unsure about how we put a new product in comparison with one that already exists.

CHAIRMAN DAUM: Does FDA or sponsor want to comment on Dr. Parsonnet's issue?

DR. MIDTHUNE: I mean, obviously it's always interesting to have comparative data, but the primary requirement is that you demonstrate safety and effectiveness, and in this case it has been done in comparison with the NAF control, and that certainly is acceptable to us.

You know, clearly, I think that some of the evolving recommendations moved ahead of the development of this vaccine, and for example, since the adult study was, you know, done, you know, the ACIP subsequently made recommendations to recommend influenza vaccine for individuals above 50 years of age.

I think had that recommendation been in place at the time, we clearly would have asked actually for a comparative study in that particular situation because, you know, clearly sort of a standard would have been to give the inactivated vaccine, but that's not where we were at that time.

And also at that time, for healthy children, there was no recommendation for administering influenza vaccine. So perhaps that puts a little bit of the history into a context.

CHAIRMAN DAUM: To this issue? Dixie and then Dr. Katz.

DR. SNIDER: Well, I'll speak, you know, with my ACIP hat on and say that I think Dr. Parsonnet's points are well taken, and they are issues that are going to be highly problematic, and I'm sure subject of vigorous discussion not only of the ACIP, but the AAP and the American College of Physicians and other professional societies that are going to have to, you know, weigh in on how they feel about the use of this particular product vis-a-vis the other available product and hopefully give some guidance to clinicians as to how to deal with the situation.

But they are very important points.

CHAIRMAN DAUM: Dr. Katz, this issue?

DR. KATZ: One, Julie, I don't think we ever demanded any studies of pneumococcal polysaccharide vaccine versus pneumococcal conjugate vaccine. The licensure of pneumococcal conjugate vaccine was on its own virtues and assets and not on comparison. That's a specific example.

A generic one is I think those of us who call ourselves vaccinologists are very interested in there being more attitude and aggression towards mucosal immunity, and I think the idea that mucosal immunization could in some ways supplant injectable vaccines is very appealing.

I have a slide that I wish I had brought with me which shows a 15 month old child being told that, well, there'll be one more injectable vaccine each year, and if you will forgive a nasty comment, this one year old is pictured going, "Not on your life."

(Laughter.)

DR. KATZ: So that I think the people who complain that we're giving kids 20 and 22 injections in the first two years of life or the first five years of life have a lot of interest in mucosal vaccines, and especially as we talk about a vaccine that, if you believe in it and you use it, is going to be administered annually. The pragmatics of it become a major issue which the pediatricians have been discussing because somebody said this morning "recommended." It isn't yet recommended. It's encouraged for children six to 24 months of age.

But if that is followed, as many people anticipate by recommendation and not just encouragement, aside from the idea of another injection, the logistics for physicians who take care of children are rather formidable if you have the window from September to November to give a vaccine.

And the question arises: does it have to be given in physicians' offices? Can it be given in day care centers?

There are a lot of other things that make us a lot more enthusiastic about the whole field of mucosal immunization rather than more injections.

CHAIRMAN DAUM: Thank you very much.

There's one more preliminary piece of business to address, and that is Dr. Mendelman asked me if he could show two slides to address the issue of Dr. Faggett, and that is minorities and others who might have been immunized with FluMist present or absent from the BLA database.

DR. MENDELMAN: Present. These are the demographic characteristics in the healthy working adult study. So most of the individuals were Caucasian, ten to 11 percent were black, and the median age was 38 across both groups.

In the 19 study, the Northern California Kaiser study, I don't know if we can move that up, but I guess you can see it. Ten, 11 percent Asian Pacific Islander, 20 percent versus 19 percent Hispanic, 55 percent Caucasian, and six percent African American, et cetera. This was Northern California. So it was primarily Oakland and the surrounding areas.

And the last one, and you have this in the briefing document from the FDA, I think the request was also to look at efficacy by race, and the efficacy, Caucasian or non-Caucasian, 94.9 percent for any influenza and 92 percent for Caucasians in the pivotal efficacy trial in children.

DR. FAGGETT: One follow-up question, if I may. Do you have any experience in terms of emergency room visits of the various populations as well? Probably not. Is that available?

DR. MENDELMAN: We have the emergency room visits from the Study 19, the Northern California Kaiser, because one of the three settings that was analyzed for medically attended events were all emergency department visits within the 42 days after vaccination, and hospitalizations and any clinic visit.

DR. FAGGETT: Thank you.

CHAIRMAN DAUM: Okay. That completes our sort of preliminary break generated housekeeping issues. What I'd like to do is just literally read the first question now and then ask the committee for discussion of things that we haven't addressed that they would like to address before we actually begin our voting process.

So the first question, as I understand it -- Dr. Mink, please correct me if I say anything wrong here -- has two parts, A and B. Part A is: are the data adequate -- it's a safety question, right? -- are the data adequate to support safety of FluMist for individuals in the three age groups that you see up there, five to 17 years, 18 to 49 years, and 50 to 64 years?

And to please consider the data relative to respiratory events, asthma, and URI, shedding and transmission of vaccine strains following receipt of FluMist, and annual revaccination.

And then Part B becomes relevant if you believe the data are not adequate for Part A, and that is: if the data are not adequate for specific age groups or if there are other safety concerns, please discuss what additional data should be requested.

DR. MINK: With the addition that it should say "healthy individuals."

CHAIRMAN DAUM: Yes. I even wrote that and I still forgot to say it. So it's normal people we're talking about in this question. So thank you.

Okay. So are there any other issues that we need to talk about to have an airing of the issues in this question? Dr. Stephens.

DR. STEPHENS: I'd like to have comments about the revaccination issue because this is going to be something presumably given every year, and yet at least from my understanding, the data for older children and adults really doesn't exist for efficacy, for example or in most instances safety for revaccination, and I'd just like to get a comment on the revaccination question because that's what we're going to be doing presumably with this vaccine.

CHAIRMAN DAUM: Let's get you a couple of comments. Let's hear from the sponsor, and let's hear from the agency as well.

DR. CONNOR: I think the primary revaccination data, as you point out, comes from pediatric trials. The data that we've provided for you shows both on the efficacy side, the second year efficacy of data, as well as on the safety side the data both from the AV006 trial and multiple revaccination years, and we also have data that looks at SAEs across those years and demonstrates that there isn't any difference and that generally things are lower in the reactogenicity cycle following multiple years.

It is a setting in pediatrics where there is more reactogenicity generally than in the older population. So we believe that that represents the opportunity to best look at the issues of revaccination, but there's not specific revaccination data in the older populations. I guess those are just the SAE data.

CHAIRMAN DAUM: Dr. Mink, do you or someone else in the agency want to comment on that issue from your perspective?

DR. MINK: On the slides that I showed, slide number 18 shows the total database. It shows that across all age groups there was 7,354 second dose experienced with FluMist. I believe it's page 5 on your handout, slide 18. Those are the revaccination data for second dose.

You can see that about 3,000 of those are in one to four years of age; 2,600 are from five to eight years of age; and just over 1,000 are in children from nine to 17 years of age. So those are the total number of subjects in the database with repeat vaccinations.

However, in the second dose for some of the kids, those will be the same dose in the same year. They're not all a second year or a second season.

For subjects in the AV006 trial, we do have the repeat vaccination data for safety and efficacy in year two and safety data from year three, which I presented, and then also for individuals over ten years of age there are some vaccinees in study AV012. I believe there was a total of 2,100 subjects in AV012 who received doses in two years. We have SAEs mostly from those individuals, and there was no increase in the repeat vaccinees.

CHAIRMAN DAUM: I trust that reviews the available data and addresses Dr. Stephens' question.

Other questions specifically focused on question one?

(No response.)

CHAIRMAN DAUM: Okay. So, Dr. Stephens, you are in the hot seat for the last time probably and perhaps you could initiate our discussion of question one.

Now, I should say as a procedural item that I have a voting sheet somewhere, right in front of me, and that we will record your vote separately for each of these age groups so that when you're finished speaking I'll know how you felt about five to 17 years, 18 to 49, et cetera, and then your comments regarding question B will be recorded as well.

Then once we complete this, we'll then put up question two and again look at discussion issues and we'll repeat the process till we're done.

Dr. Stephens.

DR. STEPHENS: A couple of kind of opening comments. One, I think there's been a lot of progress made since we first heard about this vaccine a year and a half or two years ago. I think there's pretty convincing data in terms of the new analyses that the vaccine does have a problem in kids under five in terms of potentially probably at a small rate increasing the incidence of asthma and croup. I think that's certainly borne out by the new analysis of the data.

In older individuals, I think that's probably not the case. There is evidence of viral shedding. There is evidence of transmission, but I am somewhat relieved by the data that's presented today in terms of certainly the transmission issue.

The issue of reassortment, I think, is still on the table, and I preface all of that by commenting on the safety categories.

I think the one kind of issue that we're facing is that this was a vaccine largely designed initially for children, young children, and now it's kind of being reassessed, if you will, and relooked at for older groups, and I think that the problem that we run into in some of these age categories is lack of specific data about the specific vaccine.

So from my perspective, and I'll start with what I think are the easy categories first, the 18 to 49 year old, I think there is data that is adequate to support the safety of FluMist in healthy individuals in that particular age group.

I think also that there is reasonable data in the five to 17 year old age group. My concern is actually that five to nine group where the issue of asthma may not be completely solved at least in my view, but the data would suggest that in all likelihood, and I would probably vote yes, that in five to 17 year old individuals that there is adequate data for safety in the individual, in the healthy individual.

I don't think thought that with an n of 500 or so, 511 I think it is, that there is good data in the 50 to 64 year old age group for safety, and I think that's largely an issue of numbers.

For individuals undergoing revaccination, we just heard that data represented. I think it's more of an issue of efficacy than it is safety. So I would probably vote yes in terms of safety for this category.

I think there continues to be though concerns about this attenuated virus in terms of reassortment issues, in terms of introducing this into a very large population, a very large amount of this vaccine being administered to the population, and we'll come to -- so that remains an issue in my mind.

And I think I'll stop there and turn it over to Dr. Katz.

CHAIRMAN DAUM: Before we leave you, I heard three yes votes for the different categories.

DR. STEPHENS: You heard yes, yes, no, and yes.

CHAIRMAN DAUM: I end up confused. I apologize. No, it's good we get this straight in the beginning and then it will go well.

DR. STEPHENS: For the age group -- there are four questions here as I read it -- five to 17 year olds.

CHAIRMAN DAUM: Correct. Eighteen to --

DR. STEPHENS: And I would vote yes.

Eighteen to 49, yes; 50 to 64, no; individuals undergoing annual revaccination, yes.

CHAIRMAN DAUM: No.

(Laughter.)

CHAIRMAN DAUM: What I think we've been asked to do is --

DR. STEPHENS: It looks like I'm working from an older version of this particular question.

CHAIRMAN DAUM: Yes, the newer version is what we were showing this morning.

DR. STEPHENS: Okay. Three age groups: yes, yes, no.

CHAIRMAN DAUM: Okay.

DR. SACHS: Everybody has a newer version in your packet, your folder, your blue folder. So it should be more than one page, and we'll all work from the same version.

CHAIRMAN DAUM: I think we're there now. Now I understand what happened. Okay. So it's good to straighten these things out early because then we streamline the process and it gets much easier.

So Dr. Katz.

DR. KATZ: Dr. Katz has an initial question, which is: who came up with these age divisions and on what basis?

I don't understand how we divided the life span from five to 64 into these three groups. I don't know of any other vaccine where we've ever studied in that particular way.

I tried to think. This is a live attenuated vaccine, and I tried to think of the other live attenuated vaccines we use: measles, mumps, rubella. No one has ever shown any difference in age groups with measles or mumps. With rubella, yes. Post menarcheal females are more apt to have arthralgia. That's the only thing I can think of.

OPV, when we used oral polio vaccine, which as my colleague has pointed out we don't use anymore fortunately, there was a suggestion in very early years that perhaps people over 18 were more likely to develop vaccine associated paralysis, though that was never borne out.

Yellow fever vaccine? Dr. Snider has brought us information at the last meetings on adverse events in adults, but I don't think many children get the vaccine. So the denominator doesn't give us any data on which to base. So I'm left with a basic question: why should I have to worry?

Vaccinia, which we're thinking very seriously of using again. The very young child in the first two years of life perhaps has a higher rate of adverse events, but you get above that and there's no difference among different age groups.

So I find it very difficult to get excited about differentiating. So I'd vote yes for all three.

DR. MINK: Dr. Daum, may I answer?

CHAIRMAN DAUM: But of course.

DR. MINK: The reason the 50 to 64 year age group has been divided out is because of the CDC acknowledgement of them being high priority because of the significant percentage of those --

DR. KATZ: Yeah, but there's nothing to suggest that they're at higher risk of adverse events.

DR. MINK: I'm just explaining to you why the questions are presented to you by these age groups.

DR. KATZ: Okay. I don't accept that as justification.

(Laughter.)

CHAIRMAN DAUM: I have a suspicion that we are not going to resolve this issue.

(Laughter.)

CHAIRMAN DAUM: Right here and right now, and so we're going to move on and hear from Dr. Edwards.

DR. EDWARDS: I think for many of the same reasons that David articulated, I think that the five to 17 years' safety data is quite extensive, as is 18 to 49, given the data of the effectiveness trial.

I'm less comfortable, however, with the data from 50 to 64 in terms of safety primarily because of the numbers and because that this has been a group that has been looked at and targeted as many of their members in higher risk groups.

So I would like to vote yes, yes, and no.

CHAIRMAN DAUM: And can you address Part B also? I think I mischarged the committee a little bit because you can answer Part B if there are other safety concerns, even if you felt the data were adequate.

So do you have any Part B issues?

And, Dr. Katz, if you did, you could pipe up as well.

DR. KATZ I think with Part B I would only add that there should be continuing studies of transmission.

CHAIRMAN DAUM: Thank you.

Dr. Edwards?

DR. EDWARDS: I think the transmission issues obviously need to be looked at, and I think as we're going to come back to in post marketing, there has to be attention to reactive airway disease and asthma in post marketing.

But I think that other than that currently we're safe.

CHAIRMAN DAUM: Thank you very much.

Dr. Snider.

DR. SNIDER: I think I'd generally agree with my colleagues who have spoken thus far. I think the problem of asthma, reactive airways disease is potentially a problem for those less than five, and it may be a problem of lower frequency in older age groups.

At the same time, as I mentioned earlier, unfortunately we don't have the data to know whether what might be precipitated by FluMist is less than what would occur with natural infection. We don't know if TIV, as far as I know, really protects against asthma reactive airways disease as well as FluMist might do it because we don't have a head-to-head comparison on that.

So I think the point is it's an open issue and one that needs to be studied in more detail.

As far as the specific questions about the age groups, I'm comfortable with the safety data for the five to 17 year age group and the 18 to 49 year age group. The 50 to 64 year age group, I think it's 511, and I would prefer to see a larger population.

I think with regard to additional concerns I don't have huge concerns about transmission. I don't have huge concerns about reassortment. There are these things we've talked about in terms of inadvertent administration to people in whom it would be contraindicated according to the current application and high risk people who might receive it because of transmission.

So I think additional studies there are indicated, but I don't have a high level of concern about it.

The revaccination issue has already been clearly outlined. I think the data there indicate a high level of safety in the younger age groups. I can't think of biologically plausible reasons why that would be a major problem in the older populations, but the fact is if we look at the data submitted under the BLA, we don't have data, extensive data, on revaccination. So that's additional data that would be nice to have somewhere along the way.

And that's all I have to say about question one.

CHAIRMAN DAUM: Dr. Snider, so I understand you, in the 50 to 64 you voted no?

DR. SNIDER: Yes, yes, no.

CHAIRMAN DAUM: Thank you very much.

Dr. Hamilton.

DR. HAMILTON: Yes, yes, no.

With respect to considering the additional data, I think that one merely has to look at the numbers in the study design and the confidence is related to that, but more of the studies were designed to look at respiratory events, and by necessity perhaps shedding and transmission received a lot less attention, as did annual revaccination. And perhaps additional data should be generated to reflect that.

CHAIRMAN DAUM: Thank you very kindly.

Dr. Eickhoff, the time has come.

DR. EICKHOFF: A year and a half ago I think I voted no on the safety issues, and I'm going to change my vote this year.

For the shedding and transmission data, sure, I think that should be studied. We need further studies of that. I'm less concerned about transmissions to the high risk host simply because wild type influenza by and large is not seen as a problem in immunocompromised hosts, including those who are organ transplant recipients and including those who have AIDS, for example.

We don't see serious disseminated disease in the same way we see serious disease if they are exposed to other live attenuated viruses like MMR or varicella or perhaps all too soon vaccinia. So I am less concerned about transmission to a high risk host. It's going to be less of a problem than it is with wild type influenza.

Annual revaccination? Yes, again, I would like to see more data about that as the years go by. There is some data with regard to annual revaccination of TIV, but again, not that much. So as cold adapted influenza virus comes along, I think that's a subject for further study.

Asthma has been identified as a problem in the one to five year olds. There may be a problem. I'm not convinced it is a problem, but the answer will come from comparable studies of wild type influenza studied in the same way as these individuals who have received CAI vaccine were studied.

It may not be intrinsic to this vaccine. It may be intrinsic to influenza viruses in general.

Given all of that, for safety in the first two age groups, I will say yes to both of them. The 50 to 64 year old age group, if I could think of any biologically plausible reason why they might behave different immunologically from those ten years younger, I might hesitate, but I can't think of such a reason. So I will vote yes.

CHAIRMAN DAUM: Thank you very much.

And we'll go on to Dr. Cox.

DR. COX: Yes, I would like to, first of all, say that both the sponsor and the FDA have made our lives so much easier. Although we're awash in data, we have been pointed in the direction, and it's just much easier than dealing with the information that we had last time.

So for the first three questions, I would say yes, yes, and yes.

With respect to the 50 to 64 year old age group, I agree with Ted and with Sam. I cannot think of a biologically plausible reason why the safety would differ in this age group.

I think that the additional studies that I would like to see done have to do, first of all, with developing tools that could be used by practitioners to screen out those with high risk conditions so that we can be sure that those with high risk conditions are receiving trivalent inactivated vaccine.

I think we do need continuing studies on transmission, reassortment, and genetic stability, and I would also like to see additional information generated on annual revaccination both in terms of safety and efficacy.

CHAIRMAN DAUM: Thank you so much.

Before we go on, I'd just like to remind everybody that there is some confusion, and the preliminary questions are not what we're using today. We're using the questions that are here. So there are really only three age groups to address.

And I presume from the global tone of your comments that you voted yes on all three, but please correct me if I'm wrong.

DR. COX: Yes, I voted yes on all three.

CHAIRMAN DAUM: Yes, but I think David Stephens got a little confused by that as well, but were using this version of the questions today. So committee members could try and remember that.

Dr. Gellin.

DR. GELLIN: I'll assume that was so I didn't screw this up as well.

Since the question is are the data adequate, I'll stick to that question, and for that reason will vote no on the older category, 50 to 64, but yes on the two younger categories.

I think, I mean, somewhat along the lines of Sam about the age stratifications. I actually would like to see subsequent data better represented in children nine and above and less than nine, particularly since the recommendation for those less than nine will be two doses. So I think that has some relevance to the annual revaccination issue.

CHAIRMAN DAUM: Thank you very much, Bruce.

We'll move on to Dr. Steinhoff, please.

DR. STEINHOFF: I'm going to vote yes on all three groups, but I want to raise an issue that's been raised before in a slightly different way, and I'm concerned about the issue of transmission to household members who might be at high risk, and I don't know quite how this should be dealt with, but one approach might be the obverse of the recommendations for the inactivated vaccine is that it should be given to healthy persons who are in a household with persons at high risk who would also get the vaccine.

You might want the observe of that, that this vaccine maybe should not be given to persons who have household exposure to subjects at high risk.

That's not too confusing, right? Who are also unvaccinated. What I'm trying to say is we need some more data on the actual transmission and risk in the likely common household exposures to people who get this vaccine.

CHAIRMAN DAUM: Mark, thank you.

We'll move on to Dr. Myers.

DR. MYERS: Well, I'll start off by saying I don't like the questions because I agree. I think the age stratification is not the way it should be. It should be eight and below and eight to 17 and 18 and above.

With that said, like Mark I have concern about the transmission within households when we have remarkably little data on people with underlying diseases, and so let me take 1(b) first.

I think there needs to be more safety data collected on the 50 to 64 year age group. I think there ought to be specifically data collected on safety in individuals with chronic diseases, particularly chronic lung disease.

I think there should be a comparative trial to trivalent inactivated vaccine, and the question I would like to have been asked but we weren't asked was do I think there's sufficient data on annual revaccination. My answer would have been no, that I don't think there is.

We're talking about this vaccine used year in and year out, and we just have no data except for two or three doses to children. So I wasn't asked that question, but I'll give you the answer anyway.

CHAIRMAN DAUM: Thank you very much.

DR. MYERS: So my voting would be I believe the data does support the safety of FluMist in healthy individuals five to 17, 18 to 49, and 50 to 64, and for the same reasons that Sam does, I can't think of any reason why 62 year olds are different than 49 year olds.

CHAIRMAN DAUM: Thank you very much, Dr. Myers.

Did you want to say something?

DR. MYERS: No.

CHAIRMAN DAUM: Okay. Moving right along, Dr. Overturf.

DR. OVERTURF: I would vote yes, yes, and yes, based upon the fact, again, I do not feel there's any biologically plausible reason to expect differences in the 50 to 64 year old age group.

I also think that for all the questions below regarding data for respiratory events and shedding and transmission of the vaccines and annual revaccination there's a critical need for a lot more data and will have to be a critical part of the post licensure evaluation of this vaccine.

I'm also a little bit concerned about what the demand for this vaccine might or might not be, and that it would be interesting to see in the future how the production of the vaccine which has to be done on an annual basis -- it seems rather complicated to me -- will be able to keep up with that demand.

Obviously part of that will be dependent a little bit upon how the ACIP and other professional groups decide how to recommend this vaccine or whether they choose to recommend it or let people decide for themselves whether they want to take this vaccine.

I think there will be, because of a lack of education, a lack of data, a critical need for serious education of individuals who choose to take this vaccine over another vaccine which already has a safety and efficacy profile defined for it. So I think that's going to be a critical role for professional bodies in the future to try to define this for potential vaccinees.

CHAIRMAN DAUM: Dr. Diaz

DR. DIAZ: I would vote yes, yes and no, and purely from taking the purist standpoint in answering the question is the data adequate to support safety. I think we're splitting hairs in a sense because it really comes down on that last question to whether one is willing to extrapolate from the data that's presented.

I think the sponsor did a good job of looking carefully at the data in the 50 to 64 year old age group that they had, and in my mind I don't think that there's any reason to suspect likewise that there would be any ill effects safety-wise in that age group.

But based upon the question and the small n, my vote would be yes, yes and no.

I do think there's more need for information on transmission and reassortment. In particular, I would very much like to see a movement toward capturing information on annual revaccination not only with this vaccine, but also with the inactivated flu vaccine because there isn't much data, and we're using it in younger and younger populations where the accrual rate over time in terms of revaccination events is going to be much higher than it was in the past.

So I think there is a need to get information on revaccination on an annual basis.

CHAIRMAN DAUM: Thank you very much.

Dr. Faggett, please.

DR. FAGGETT: Yes. Starting with B first, I really would like to see some other data relative to populations in other geographic locations, such as the TennCare population that Dr. Edwards mentioned earlier. I think that there are a lot of differences not just with ethnicity, but socioeconomic status that you probably won't capture in a closed system, health care provider population.

Saying all of that, I do vote yes, that the data is barely adequate for five to 17, 18 to 49 and 50 to 64. So yes in all three.

We do need more studies in terms of shedding and transmission of the vaccine strains. I think the jury is still out in terms of annual revaccination.

CHAIRMAN DAUM: Dr. Faggett, thank you very much.

We'll go on with Dr. Markovitz, please.

DR. MARKOVITZ: Yes. For the reasons outlined already by Drs. Katz, Eickhoff and Cox, I'd vote yes, yes, and yes. In 1(b) for data that we need, clearly besides whatever else we need for these age groups, and I do believe it's very important to vaccinate healthy people in these age groups, we need a lot more data on older people and people with underlying problems who, of course, are the absolute top priority for flu from a public health point of view. So that's what I'd like to see more of.

CHAIRMAN DAUM: Thank you so much.

Dr. Parsonnet.

DR. PARSONNET: I'm also going to vote yes, yes, and yes. I think they looked at the study to look at 18 to 64 and showed that it was safe from 18 to 64 and doing a post hoc subgroup analysis of 50 to 64, they just need to really show that it's comparable in those age groups, and otherwise they could take out each individual year. Fifty-one, is that safe? Fifty-two, is that safe?

I mean, you can't even looking afterwards for these post hoc analysis, I think it should be -- unless we have some reason to think otherwise, it should be considered safe in those groups.

I'd like to see more comparative data with the currently available vaccine in terms of safety because I think it will help in making decisions for various groups in using these vaccines in the future. I'd like to see more data on smokers and safety in smokers.

And I don't think biologically it's likely that there are really going to be safety issues with revaccination, but it would be nice to see more revaccination data and more data in the elderly population.

CHAIRMAN DAUM: Thank you very kindly.

Ms. Fisher.

MS. FISHER: No, no, and no.

The data are inadequate to support the safety of FluMist for individuals five to 64 years of age. The increased risk of asthma in young children and the increased risk for some children in these studies for upper respiratory infections, musculoskeletal pain, otitis media and croup, as well as upper respiratory symptoms in adults suggest that an unknown number of health, but perhaps genetically vulnerable individuals across all age groups will not be able to handle this vaccine well, and this will over the long term also lead to the public perception that when you get the flu vaccine you get the flu.

And this is an important consideration long term because when you make healthy people sick after they get a vaccination, whether it's with live virus polio vaccine or live virus flu vaccine, when you have inactivated vaccines that do not cause disease symptoms, you're going to pay a price in terms of the public perception of the risks associated with vaccination.

You were able to successfully make the argument to prevent polio, but as I said, flu is not polio, and because most healthy children and adults are not permanently injured or die from the flu, I think careful thought needs to be given to this issue.

The fact that live vaccine flu virus is shed in 80 percent of recipients poses an additional risk for our population at large, particularly for immune compromised individuals across all age groups.

The outstanding questions about the true rate of transmission of vaccine strain viruses among children needs to be clarified, as does the retention of the attenuation of the shed viruses and the high frequency of nucleotide changes. Because this live virus nasal vaccine is not indicated for high risk health groups, which have historically been the targeted population to receive the flu vaccine, it's a very serious step to move to use of a live virus vaccine for the majority of healthy individuals, and a standard for proof of safety must be very high.

I don't think that standard has yet been met by the data which have been presented so far. I'd like to see a trial of a genetically diverse group of American children and adults which addresses safety and efficacy of simultaneous vaccination with FluMist and other vaccines, revaccination, vaccine shedding, and the rate of household transmission to the unvaccinated individuals, as well as genetic stability.

CHAIRMAN DAUM: Thank you, Ms. Fisher.

We'll move on to Dr. Goldberg.

DR. GOLDBERG: Okay. I guess yes for five to 17 years, recognizing that I think that in the younger age groups here it's not clear to me that there may not still be some effects that are continuing on from what we saw in the younger children. I do believe we made the right recommendations the last time, that there were problems with the younger children that have been borne out.

Eighteen to 49, yes.

Now, I don't believe in post hoc subgroup analysis, but that said, I do think that there really are inadequate data for 50 to 64 if we're going to split the hair and label it in that way. So my answer would be no.

I think we need more information, more trials on shedding and transmission and reassortment, as well as annual revaccination studies.

CHAIRMAN DAUM: Dr. Goldberg, thank you.

I guess it's my turn. I'd like to compliment the company and the agency team working together on the progress that's been made sine the last time we've heard about this vaccine. Many of the anxieties and concerns, not all, have been addressed adequately, in my opinion.

I think that the decision to move the requested indication to age five, an age when we currently don't actually immunize healthy children against influenza, was a bit of a master stroke in terms of corporate strategy because a lot of the issues in children under five were swept off the table.

I believe that the answer should be yes on the safety data across the board. Having said that, I would like to see more data generated by the company's sponsorship working together with the agency's guidance.

The shedding issue is an important one to me, and I think we don't know enough about it, and we don't know enough about the impact on the people on whom the virus is shed.

Having said that, I'm persuaded that this is a relatively low frequency event, a relatively low inoculum event, and is, after all, an attenuated virus.

The concurrent vaccine issue, of course went by the boards by and large with the moving the lower age to five, but there are adults who will get simultaneous vaccines, particular vaccines against pneumococcal disease, and so there are some issues there that I would like to see explored, but I don't think that's an issue for holding this up at this point.

The annual revaccination issue that several have addressed is obviously a very important one and needs additional information. I'm always persuaded by Dr. Faggett's argument that there aren't enough minorities. We saw especially in populations where people were likely to come back and likely to comply, and I think that the real world contains large segments of people who aren't, period.

And we need to make sure that the vaccine performs in those settings as well, and I'd like to at least have some safety data generated in the future about those.

I'm also intrigued by and persuaded by Dr. Cox's suggestion that we know more about the molecular documentation of the transmission and genetic stability of this virus, but I think we saw enough data that I feel confident that this won't be a deal breaker, so to speak.

So I'm willing to vote yes on all three of these issues, and with those 1(b) issues that I raised.

And that brings question one to a close.

Say it again, please.

DR. FREAS: Industry's position on record.

CHAIRMAN DAUM: Industry's position for the record, but of course.

Dr. Decker, would you give us industry's position, please?

(Laughter.)

CHAIRMAN DAUM: As best you can.

DR. DECKER: No, I can't do that, and let me take this opportunity to clarify. I'm the industry representative, but of course, I'm Michael Decker. I'm not some distillation --

CHAIRMAN DAUM: We know that.

DR. DECKER: -- of industry. All right. So let's be clear on that.

Also, because it did come up, let me comment that you can't have an industry rep. who represents the industry who is not involved in industry. So it is inevitable that as long as there's an industry rep. there will be discussion here of products that that representative has something to do with.

In this case, I think everybody knows that I work for Aventis Pasteur, which is the world's largest manufacturer of inactivated influenza vaccine.

This arises all the time at these meetings. So I also hope everybody knows that when I make comments, at least I think they're solidly grounded in the scientific issue, and they're not simply trying to stroke either Vanderbilt, one hat I wear, or Aventis Pasteur, the other hat I wear.

So with that preface out of the way, my comments have all been articulately spoken by the members of the committee. I really truly have nothing to add to what's been said here.

CHAIRMAN DAUM: Most unusual in my experience.

(Laughter.)

DR. DECKER: It's a going away present for you, Bob.

CHAIRMAN DAUM: Michael, we are mindful of the difficulties that you find yourself in, speaking for, quote, all of industry and at the same time obviously a member of one specific company, and we appreciate your candor with respect to that.

I know Dixie wants to make a comment, but before I call on him, I'd like to announce the results of the vote on question one. The question concerns the adequacy of safety data for individuals in three age categories.

For the first category, age five to 17, the committee voted 17 to one that the data are adequate.

For the second category, 18 to 49 years, the committee voted 17 to one that the data are adequate.

For the third category, 50 to 64 years of age, the committee voted ten to eight that the data are adequate.

That is the vote on question one, and before we move on, Dixie, let's hear your comment, please.

DR. SNIDER: Well, my comment had to do with the 50 to 64 year age group, and what I wanted to make clear is that from my perspective if we're talking about this question in a purist sense, in a sort of abstract way, in other words, healthy people 50 to 64 years of age, I would have no problem voting yes.

The difficulty I see, being a pragmatic person, is that when people were talking about biologic plausibility, people 50 to 64 years of age begin to get in significant proportion a number of chronic health problems, and the question becomes: can you effectively screen those people out?

And I would submit that you don't always successfully do that. And so from a pragmatic standpoint, I'm comfortable with those numbers, realizing the pragmatic difficulty of trying to actually identify a true healthy population.

CHAIRMAN DAUM: Thank you very much, Dixie.

I'd like to ask that question two -- oh, look at that. Okay -- be put on the screen and begin by asking committee members if there are issues that they feel like haven't been discussed today that they need clarity on before we start the voting process. Question two, are the data adequate to support efficacy of FluMist in individuals in the same age groups as we saw before, Dr. Katz demurring perhaps and others?

And then if the data are not adequate, what additional data should be requested? So are there discussion issues unaddressed?

Dr. Edwards.

DR. EDWARDS: I guess that goes to say to support efficacy/effectiveness, right?

CHAIRMAN DAUM: FDA, comment, please.

DR. MIDTHUNE: Yes.

DR. EDWARDS: Yes.

CHAIRMAN DAUM: Dr. Midthune says yes.

And also we add the word "normal" or "healthy" to the question again, correct? Again, yes.

Okay. So there's two clarifications there. All right. David, let's go.

DR. STEPHENS: Thank you, Bob. I appreciate it.

CHAIRMAN DAUM: If I ever come back as a guest, you know where I do not want to sit.

DR. STEPHENS: You know where you're going to sit though.

(Laughter.)

DR. STEPHENS: My votes are yes, yes, and no on these three age groups. I certainly think based on the effectiveness data that the 18 to 49 year old group, there is evidence of effectiveness, and I interpret that to be efficacy.

I'm bothered by that to some degree, but I'm willing to accept it for that age group.

I'm a little more concerned about the younger age group. We really don't have a lot of data on the, say, ten to 17 group. In fact, there's very little data, but I'm prepared to, looking at the efficacy data across the board in the older studies to accept the fact that for this age group that there is reasonable data to suggest or indicate that there would be efficacy of FluMist in healthy individuals.

I am concerned about this issue which has been raised about whether we would use FluMist or whether we would use an inactivated vaccine in this particular cohort of individuals, and we have no head-to-head comparison, and I think that, again, the data on 511 individuals regarding efficacy -- and I appreciate the arbitrariness of breaking out this age group, but that's what we've been asked to do, and that's what the ACIP has now done in terms of the recommendations for inactivated influenza in this particular group.

So I appreciate that, as he's saying, without any doubt. But I'm --

(Laughter.)

DR. STEPHENS: I remain concerned that this is a group that we do not have adequate efficacy data, and rather than give FluMist versus the inactivated, I think we need a study to address that particular issue.

CHAIRMAN DAUM: Thank you very much.

Dr. Katz, please.

DR. KATZ: I may surprise Dr. Stephens by somewhat agreeing with him because now I turn the question not to safety, but to efficacy, and there certainly is suggestion that with advancing age there is loss of immunologic responsiveness to various antigens.

I'm very comfortable in the younger groups beginning at age five. A lot of data on five to six years of age is all we know about the immune system. It's mature at five years of age, and I think we can, to my way of thinking, extrapolate from that on through healthy adult life.

When you get up into -- and again, it has been an arbitrary decision which I had no vote on -- but when you get up into the 60s, you begin to get people who don't respond to vaccines. Again, we're talking about a live vaccine, not an inactivated.

Who doesn't respond to Hepatitis B vaccine? Inactivated admittedly, not live, but it's the older age group. As you get older, as you smoke, as you're fat, you're less likely to respond to inactivated antigens.

I don't know about this. So that I guess I would vote yes, yes, no, requesting more data on immunogenicity. I'd settle -- you should forgive me -- for antibody data.

(Laughter.)

DR. STEPHENS: Which may be easier to collect than efficacy data.

CHAIRMAN DAUM: Thank you very kindly.

Dr. Edwards.

DR. EDWARDS: I have similar conclusions. I think that the revaccination issue is more important for efficacy effectiveness because I think we do know that the more antibody that you have either locally or humorally, the less you're going to respond to the cold adapted vaccine. So I think that really does need to be looked at, the revaccination question in terms of efficacy and effectiveness.

So I would vote yes, yes, and no in terms of the efficacy/effectiveness.

CHAIRMAN DAUM: Moving right along, Dr. Snider.

DR. SNIDER: I would agree and vote yes, yes, and no, and point out that if you're going to do that study, you can also get some additional safety data in 50 to 64 year olds.

CHAIRMAN DAUM: Could you clarify what study you mean, for the record, Dixie, please? The question contains a hook at the end of it. What additional data are requested? And you said if you're going to do that study. So what --

DR. SNIDER: If you're going to study efficacy of the vaccine against placebo and/or with the inactivated vaccine, it gives you an opportunity to look at additional safety data in that large group.

CHAIRMAN DAUM: Thank you.

Dr. Hamilton.

DR. HAMILTON: Yes, yes, no, and I agree to study older people.

CHAIRMAN DAUM: Dr. Eickhoff?

DR. EICKHOFF: Yes, yes, and yes. Yes on the third age category simply because I think the distinction between those two adult categories is artificial and not biological.

Just one other comment. I appreciate Dr. Nichol's justification for the effectiveness study. By and large in pre-licensure studies, however, I vastly prefer efficacy studies.

CHAIRMAN DAUM: Thank you. Before I call on Dr. Cox I'm going to call on Dr. Overturf because we're beginning to run into airplane schedule problems and ask him to weigh in next.

DR. OVERTURF: I would vote yes, yes, and no, based upon I think there are some biological differences in the older adult, and I think there's precedent in other studies. So I think there needs to be more data.

I think the data could be more easily obtained by efficacy rather than effectiveness data and could be done in a smaller group probably. So although I think either study would be useful in that group, certainly I think true efficacy would be a better study and easier to obtain perhaps.

CHAIRMAN DAUM: Thank you, Dr. Overturf.

We're going to return now to Dr. Cox. Sorry and thank you.

DR. COX: Sure. I would vote yes, yes, and yes. I think that while there are immunologic differences in older age groups, I think that they apply both to an activated vaccine and to the live attenuated vaccine.

And I would suggest that annual revaccination studies are particularly important with respect to efficacy.

CHAIRMAN DAUM: Excellent. Thank you.

Dr. Gellin.

DR. GELLIN: I seem to be hung up on the E words here of efficacy, effectiveness, and extrapolation, and while I was tempted, you know, the changing of the question to make it, slash, effectiveness, I think it's actually a pretty significant inclusion in the question, particularly for a new product, and it strikes me that this is a precedent setting inclusion in that question.

Nevertheless, my vote would be yes, yes, and no because I know the effectiveness data is interesting. I'd like to see a formal efficacy study of all the population.

CHAIRMAN DAUM: Thank you, Bruce.

Dr. Steinhoff, please.

DR. STEINHOFF: I would mark this yes, yes, and no.

CHAIRMAN DAUM: And therefore, we ask for Part B comments.

DR. STEINHOFF: Right. Additional data on the efficacy in the older group.

CHAIRMAN DAUM: As opposed to effectiveness. Thank you.

Dr. Myers.

DR. MYERS: I'm going to vote no, yes, and no. I think the efficacy data is for five and six year olds. There is no data for seven or eight to 17 year olds. I don't think you could extrapolate from a 49 year old effectiveness data into the eight year old population or to children who are getting two doses of vaccine, and so I don't believe that effectiveness or efficacy has been demonstrated for that age group.

Yes on the 18 to 49. I think Dr. Nichol made a good case for the effectiveness study.

No, I think there were several other people who have already made the point that I think that the 50 to 64 year age group could be less immunologically responsive.

So I think we need studies of the eight year old to the 17 year old for at least effectiveness or at least seroconversion, and I think there needs to be studies of trivalent vaccine versus FluMist versus placebo in the 50 to 64 year age group.

CHAIRMAN DAUM: Thank you very much, Marty.

Dr. Diaz, please.

DR. DIAZ: Yes, yes, and no, for the same reasons that Dr. Katz raised, especially in an older age group where the efficacy of the product, it would be nice to have some comparison data, a little bit more data in that age group in comparison to the efficacy in that age group for the inactivated vaccine.

CHAIRMAN DAUM: Dr. Faggett, please.

DR. FAGGETT: Yeah, this is becoming the cat's corner. I vote yes for the five to 17; yes for the 18 to 49; and no for the 50 to 64. I do agree we need more studies in the 50 to 64 age group and more comparative studies, as well.

CHAIRMAN DAUM: Dr. Markovitz.

DR. MARKOVITZ: Yeah. On the easy ones I'd like to vote yes for five to 17, yes for 18 to 49. For 50 to 64, while I believe more data would be helpful, and especially a direct comparison, again, in this age group and people who are yet older with the inactivated vaccine, I think if one had to extrapolate, there is, again, I don't think any biological reason why a 49 year old and a 55 year old are that different.

I also think that if anything, an older person is more likely to respond to the live vaccine than they would to an inactivated vaccine based on sort of general immunologic principles.

And then lastly, there's certainly old data that we haven't seen as a committee, but there are old papers showing the efficacy of previous iterations of this vaccine that's quite efficacious in yet considerably older people than 64.

So I believe that it's quite likely to be efficacious in the 50 to 64 age group, and while I'd like to see more studies, I vote yes in that age group also.

CHAIRMAN DAUM: Thank you, Dr. Markovitz.

Dr. Parsonnet, please.

DR. PARSONNET: I agree with Dr. Myers. No, yes, no. I think if you're thinking about potentially giving this vaccine to millions of children, I'd like to see some data in that age group, and there's really no data in children between the ages of six and 17, and so I'd like to see some data about efficacy in that group.

In terms of the subgroup analysis from 50 to 64, again, I'm not a big fan of subgroup analysis, but when you have a subgroup analysis that actually kind of pushes you in the opposite direction of the main group analysis, you have to take it somewhat seriously, and so I say no to 50 to 64.

CHAIRMAN DAUM: Thank you very much.

Ms. Fisher.

MS. FISHER: I think the data support efficacy for children ages 60 to 72 months, but are inadequate to demonstrate efficacy for healthy children and adults older than 72 months. I think another trial including healthy subjects in all age groups should be held, and ideally it should compare the efficacy of the live virus vaccine to the inactivated vaccine, including what happens after revaccination.

CHAIRMAN DAUM: Thank you, but we will need your vote on the question. So I'm going to --

MS. FISHER: No, no, no.

CHAIRMAN DAUM: Thank you.

Okay. Dr. Goldberg.

DR. GOLDBERG: Okay. No for the five to 17 year olds based on the need for that extensive extrapolation beyond the 72 month data.

For the adult trial, I'd just like to make a comment. The primary endpoint that was specified was any febrile illness. In neither the entire cohort or the 18 to 49 or the 50 to 64 met that standard compared to placebo as being significantly better than placebo.

That said, all of the supported endpoints do hang together and do support effectiveness in the 18 to 49 year olds, but I don't believe they do in 50 to 64 year olds. So it's yes and no.

And I think we need a study comparing FluMist to the inactivated vaccine in the 50 to 64 year olds, and we need to think very carefully about what the endpoint does need to be. I certainly, based on these data, would not recommend that we do another study on any febrile event.

So that said, and then in the younger children you need to do an efficacy study.

CHAIRMAN DAUM: So to make sure I have it right, you're no, yes, and no.

DR. GOLDBERG: It's no, yes, no.

CHAIRMAN DAUM: Okay. So I guess I'm the last one here, and I'm concerned about a number of things. One is that there aren't any real world data with H1N1 viruses in any population.

Secondly, that we don't have good efficacy data in the six to 17 month old age group. I think we do have good efficacy data for the 18 to 49 year old group, and I can solidly vote yes on that part.

So I come to I think it's -- I don't remember whose issue it was -- but whether the -- Dr. Katz's issue -- whether the younger children, the five to 17 year old ones, can be extrapolated knowing that there's efficacy in even younger children and in older people.

And I think that the answer to that is yes, that they can be, and so I'm going to vote yes on that, although I would also prefer to have had more direct efficacy data in that age group, and perhaps one way to reaffirm that my vote is correct would be to get some antibody data in those children.

Then comes the 50 to 64 year old age group, and here I really am a little more nervous about assuming efficacy. On the other hand, we've been asked about effectiveness as well, and I think effectiveness was demonstrated.

So that I'm going to vote yes for that older group, but it's a back door yes, reasoning leaning on the effectiveness issue rather than efficacy.

So I'm going to end up with yes on all three, but I would really like to have more data, particularly in the elderly -- 50 to 64 is elderly. I hope I'm not offending anybody -- age group, and I think that antibody data would be very helpful, and perhaps it could be constructed to bridge to the inactivated influenza vaccine.

So that concludes, I think, the committee's vote, and let me just rub shoulders with Jody here for a second.

(Pause in proceedings.)

CHAIRMAN DAUM: Michael, I apologize. Let us hear from our industry representative, Dr. Decker.

DR. DECKER: In this case, the industry representative does have a specific comment, which is although I was overseas at the time of the last meeting on this, if I recall my briefing materials and reports correctly, there was a vote on efficacy of this product at the last committee meeting, and it prevailed for all of the age groups, for the 50 to 64 age group, among others. Am I right?

Could I ask FDA to comment? Was the issue of 50 to 64, was that age group included in a prior vote of this committee on this product?

DR. MINK: That was not. The vote from the previous VRBPAC was across the ages.

DR. DECKER: All ages.

DR. MINK: For adults.

DR. DECKER: Right.

DR. MINK: It was 18 to 64.

DR. DECKER: Right, and 50 to 64 is a subset of that.

DR. MINK: And the 50 to 64 data was presented briefly by Dr. Nichol at that time, but the subset comparison had not yet been submitted to the agency. So this is the first time that these data, though it's a post hoc analysis, it's the first time these data have been presented to you.

DR. DECKER: Okay. So it's a murky issue. The point that I'm trying to raise though is that I think one thing. There's a couple of unfortunate things in the data in the presentation here. One, of course, is that in retrospect I think many people wish it was an efficacy trial and not an effectiveness trial, although effectiveness ought to be adequate.

The second thing is that I think the sponsor suffered a little from the effectiveness trial in that I suspect because of the ancillary measures all being consistently in favor of effectiveness, there's probably not biological meaning to the absence of demonstration of reduction in illness, but yet they're saddled with that outcome, which is unfortunate.

But then the third thing is procedural. In essence, they thought already home free on efficacy in 18 to 64 and here it is addressed again and they're shot down on 50 to 64. I think having been at this committee for a couple of years, my experience is that sometimes we don't always honor our prior decisions.

I don't know if this exactly fits that, but I thought it was worth mentioning.

DR. MYERS: The question that was framed at the --

CHAIRMAN DAUM: Are you going to -- tell me what you're going to do.

DR. MYERS: I was going to just comment.

CHAIRMAN DAUM: Please go ahead. I'd like to announce the vote as soon as you -- go ahead and make your comment.

DR. MYERS: I was just going to say you weren't at the last meeting, but the question as formulated did not include the breakdown of the subgroups. It was 16 to 64, not broken down the way it is.

CHAIRMAN DAUM: Okay. Thank you, Dr. Myers.

The committee had voted, and the issue of efficacy for ages five to 17 years, 14 members in favor, four opposed.

For the issue of efficacy in 18 to 49 year old folks, 17 members in favor, one opposed.

For the issue of efficacy in 50 to 64 year old folks, four members in favor, 14 opposed.

And that concludes our deliberation on question two. I'd like to move on now to discussion point three. I suspect that the agency has already heard many of the issues that we would raise vis-a-vis question three, discussion point three. This is not a voting question. So we can have some free discussion about this if people wish or we could just ask people to comment on this question.

If anyone wishes to open the discussion. Dr. Katz.

DR. KATZ: I don't want to open the discussion. I want to ask a question of the FDA, and that is: what is done each year with the inactivated vaccine when new strains are incorporated into the vaccine for that year? Are there human trials done? And if so, what's their magnitude?

DR. LEVANDOWSKI: No, there are no human trials that are done for inactivated vaccine each year, but the purpose for this study is somewhat different in that it's looking at a safety parameter of a new live virus.

DR. KATZ: I wasn't asking for justification, Roland. I was just asking for the record.

So for the record, there are no trials done with the new vaccines.

CHAIRMAN DAUM: Okay. Dr. Edwards and Dr. Decker.

DR. DECKER: Wait a minute. Can I clarify that? Because that's not correct. There are no trials required. They are done. There are trials done every year.

The FDA is not one of the licensing bodies that requires those results. Other licensing bodies elsewhere in the world do require them, and the trials are done.

CHAIRMAN DAUM: Can in about 30 seconds you tell us a little about those trials?

DR. DECKER: Virtually identical to what's proposed for this.

CHAIRMAN DAUM: Thank you.

Dr. Edwards.

DR. EDWARDS: I think one of the interesting features about the cold adapted vaccine is that in certain years the H1N1 may look more immunogenic than in other years, and the same with the H3N2. So that I think that at least in our trial, which is not FluMist, but is from the same mother or father, and preferably mother, that you would notice that there may be some need to look at the intrinsic immunogenicity of each strain because there is some difference.

And so I wonder if that should be something that would be asked more routinely than with the inactivated because with the inactivated the immune response is pretty much comparable each year, in that same ball park, but that may not be the case.

I mean, certainly with Dr. Belshe's studies, the H1N1 strain was less immunogenic, and in our studies the H1N1 was more immunogenic. So there may be intrinsic differences when you give a different code to the same virus.

CHAIRMAN DAUM: Thank you.

Other comments about discussion point three? Dr. Stephens, would you care to offer any comments about this?

DR. STEPHENS: Well, I'm in agreement that the studies as proposed should be done. I think that without question they should be done given some of the safety concerns that were mentioned.

It would also be nice, going to Dr. Katz's repeated comments, to obtain some additional immunological data on the new reformulated vaccine each year as a correlate, obviously appreciating that the surrogate of protection isn't fully, but I think more immunological data and more testing in animal models is also something that I would suggest.

CHAIRMAN DAUM: Thank you.

Dr. Snider?

DR. SNIDER: I had a similar question in looking at the annual clinical release testing that was the data we were shown and the indication that it demonstrated the feasibility of annual testing.

The question of, you know, what endpoints might be useful that could be gathered really rather quickly that would be of utility to the manufacturer, FDA, and all of us involved raises the question of whether we could get some serological data, and so I guess I would ask Nancy or Roland or someone who is more familiar with this if that would be a problem or something that would be doable.

CHAIRMAN DAUM: Dr. Cox or Dr. Levandowski, do you want to comment on Dr. Snider's idea?

DR. COX: Well, I think it's a good idea, and it certainly is doable. So that's something that's desired.

DR. SNIDER: I would concur and recommend it, and since it seems feasible.

CHAIRMAN DAUM: Roland?

DR. LEVANDOWSKI: Well, I actually was having another conversation while you were making your comments. So I'm not quite sure I got it all, but if the question was, you know, why not collect information on the immunogenicity of the vaccine at the same time, of course, that could be done. It just adds another parameter of difficulty in getting things done in a fairly quick period of time.

I think everybody is aware that for influenza vaccines to be useful they have to be available, and part of the point of doing this study, as I mentioned, is really looking at safety parameters for a new strain that might be incorporated into the vaccine.

That by itself takes some period of time to do just that clinical study, and although the information could be available at some point, it's likely that it would not be available at any time that you could use it for anything meaningful related to production of the new vaccine or what's going to happen with it.

I think it's a very similar situation that we find ourselves in with making inactivated vaccines where by the time they're available to do a clinical study, they really pretty much have to be used.

So I think there are some logistical issues that would, although the information could be useful, ultimately would probably not be useful in a fashion that you would have it before the vaccine could be made.

CHAIRMAN DAUM: Dr. Diaz, this issue?

DR. DIAZ: Actually it's somewhat related to that. I was curious to the manufacturer's comments about the annual time frame. Every year we sit here, and we march through in getting to the next season's flu vaccine under a fairly tight time frame based upon making an inactivated influenza vaccine. I'm wondering how that time frame compares to what you have to go through or would have to go through with an annual cold adapted flu strain.

CHAIRMAN DAUM: Manufacturer like to speak to that? I think we've heard Dr. Levandowski on this point, but of course, he's welcome to say it again if he wishes.

DR. YOUNG: Yeah, I think if I understand your question correctly you're wondering in terms of when VRBPAC actually selects the strains for the upcoming season's vaccine do we actually have enough time to make it --

DR. DIAZ: Right.

DR. YOUNG: -- and go through the process of --

DR. DIAZ: Exactly. I mean, Dr. Levandowski was addressing doing safety studies, but I'm just looking at the pure manufacturing aspects.

DR. YOUNG: Yes. I think, frankly, the issues are similar with either the inactivated vaccine or the cold adapted vaccine. We actually are already making cold adapted strains for the Brisbane, for instance, that has been identified in the southern hemisphere as a potential candidate for next season's vaccine. We try to stay ahead and obviously monitor the discussions of all the agencies around the world who are monitoring flu variability around the globe.

But in terms of once the actual strain is selected, if we haven't as yet started making that new master virus strain, it takes about four weeks to make that strain, and frankly, I think the inactivated vaccine manufacturers have to get a PR8 recombinant that grows in eggs well from the agencies. They make a reassortant for that wild type virus as well.

So I don't think that the timing is really that much different. The actual amount of manufacturing, the amount of eggs used, for instance, is quite a bit different for the inactivated vaccine compared to the cold adapted influenza vaccine because we actually rely on the nose to make a lot of the vaccine for us. So we put far less virus into the nose than you actually do when you inject it into the arm.

So the level of eggs that we need to make the same number of doses is probably ten to 100 times less.

CHAIRMAN DAUM: Thank you.

Dr. Hamilton, any comments on discussion point three?

DR. HAMILTON: It's already been stated.

CHAIRMAN DAUM: Good.

DR. HAMILTON: We're interested in efficacy data.

CHAIRMAN DAUM: Thank you very much.

Dr. Eickhoff.

DR. EICKHOFF: A comment. Only, again, the efficacy data, and that could be purely a subset of those 300 adults. Twenty-five or 30 individuals probably would suffice.

That's all I wish to say.

CHAIRMAN DAUM: Dr. Cox.

DR. MINK: Discussion point number three is about the committee's input for the design endpoints of the clinical study for release of new strains.

CHAIRMAN DAUM: Right. so it's not building on what you told us about questions one and two. It's anticipating, I guess, a world where this vaccine were on the market and being revised every year with new strains, and then what studies would you like to have on those new strains each year.

DR. MINK: Right.

CHAIRMAN DAUM: Do I have it right?

DR. MINK: And the study that was performed was a safety trial in adults using you saw about 330 or 300 or so adults. So the next discussion point is about post marketing studies.

CHAIRMAN DAUM: Thank you.

Is that clear or are people still mystified by that?

DR. SNIDER: I'm still mystified.

CHAIRMAN DAUM: Mystified. Okay. Check in here, Dixie.

DR. SNIDER: Sam brought it up earlier, but I mean, I seem to hear something different from Nancy and Roland about the feasibility of doing some serological studies, and I seem to recall a few years ago we had a problem around some elderly patients with flu vaccine that didn't give us serological responses that we had hoped for in a nursing home, and so I'm confused at this point about the role of serological testing in terms of annual evaluation, as well as even its potential role in helping us sort out whether the 50 to 64 year olds are going to respond as well as younger adults.

So if someone could clarify for me a little bit more about serologies, I understand about serologies in general not necessarily having a surrogate marker, but I also seem to hear that there's some utility to it, especially in terms of relative responses in different age groups or from year to year.

So if I could get some clarification I'd appreciate it.

CHAIRMAN DAUM: Dr. Levandowski and Dr. Decker wanted to comment on this as well.

DR. LEVANDOWSKI: Okay. Well, I'm not sure I'm going to clarify much, but I just wanted to comment on a couple of things. The study or the experience that Dr. Snider was referring to with the vaccine where there were low responses, I think you're referring to the vaccine that was the vaccine that was recalled in 1996, and the concern there was really that the vaccine itself was so potent. So there was an attempt made to identify whether there was a good antibody response in the recipients of that vaccine or not and whether a recommendation should be made to revaccinate.

That is a little bit different situation from, I think, what we're dealing with generally from year to year in terms of immunologic studies. We don't have a requirement for doing a clinical trial in the United States for inactivated vaccine, but we're glad to have studies done so that we can obtain sera from people who are being immunized with the most recent, current vaccines that aren't available until the fall each year when we can get those materials from a clinical trial.

We generally use that information to help us in trying to predict whether the current vaccines will produce antibody responses that will cross-react with the newer strains that are out there. We're not really looking to compare the vaccines one to another, nor are we looking to compare what happens with the current vaccine with what happened in terms of immunogenicity with the previous year's vaccine because the trials, first of all, can't be probably big enough to do what we would need to do, and the funding certainly isn't there to do really massive kinds of studies.

But what we're generally trying to do every year with that information really is to help us with vaccine strain selection, and my comments earlier about timing. I think there could be information that could retrospectively or, you know, would be retrospective data by the time we got it. I don't think it would help us with the current vaccine, but could help in the long run with understanding immunologic responses from the attenuated vaccine, if that's what the intent of the questions and comments was earlier.

But I still don't know and others may want to comment whether that could be used in a prospective fashion to say anything about what the likelihood of live attenuated vaccine was having in terms of being immunogenic and effective. And I think we still don't have full understanding on what the meaning of antibody responses from the live attenuated vaccine would be and only because there are multiple pathways by which the immune response may be the systemic antibodies, the local antibodies, secretory antibodies, and also probably some cell mediated responses that help in that.

I think we won't get the full facts from any one measurement on that, and it's something that needs to be done in a better way than just as part of a study that we're really trying to use to identify safety parameter for a new strain.

CHAIRMAN DAUM: Thank you, Roland.

Is this a very, very brief comment? Because we're really beginning to get some time pressure here.

DR. CONNOR: I just wanted to clarify the intent of the study just to make clear that the trial that we have done and are proposing to do is a safety release trial for the vaccine each year. So obviously as people pointed out, the goal is to get the vaccine tested and released quickly.

The other thing is that the adult population is probably not the best population to be looking at the immunogenic response and that we usually can't detect immune response in most of the adult patients post vaccination.

So that would have to be done in a pediatric population or some other setting in some other venue, I guess, is my point.

CHAIRMAN DAUM: Thank you very much, Dr. Connor.

I would like to continue just polling the troops here and ask people to refocus themselves on discussion point three, which is the clinical studies that you would like to see done/required for the release of new strains.

And I think we left off with Dr. Cox next up. Maybe you've already said your piece.

DR. COX: I think that the study that was presented here is really adequate. We are holding the live attenuated vaccine to a higher standard than we do the new trivalent strains, and I think that that's appropriate, and that the proposal here is a good one.

CHAIRMAN DAUM: Can you just say in about six words which proposal you mean?

DR. COX: Where they looked on page 22, the FDA's slide set, there is a slide that talks about annual clinical release testing methods.

CHAIRMAN DAUM: Thank you.

Dr. Gellin.

DR. GELLIN: I have no additional comments on that.

CHAIRMAN DAUM: Thank you.

Dr. Steinhoff, where are you? Gone, not forgotten.

Dr. Myers.

DR. MYERS: I think it's an appropriate study.

CHAIRMAN DAUM: Dr. Overturf is gone.

Dr. Diaz has spoken. Do you want to speak again to this?

DR. DIAZ: I haven't really weighed in on this particular issue.

CHAIRMAN DAUM: Weigh in, please.

DR. DIAZ: I would consider doing essentially what they have described, although I'm not really clear as to what the appropriate study would be. Something along these lines, and perhaps assuring that there is no other adverse events that are not mentioned here that would be considered potential serious adverse events. Obviously it goes without saying.

CHAIRMAN DAUM: Thank you.

Dr. Markovitz.

DR. MARKOVITZ: I have nothing to add.

CHAIRMAN DAUM: Dr. Parsonnet?

DR. PARSONNET: I have nothing to add, but could I address the next question because I also have to leave?

CHAIRMAN DAUM: Yes.

DR. PARSONNET: For the next question which is on post marketing trials, I'd like to see more comparative studies with the inactivated vaccine and also more data specifically on efficacy in the age groups that we have been talking about.

CHAIRMAN DAUM: Thank you very much, and please travel safely.

Ms. Fisher.

MS. FISHER: Well, it sounds like we need to know more about the biological mechanisms for vaccine induced immunity and the correlates for immunity, and so I haven't got a clue as to what the endpoints would need to be.

CHAIRMAN DAUM: Okay. Thank you, Ms. Fisher.

Dr. Goldberg?

DR. GOLDBERG: I think what's being proposed is probably fine, assuming that all the other adverse events are recorded. My only question is if you do this shortly before you're going to develop and produce the vaccine for that year, what does this do to the production timetable. I mean if you had a result that you now had to deal with, when do you do this?

DR. YOUNG: Dr. Daum, would you like me to comment on that?

CHAIRMAN DAUM: I believe I would.

DR. YOUNG: Thank you.

Basically what we do is once we make the MVS, we expand that one more passage and make a clinical trial lot as we move into production. So the clinical trial lot is made in parallel with production of the commercial material, if you will, and while we're making commercial material, we do the safety study at risk as a release test.

DR. GOLDBERG: That was what I was wondering, but it could theoretically impact your production for the year; is that right?

DR. YOUNG: We certainly don't wait for the results before we start production.

DR. GOLDBERG: Yeah, but supposing you're at risk.

DR. YOUNG: Oh, absolutely. If we have a hot strain that has developed, we're -- I don't want to use "dead in the water." That's not a good term to use around you all.

CHAIRMAN DAUM: We can't handle it.

(Laughter.)

CHAIRMAN DAUM: Dr. Steinhoff, did you want to comment on issues related to discussion point three?

Nothing to add. So I'm the last person, and I would like to see. I think the FDA proposed plan is sound. I will make my pitch for annual monitoring of efficacy once the vaccine is deployed, and one idea I had about how to do that, which would obviously take a lot more thought than my comment here, is in areas where there's influenza surveillance to perhaps develop a case control technique to assess efficacy of the vaccine each year.

I find myself wanting to know that about influenza vaccines of any sort and would love to see data about that.

Also, if I hear the drum beats correctly, we're moving more and more toward immunizing children not in the near future, but certainly in the future is my suspicion, and if that's so, the plan to do testing on adults alone that was proposed won't be enough, and there will have to be children included in that annual evaluation as well.

And, Michael, before we leave discussion point three, would you like to give us industry perspective?

DR. DECKER: Very briefly.

CHAIRMAN DAUM: Thank you.

DR. DECKER: Concur with the proposal, and with respect to the questions about additional data, I think it has become clear now that this test can only be -- it's the last safety check for a hot lot. If you failed the test, which hopefully no one ever will, you've lost your production. You may not even be able to go back and make any more in time.

The thought that you're going to get data out of this then allows you to adjust what you're doing that year is simply not supported. You can't do it.

CHAIRMAN DAUM: Thank you very much.

We move on to discussion point four, but I'd like to preempt the discussion by at least pointing out that I believe that a great majority of the things that we'd like to see done have been addressed, but we will survey the committee to make sure there's nothing else that people want to say.

Of course, discussion point four deals with if the data are adequate to support safety and efficacy, to discuss what additional information, if any, should be requested from post marketing studies.

One point to keep in mind as we go around on this is that if you believed that the data were not adequate, then perhaps this question is moot, at least as I understand how it's written. FDA people agree?

DR. MINK: Actually, yes, I agree. Most of the deficiencies from the committee members who voted no they stated at the time they voted.

CHAIRMAN DAUM: I believe so. So we'll just run around quickly and make sure there's no other comments, and then we can consider adjournment.

Dr. Stephens, anything else?

DR. STEPHENS: Well, I want to be sure the list is full and long because I think there is a long list of post marketing issues that need to be addressed.

The revaccination safety issue, for example, and efficacy issue I think is one that needs to be addressed. The whole issue of asthma in children needs to be better addressed and understood because, you know, it was the clear hope that this vaccine would be used in younger children, and I don't understand the asthma issue all that well, nor do I understand it in influenza in general. It needs to be addressed.

The viral shedding and reassortment issue, I think, is still out there and needs post marketing assessment. The transmission high risk individuals, we've discussed that today. It's still an issue that's on the table and needs additional post marketing studies.

Dr. Parsonnet mentioned the issue of comparative studies with the inactivated influenza vaccine, and then the age group issues that we've already discussed in depth.

I would urge also because I think it would really help for future vaccines of this nature to understand the immune correlate of protection. We simply do not understand that for this vaccine. We have just recently been told that antibody is a worthless guide for this particular vaccine, but we need to understand how this vaccine works in terms of future vaccines.

CHAIRMAN DAUM: Thank you.

You'll be pleased to know that those of us that were keeping a list of things that were made before had a pretty good cross-reference with your list, but it doesn't hurt to review, and almost all of those things are actually in the record, which is good, and the agency, I think, has heard, but let's make sure.

Dr. Katz?

DR. KATZ: It seems to me I don't know if it fits under the purview of question four, but what I would like to see is a study of the vaccine used in high risk individuals. We heard a little bit about HIV infected individuals, but I think there are many other groups whom we list as high risk.

I don't know if that's something the company is interested in doing, if it's something that FDA would endorse with an IND, but one way or another I think that questions needs to be answered.

CHAIRMAN DAUM: For several reasons actually.

Dr. Edwards?

DR. EDWARDS: I think that as a pediatrician the delivery of inactivated vaccines currently even to our high risk groups is not very good. I think flu vaccine, internists do much better than pediatricians. That's probably the only vaccine, however.

But I think that the delivery system is an interesting one and does perhaps open much broader array of opportunities for young children to be immunized. So I would really urge that the additional studies on the mechanism of the reactive airway disease be done so that ultimately this vaccine might be delivered to the children that were supposed to be the target population in the beginning.

CHAIRMAN DAUM: Thank you very much Kathy.

Dr. Snider.

DR. SNIDER: Well, with regard to the reactive airways disease, I mean, I think we need to keep in mind that given the numbers, there may be problems in persons older than 60 months of age, and so we need to be attentive to that.

And also, as I mentioned earlier, I think we need to look at this issue of asthma or reactive airways disease in the larger context in terms of what might be induced by FluMist versus what might occur as a result of natural infection and what might be prevented or not prevented by the inactivated vaccine.

And so I think those are interesting questions. I think it has probably already been mentioned that we need to continue to look at the transmission issues, both the inadvertent and otherwise. The reassortment issue obviously needs to continue to be monitored.

Revaccination has been mentioned. Risk groups, safety and efficacy has been mentioned. The age group is not included, may or may not have been mentioned, but I hope they're on your list of things that should be looked at.

And immune correlates have been mentioned. Operational issues also would be of interest, although I'm not sure they're necessarily post marketing studies in terms of things you would lay on the manufacturer, but in the broader context if you interpret post marketing as things that might be done in the context of after licensure by someone, the thing that was mentioned earlier in terms of who will administer this vaccine and can it be self-administered or would it be administered by ancillary medical personnel or pharmacists and so forth; I think those things are interesting topics to look at down the road.

CHAIRMAN DAUM: Thank you, Dixie.

Dr. Hamilton, please.

DR. HAMILTON: I have nothing to add.

CHAIRMAN DAUM: Thank you very much.

Dr. Eickhoff.

DR. EICKHOFF: Nothing further to add.

CHAIRMAN DAUM: Dr. Cox.

DR. COX: Nor I.

CHAIRMAN DAUM: Dr. Gellin.

DR. GELLIN: I want to pick up a little bit where Dixie left off because I think the post marketing studies might depend or will depend on how this vaccine is marketed, and because it has the potential to be self-administered, you could envision how this might be something that is given as a prescription and somebody goes and gets this at a pharmacy and then does their own thing with it.

And given that, I think there may be implications for both analyses of safety and effectiveness when it's self-administered because it's not necessarily something you can do without some instruction.

CHAIRMAN DAUM: Thank you, Bruce.

Dr. Steinhoff.

DR. STEINHOFF: I just want to underline a point that's been made, to ask for additional data on high risk groups, which would also speak to the issue of transmission to that high risk group.

CHAIRMAN DAUM: Dr. Myers.

DR. MYERS: All been said.

CHAIRMAN DAUM: Dr. Diaz.

DR. DIAZ: Nothing to add.

CHAIRMAN DAUM: Dr. Markovitz.

DR. MARKOVITZ: Nothing to add.

CHAIRMAN DAUM: And Ms. Fisher.

MS. FISHER: Nothing to add except I can't imagine that we're going to have people administering this to themselves or to other people. That sounds like a nightmare, prescription for a nightmare to me.

CHAIRMAN DAUM: Thank you.

Dr. Goldberg.

DR. GOLDBERG: Nothing to add.

CHAIRMAN DAUM: And I have nothing to add. Everything has been said.

So with that, two things before we start making noise. One is for committee members. There's a van downstairs at 5:00 p.m. to go to airports.

And, two, I want to thank the committee first, the sponsor second, and of course the agency third for a day of respectful and, I think, good scientific exchange.

Thank you. We are adjourned.

(Whereupon, at 4:42 p.m., the Advisory Committee meeting was concluded.)