P-R-O-C-E-E-D-I-N-G-S

(8:34 a.m.)

CHAIRMAN DAUM: Good morning and welcome. I would like to begin by asking committee members, old and new, and temporary voting members, all those people at the table really, to introduce themselves.

Dave, if you're up for it, we'll start up at your end, please.

DR. STEPHEN: David Stephens, Emory University and other places in Atlanta.

DR. KATZ: Sam Katz from Duke University.

DR. HAMILTON: Holli Hamilton, DMID, NIH.

DR. GLODE: Mimi Glode, pediatric infectious disease, University of Colorado.

DR. OVERTURF: Gary Overturf, University of New Mexico.

DR. FAGGETT: Walt Faggett, D.C. Department of Health, Private Practice Pediatrics, Washington, D.C.

DR. GRIFFIN: Diane Griffin, Johns Hopkins School of Public Health.

DR. WHITLEY: Rich Whitley, University of Alabama at Birmingham.

DR. DIAZ: Pam Diaz, Chicago Department of Public Health.

DR. GOLDBERG: Judy Goldberg, New York University School of Medicine.

DR. MARKOVITZ: David Markovitz, University of Michigan.

DR. PARSONNET: Julie Parsonnet, Stanford University.

DR. DECKER: Michael Decker, Aventis Pasteur and Vanderbilt University.

DR. KOU: Jingyee Kou, FDA.

DR. PRATT: Douglas Pratt, FDA, Office of Vaccines.

DR. GOLDENTHAL: Karen Goldenthal, FDA.

CHAIRMAN DAUM: I'm Robert Daum from the University of Chicago.

DR. SACHS: And I'm Jody Sachs with the FDA, the Executive Secretary for VRBPAC.

CHAIRMAN DAUM: There are a number of people at the table for whom this is their first meeting, including, of course, Dr. Sachs, who has taken over the Executive Secretary role from Nancy Cherry. Tough shoes to fill, but Dr. Sachs is up to the task and I have no doubt will be steering us through with the same aplomb as Nancy Cherry used to do.

In fact, we'll now turn the floor over to her, please, for a conflict of interest statement.

DR. SACHS: Thank you.

I want to welcome everybody, and I'd like to read the conflict of interest statement for the record.

The following announcement addresses conflict of interest issues associated with the Vaccine and Related Biological Products Advisory Committee meeting on May 21st, 2002. The Director of the Center for Biologics Evaluation and Research has appointed Dr. Mimi Glode, Holli Hamilton, and Dixie Snider as temporary voting members for the discussions during this meeting.

In addition, the Senior Associate Commission for Communications and Constituent Relations has appointed Dr. Richard Schwartz as temporary voting member.

To determine if any conflicts of interest exist, the agency reviewed the submitted agenda and all financial interests reported by the meeting participants. As a result of this review and based on the FDA draft guidance on disclosure of conflict of interest for special government employees participating in an FDA product specific advisory committee meeting, the following disclosures are being made.

Dr. Richard Schwartz has been granted a waiver under 18 USC 208(b)(3) and under 21 USC 355(n)(4), Section 505 of the Food and Drug Administration Modernization Act for stock in competing firm valued you at $5,001 to 25,000. Dr. Schwartz may participate fully in the discussions of the safety and efficacy of Prevnar for acute otitis media indication.

We would like to note for the record that Dr. Michael Decker is participating in this meeting as an industry representative acting on behalf of regulated industry. Dr. Decker's appointment is not subject to 18 USC 208. Dr. Decker is employed by Aventis.

In the event that the discussions involved specific products or firms not on the agenda and for which the FDA's participants have a financial interest, the participants rare reminded of the need to exclude themselves from the discussions. 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 or products you wish to comment upon.

A copy of the waiver addressed in this announcement is available by written request under the Freedom of Information Act.

And I also would like to ask as a courtesy to the committee discussion and your neighbors in the audience please put your cell phones and pagers on silent mode. If you need to use your cell phone, please step out in the hall.

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

Thank you.

CHAIRMAN DAUM: And those that have just turned their cell phones and pagers off, we thank you.

I think we'll try and zip right along here and turn to business at hand. The first item for discussion today is an open session. We are discussing the role of Prevnar for an acute otitis media indication.

And we will begin with a two-part, as I understand it, sponsor's presentation, beginning first with Steve Black, which will give us a Prevnar update.

Welcome, Dr. Black.

DR. BLACK: Good morning. I've been asked to give an update on an ongoing post marketing, Phase IV study that we're conducting within Northern California, Kaiser Permanente, of the Prevnar vaccine, and I'll give you an update which includes an interim analysis on safety and results regarding the changes in epidemiology that we have observed of pneumococcal disease in our population.

The post marketing study that I'm going to describe to you, let me give you a little bit of background on that. The vaccine Prevnar was licenses in February of 2000, as you know, and post marketing surveillance began in our population very shortly thereafter with general availability of the vaccine in April.

And the vaccine is being given now routinely to children concomitantly with other vaccines.

What I'm going to describe to you this morning in terms of safety is a second interim look on data through December 31st, 2001. There was an earlier interim look through December of the year prior to 2000, which has been submitted to the FDA, and they've had time to review, and this, I should say in fairness to them, has only recently been submitted to them for their review.

Following the review of the safety, I'll talk to you about the impact of the vaccine and present what we think is exciting data on the changes of epidemiology that we've seen, which includes data through the end of the first quarter of this year.

Okay. So this shows you what happens if you keep tinkering with slides, but what I will show you here is that there are two cut points. One is December 2000 and December 2001, and what you can see here is that as of 2000 in the post marketing study or what you can't see -- I'll read it to you -- is that there were about 22,000 first doses given, whereas through December 2001 there were 54,000 first doses given, and there were only 85 fourth doses in the initial look, where there's 17,000 in the second look. So there's substantially more data there.

So back to visible slides now. The way this is set up, and since this is a post marketing study is that there is no control group, and what we're doing is comparing rates of medical utilization within a defined time window, exposure window following vaccine to a control period in the same individuals.

And the exposure window is 30 days for hospital ER and clinic, and there's an additional window in the clinic of three days that we've used to evaluate possible allergic reactions, for example.

And the control period in these comparisons that I'm going to describe to you is 31 to 60 days following vaccine for all settings.

Also what I'm reporting on here is the subset of children who received the first dose of vaccine at less than 120 days of age. In other words, catch-up and children who started late are not included in this analysis.

And the way we did this is we extracted all diagnoses for medical utilization in the clinic, emergency and hospital, from automated databases that exist at Kaiser Permanente and then rate comparisons were made for all diagnostic categories in the ER and the hospital, and for pre-identified clinic diagnoses as specified in the protocol for the clinic.

In addition, because of concerns expressed regarding a possible association of seizures with receipt of vaccine, we have conducted a review of seizure outcomes using medical record review, and I'll report that separately to you.

To give you an idea, not that you need to read this, this just gives you an idea of the number of diagnoses that were reviewed in the emergency hospital and clinic. These were basically, as I said, for the ER and the hospital all diagnoses, and it's important to be aware of this number because the statistics that I'm going to be describing to you are not adjusted for multiple comparisons.

And I hope there isn't too much information over here on the right. We'll try to capture that, but what this shows you are the diagnostic categories with elevated risk in this comparison.

This is a hospital setting, an emergency setting, and clinic setting, and then which series: the primary series or the booster dose? And for this analysis the primary series was analyzed as a unit, all three doses together rather than looking at each dose separately.

And what we see here is the outcome and then the rate ratio with a confidence interval and part of the P value here.

And what you can see is really there are two groups of diagnoses. These three, GE reflux, pyloric stenosis, and formula intolerance as a diagnosis.

The rate ratio is here indeterminant because there were no cases in the control group, and then these febrile illness in the emergency room, in the clinic, and fever related diagnoses, which was a predefined diagnostic category in the clinic also showed up, and this entity is basically febrile seizures plus fevers. Febrile illness is pretty much driven by the febrile illness as you can see.

Next slide. Oops, that's me.

Okay. So these are the diagnostic categories with decreased risk. To give you an idea, there are actually more of them than the ones with increased risk, and we really attribute this to the multiplicity of comparisons rather than any protective effect for otitis media, for example, because remember the control period here is in the same children. So that wouldn't really make physiologic sense.

So we looked at these, and the elevated relative risk in a little bit more detail, and this is one of these, febrile illness in the emergency room after the booster dose, and this is the n, the number of events here, and this is the days since vaccination, and the 30-day exposure window.

And what you can see here is what we look for in this type of analysis when we see something that we might think might be physiologic, and that is a clustering of events at one time period, and we see these are eight to ten days following receipt of these vaccines.

If you remember, the booster dose is given concomitantly with MMR in the vast majority of these children, actually more than 90 percent, and we attribute this to the well described fever associated with MMR at this same time interval rather than the fever that we described in telephone interviews where we were actively looking for this and during the trial with Prevnar which was seen earlier on. So we're not really seeing that blip here.

Similarly, in the clinic, we see the same thing with the same time clustering of these events for febrile illness in the clinic and only after the booster dose.

In contrast for GE reflux, what we really see is not that. We see really pretty much a uniform distribution of these events spread out over this time window, and similarly for pyloric stenosis the data is much more sparse, but there really is no time clustering of the event or interpretation either.

Similarly, with formula intolerance as well.

So although seizures did not show up as a positive analysis in these reviews that I showed you, we had planned before doing this interim analysis report to do the seizure review, and let me describe that to you.

What we did is attempted to identify all possible seizure events in automated data by looking for seizure, possible seizure, epilepsy, spasm, shaking or suspicious movements, and those were then reviewed in a manner that was blinded as to whether they were in the exposure window or the control window by trained medical record reviewers using a standardized instrument, and they were classified as definite, probable or possible seizures or the other category was not seizures at all. There was a group of children who were there for maintenance or for assurance or for other things that were really not acute events.

But acute events were classified in one of these categories. Based upon what the physician wrote in the chart, if they described a definite seizure event or one was described then that was classified as definite, and if the physician's interpretation was that this was a probable seizure, then we took that at face value.

But if it was something that was included as part of a broader differential and they really weren't sure, and there were no confirmatory tests, and no medication was given, we thought it was less likely and that was classified as possible.

So a priori before doing the analysis we had decided we would want the definite and probable seizures together as a group and then analyze them as events, and those were classified as febrile or afebrile based upon, one, whether it's two possible criteria.

One is if it said they were febrile on the chart, we counted it as febrile, or if there was actually fever recorded by one of these two criteria, and our physicians are a little schizophrenic as to which temperature scale they use. So we had both criteria.

And these are the results for seizure, and I'm sorry this is complicated, but if you slide and dice things enough, this is sort of what happens. This is the hospital setting again, the emergency setting, the clinic, and then the series for this comparison, primary and boosters, primary and booster, primary and booster, and then the outcome, afebrile seizures or febrile seizures.

This is the exposed rate. This is the control rate, and then this is the rate ratio with a confidence interval, and then the P value.

To make a long story short, seizures were uncommon in either window and there was no statistical difference for any of these rate ratios. And furthermore, as you can see, there are a fair number that are below one, a fair number that are above one, and there is really not even any suggestion of a pattern here. So we found that quite encouraging in terms of the safety of the vaccine.

And we also looked at, to give you an idea of what these look like, these are emergency visits for febrile seizures after the primary series.

There isn't really any clustering of this, surely not within the first few days where fever is observed with Prevnar.

And after the booster dose, this is not statistically -- there is no statistical clustering here, but we do see that there are more of these events at the same time period where we saw fever in the emergency room as well.

And, again, if there's anything here, we would probably attribute that to the fever of MMR rather than Prevnar.

So a summary of the safety analysis to date, and I would like to emphasize that this is ongoing and not the final results by any means, is that our analysis showed an increased rate of utilization for febrile illness following the booster dose, and the timing of this fever suggests a relationship to concomitant MMR.

Other events observed with an increased risk, including GE reflux, pyloric stenosis, and formula intolerance, were not felt to be physiologically likely. The analysis, as I said, and data collection are ongoing.

And furthermore, the results are consistent with the first interim analysis which the FDA has had more time to review, as well as with pre-licensure data from our own infancy trial.

So that's the safety data I wanted to share with you, and now I'd like to share some exciting information; at least we think it's exciting vis-a-vis what's happening with disease epidemiology in our population since introduction of the vaccine.

Again, Prevnar was still licensed in February of 2000, and general use began in April. For the evaluation of effectiveness case ascertainment, it's important to emphasize here it was for the whole Kaiser population. One, children and adults, and both vaccinated and non-vaccinated.

So unlike the efficacy trial data we showed you where we're comparing a vaccinated/unvaccinated group, we're really looking here at the population dynamics as a whole and the effectiveness of that vaccine program.

And to look at this effect, we compared the disease risk in the two years since vaccination compared to prior years, as you'll see. All isolates, Strep. pneumoniae from normally sterile sites were identified from laboratory databases, and then the isolate was sent to Dr. Robert Austrian for serotyping.

The medical records of all the infected children have been reviewed to ascertain and confirm vaccination history and history of any underlying disease.

And then we calculated age specific disease incidence. So this is the graph I would like to show you, and I will remember if we come back next year to move things over to the left here a little bit because we won't be able to see this.

But let me orient you to this slide. This is the incidence in cases per hundred thousand person-years ranging from zero to 120 at the top, and these are years at the bottom. Each dot is a year, and the years are unusual in that they began in the second quarter of each year.

And the reason we did that is that's when the vaccine program began. So we wanted to be able to make the comparison of comparable.

And what we see here in this yellow line is children less than two years of age, and we see that prior to introduction of the vaccine to general use, the disease incidence in this group ranged between 80 and about 110-plus cases per 100,000 person-years and then falls off to virtually nothing here, less than ten disease incidents during the year beginning in the second quarter of 2001 and ending in the first quarter this year.

Similarly for children under one, the disease incidence as you know is somewhat less, ranging between 50 and almost 100 here and then falls off quite dramatically. You can see this fell off more steeply because that's where the vaccination program began, and for children under five, we see this as well.

There are five cases total that we saw during this year as compared to about 120 during years prior to introduction of vaccine. Only one of those children was vaccinated, and that child was partially vaccinated.

One of the concerns has been that we might see replacement. It's commonly said nature abhors a vacuum, and there's been a concern that other serotypes would come in and causae disease.

I guess I'd better hurry before something happens here. That's okay. I'd rather live with it this way than lose the whole thing.

What that shows in blue is the same graph that I just showed you in the different age groups, and then below these are non-vaccine serotypes, and what you can see is that, one, the incidence is lower as we all know, and if anything, there is a downward slope to the graph although that trend is not statistically significant, but there's clearly no suggestion of replacement for invasive disease up until this point in time.

And this is something that is actually quite new. This is something that we just presented at the pneumococcal disease meetings in Anchorage a couple of weeks ago, and what we did here is used the same surveillance mechanism to look at disease in older children and adults, and this is the age group here. This is the rate in the five years prior to introduction of vaccine, and this is the rate in the two years after the percent reduction, and part of the P value here.

And what we can see in yellow are shown the two age groups where there's a significant -- or three really if you count this -- age groups where there's a significant reduction in the disease, really quite strikingly dramatic, something we would not have predicted in the 20 to 39 year old age group, a 58 percent reduction in invasive disease in this age group.

Now, most of these have not been serotyped. So this is really all serotype disease. Over age 60 we see a 14 percent reduction, which is also significant, and then over age five we see an 18 percent reduction.

It's important in fairness to say that over age 60 there have been changes in terms of the polysaccharide vaccine coverage in our population which could account for part of this. We estimate there's been about an eight to ten percent increase in coverage over that time period.

But that's not true in this younger age group which we attribute this to the fact that this is the age of the parents of the children who are being vaccinated, and the children it is known -- contact with young children is a risk factor for pneumococcal disease, and we believe that this is entirely suggestive of the fact that herd immunity is operative here and is protecting these individuals.

So, in summary, we've observed a dramatic reduction in basic pneumococcal disease in childhood within our population. The magnitude of the reduction in the first year, which was much greater than the vaccine coverage, and the reduction observed in adults suggests herd immunity effect.

We've not observed any evidence of serotype replacement for invasive disease, and I'd like to say also that Dr. Cindy Whitney of CDC has results from the ABC surveillance program which are consistent with the disease reduction in adults and older children that I've shown you.

Thank you very much.

CHAIRMAN DAUM: Thank you, Dr. Black, for that update.

We have a few minutes for committee questions, if there are, or discussion points. Dr. Katz?

DR. KATZ: Steve, you mentioned the concomitant administration of MMR. Was varicella given at the same time also?

DR. BLACK: Yeah, varicella vaccine, the uptake for varicella vaccine is quite high in our group, and we looked at MMR. There's more than 90 percent of that concomitantly. Usually varicella is given at the same time, but it isn't always. We have not looked at it, but I would guess from past observations we had made it was about 80 percent.

DR. KATZ: The reason I asked is there is some indication that when you give MMR and varicella concomitantly you even further increase the febrile response.

DR. BLACK: At that same interval.

DR. KATZ: Thank you.

DR. BLACK: Actually we'll look at that. that's interesting. Thank you.

CHAIRMAN DAUM: Dr. Faggett and then Dr. Snider.

DR. FAGGETT: Steve, thank you. Those are very exciting reports. A question relative to the experience of Prevnar in the sickle patient. They were probably included under your febrile illnesses, but do you have any information on specifically how the vaccine was tolerated by sickle patients?

DR. BLACK: Yeah, we've not done specific studies on the safety of Prevnar in sickle cell patients. However, the Prevnar vaccine is being routinely used in both younger children with sickle cell disease and in older children as well, and our surveillance does include children with sickle cell disease, and we've not seen during the last two years because they were not surprisingly targeted for early immunization any cases in children with sickle cell disease.

CHAIRMAN DAUM: Dr. Snider?

Steve, I have one question. The adult data you showed on the last slide are pretty interesting. You mentioned that you haven't yet broken them down by vaccine serotypes and non. Will you be able to do so? Do you have the isolates?

DR. BLACK: No. We started collecting data at the first of this year. We're now -- Dr. Austrian, since the case load in children is reduced, is now willing to do serotyping of adults, and so beginning the first of this year, we're now serotyping all ages, but don't have that historically.

Dr. Whitney at CDC, however, does have serotype data from ABC and I think is analyzing that currently and will be reporting it soon.

CHAIRMAN DAUM: I have on other question. In the very nice curves you showed of what's happened to disease in your area since the vaccine was introduced, you broke down the data between vaccine serotypes and non-vaccine serotypes.

How would that look for the non-vaccine serotypes which did appear to be trending down? If you removed the related vaccine serotypes -- excuse me. The serotypes that are not in the vaccine but are related to those in the vaccine from that analysis.

DR. BLACK: Okay. Let me try and rephrase your question. What you're looking for are the non-cross-reacting serotypes.

CHAIRMAN DAUM: Right. Thank you for that help.

DR. BLACK: We have a slide for that here. Let me see if I can find it. We also have a million other things.

Oh, you have that somewhere else? Okay. Sorry.

The numbers are smaller and so there's more noise in this, but let me show it to you.

Yeah, okay. So what we have here is, again, the same type of graph, but you'll notice that rather than going up to 120 or 40 here, this only goes up to 20, and again, with the same age groups, under one you can see actually now has a higher incidence of these. Under two, and then under five, and you know, the overall slope here is sort of downward, although I don't understand that, and this dot, this little blip at the end here is really in the same range as these.

So so far, you know, the numbers here are a lot smaller. So it's a little bit harder to interpret, but we don't think this suggests any evidence of replacement disease because the incidence levels here are very low, consistent with what we saw before.

CHAIRMAN DAUM: Thank you.

We'll take two more comments. Dr. Snider, then Dr. Stephens.

DR. SNIDER: Steve, could you tell us what the serotypes that are vaccine related that you're still seeing are? I mean, specifically people I'm sure that have read the material have some concerns about 19F, for example.

DR. BLACK: Yeah. So the question is, you know, is 19F -- do you mean in vaccinees or in -- yeah, we've really not seen -- I mean the cases of disease that we've seen in the last couple of years since the post marketing took place have not included 19F. There's a couple of fours and one 6B, and that's really about it.

So the concern that we and others had in terms of trying to understand the difference in response to 19F, we're really not seeing that translated into breakthrough disease up until this point in time.

CHAIRMAN DAUM: Dr. Stephens.

DR. STEPHENS: Regarding the effect in young adults, is there any evidence in your health care system of off label use of the conjugate or any increased use of the 23 valent polysaccharide in individuals who may be at risk?

DR. BLACK: Well, we're encouraging increased use in individuals, you know, over age 60. so that has gone up we estimate eight to ten percent over the time period.

The older individuals where we're encouraging its use is primarily hemoglobinopathies or people who are in that category.

There has been some use in older individuals where it's not indicated, but it's very small. There's four or five individuals for reasons that we can't understand who have obtained the vaccine. Four of them are pediatricians. So maybe that's it. They're enthusiastic and want the same protection for themselves. But they're really a handful. It's very, very small.

So it's not the case in the 20 to 40 year olds. We are going to be undertaking a case control study to look at risk factors and look at this in more detail, but that will take some time.

CHAIRMAN DAUM: Dr. Katz, one last.

DR. KATZ: One quickie. In all of those things that are flashing by when you were trying to find the right slide, one that stood out in my mind was sudden infant death syndrome. That's one that in your primary series you're running through the high risk area.

Can you reassure us about that one?

DR. BLACK: Yeah, let me see if I can find that slide.

DR. KATZ: I don't need a slide. Just tell me.

DR. BLACK: Okay. I mean, the rates that we have for that are not for the last year because the state death tapes lag. So as of the interim report that we did through year 2000, the SIDS rates were about half what the state rate was, and were pretty much identical to what they were in the clinical trial, which is about .2 per thousand.

CHAIRMAN DAUM: Thank you very much, Dr. Black.

We sometimes remember and are striving to meet various bars of vaccine safety and various tests and concerns, just how wonderful vaccines are, and it's very gratifying to see this kind of information after the introduction of a new vaccine.

We will move now on to the second part of the sponsor's presentation this morning, which is concerning acute otitis media, or AOM, and we will begin with Dr. George Siber, who will introduce the topic on behalf of the sponsor to us.

Dr. Siber, welcome.

DR. SIBER: Good morning. My name is George Siber. I'm Senior Vice President and Chief Scientific Officer of Wyeth Vaccines.

Is that going to go to right for us? We'll see.

In any event, during the next hour or so we'll present series of presentations on the data and rationale underlying our proposal for an indication for otitis media for the seven valent pneumococcal vaccine, Prevnar.

I'll give a brief introduction on otitis media epidemiology and background. Dr. Terry Kilpi, who is a senior researcher and the head of the Department of Vaccines at the National Public Health Institute in Helsinki, will discuss the FinOM trial that was conducted in Finland, and then Steve Black will come back and discuss otitis media from the Northern California Kaiser Permanente trial. And then I'll give brief conclusions at the end on impact.

First of all, a quick background on clinical manifestations of otitis media or rather of pneumococcal disease in general. This pie diagram shows you the major pneumococcal syndromes and makes the point that the pneumococcus is a very important if not the most important single pathogen contributing to major bacterial infections in U.S. children, causing 45 percent of meningitis in the first two years of age, a vast majority of bacteremia sepsis, and for these two Prevnar is indicated in the package insert, but also about 60 percent of pneumonias and as much as 40 percent of bacterial otitis media.

This shows you a pyramid which puts into perspective the relative frequencies of these syndromes. Fortunately the most severe of those syndromes are the least common, with about 1,400 cases in children under five years of age of meningitis, 17,000 of bacteremia, and estimated 71,000 for pneumococcal pneumonia.

But at the base of this pyramid and really a massive number is the five million estimated episodes of otitis media each year, and although clearly a much milder disease than the others, it certainly has morbidity and has a very tremendous impact on health care and antibiotic use and so forth.

With regard to the epidemiology of otitis media, these are actually data from the Northern California Kaiser Permanente trial and the control groups looking at the age distribution of otitis media, and which show several things.

One, that in boys, in blue, the rates are somewhat higher throughout follow-up period, here to 42 months of age, than in girls. And the peak incidence is very high, and this is otitis visits per 100 children-months between six and 18 months of age, but really continues throughout the follow-up period, declining slowly but steadily with time.

A somewhat more extended age distribution comes from these data, which plot the number of visits for otitis media to physicians' offices in thousands by year from zero to ten years of age, and you can see that the peak here is 4,400,000 visits, and again, a decline over time, but continuing to have as many as five to 600,000 visits per year even out to ages nine and ten years of age.

So to summarize the impact of otitis media, this is the most common reason for sick child visits. It is also the leading cause for prescribing antibiotics during childhood, and we believe that the use of antibiotics frequently contributes to the increasing antimicrobial resistance that we have seen in this country and elsewhere.

Complications of recurrent disease and effusions lead to tempanostomy tube insertions, and this is the most common reason why children have surgery that requires general anesthesia.

The direct and indirect annual costs have been estimated to exceed more than $5 billion per year in children under five years of age. That's for all otitis media.

And this just shows you, I think, what we all know, and that is during the '90s there has been a progressive increase, looking here at pneumococcal disease, an increased rate of resistance from the low digits, five percent or so, to over 30 percent at the end of the decade.

An interesting question is whether there will be an impact of Prevnar on this phenomenon.

Importantly, the serogroups that are most likely to be resistant to penicillin and other antibiotics are serogroups that are contained among the seven valent types of the vaccine, six, 14, 19, 23, and nine. And that's true not only in the U.S. but throughout the world.

And specifically in terms of coverage for otitis media, this is an example of a study by Ellen Wald's group in Pittsburgh reasonably recently looking at serotype distribution in otitis media and suggesting a coverage of the vaccine serotypes themselves of about 70 percent.

If you assume coverage for cross-reactive types, that goes up to 85 percent, and if you only selected antibiotic resistance, you would probably get up over 90 percent in this series in terms of coverage by the vaccine types and related types.

So at the moment, you my be aware that the package insert makes no mention whatsoever about otitis media with regard to Prevnar efficacy, and we are here today to propose that otitis media be included in the package insert and that the indication be that Prevnar is indicated for active immunization of infants and toddlers against invasive disease and otitis media caused by Strep. pneumoniae due to the capsular types included in the vaccine.

And some of the reasons why we believe this is to be important is that there are now two randomized, well controlled trials that you'll hear about which show statistically significant decreases in otitis media outcomes.

Secondly, you'll hear that Prevnar immunization does have an important medical effect on otitis media disease and its implications, and that we believe it's important that this information, since it's now published and talked about in the literature, be accurately described in the label so that we can communicate appropriately information to physicians and to parents.

The trials that you will hear about just very briefly are the FinOM efficacy trial, the major trial that has been reported in the New England Journal under the direction of Juhani Eskola and Terhi Kilpi who's here with us today, and then a follow-up trial focusing most clearly on tempanostomy tube placements in Finland in the follow-up period, which Terhi Kilpi will describe.

And then Steve Black and Henry Shinefield will present the data updated on the Kaiser Permanente trial.

The two trials, just to contrast them, really they're quite different, but they give complementary data. The FinOM trial, of course, done in Finnish infants receiving a U.S. schedule of vaccination, a relatively smaller number of children, 1,600 or so, but I think what's very special about this trial is that myringotomies were performed, and we have culture specific diagnosis of the etiology of acute otitis media.

In contrast, the Kaiser Permanente trial was much larger, more than 37,000 children, Northern California. The diagnosis was made clinically rather than in a standardized way on a routine basis by hundreds of physicians and was captured from the automated databases at Kaiser.

And with that I'll ask Dr. Terhi Kilpi to come up and tell us about the FinOM studies.

CHAIRMAN DAUM: Before he does that or she does that -- I'm sorry -- does the committee want to ask any clarifying questions about Dr. Siber's presentation? Data that were unclear?

(No response.)

CHAIRMAN DAUM: Okay. Thank you very much, Dr. Siber.

Dr. Kilpi, welcome.

DR. KILPI: Good morning. I'm going to present the main efficacy results of the Finnish otitis media vaccine trial that evaluated the efficacy of two seven-valent pneumococcal conjugate vaccine for prevention of acute otitis media due to vaccine serotypes in children less than two years of age.

And this study was conducted in the Tampere area in Finland, and the clinical phase started in December '95 and ended in March '99, and during this time, we had almost 2,500 children were enrolled in the study. This is approximately 55 percent of the birth cohort in the area.

And all of these children were randomized to receive either one of the two pneumococcal conjugate vaccines used in the study, the PncCRM vaccine labeled, licensed as Prevenar or the PncOMPC vaccine or the control vaccine that was Hepatitis B vaccine in our study.

And the children received these vaccines at the age of two, four, six, and 12 months. They were followed in study clinic setting from two months to 24 months of age, and during the follow-up every effort was made to have all respiratory infections according to these children requiring medical attention evaluated and treated at the study clinics by our study physicians.

This trial was specifically designed to study otitis media, and therefore, we needed a definition for acute otitis media, and we defined that there has to be symptoms of acute infection and signs of inflammation in the middle ear.

And whenever acute otitis media meeting this definition was diagnosed at the study clinic by our study physician, myringotomy was performed and middle ear fluid aspirated for bacterial culture, pneumococcal serotyping when appropriate, and pneyumolysin PCR.

Otitis media is a condition that tends to recur in a proportion of individuals over and over again, and we, therefore, wanted to analyze the vaccine efficacy by all AOM episodes rather than just the first ones, and we, therefore, needed a definition for an episode.

And we defined that it starts at diagnosis and lasts for 30 days. And these were the endpoints we looked at, and these were defined in the protocol and in the analysis plan. The primary endpoint was all AOM episodes due to vaccine serotypes.

The secondary was first and subsequent AOM episodes due to vaccine serotypes, and we also looked at all pneumococcal AOM episodes, at all AOM episodes, and recurrent AOM.

We have late also performed some additional analysis, looked at endpoints of special interest, namely AOM episodes due to vaccine related serotypes, due to serotypes unrelated to vaccine types, and also calculated the vaccine efficacy against AOM episodes due to individual pneumococcal serotypes.

And from now on, I will present the results for the PncCRM group of this study as compared to the control group and forget about the third arm since this is the vaccine we're talking about today, and to start with, I hope this slide will demonstrate to you that our trial was very successfully conducted.

Of the 1,662 children enrolled in these two groups, as many as 1,580 completed the trial without critical protocol violations. That is, 95 percent of the children originally randomized. So we feel pretty comfortable with the results.

And now to the results. During the protocol follow-up period that lasted from 6.5 to 24 months of age, there were 107 AOM episodes due to the vaccine serotypes in the PncCRM group as compared to 250 episodes in the control group.

And this means that the vaccine efficacy against the primary endpoint, all AOM episodes due to vaccine serotype was 57 percent, and this efficacy was statistically significant as indicated by the confidence interval here.

And to the secondary analysis, the vaccine efficacy against AOM, first episodes of AOM due to vaccine serotypes was 52 percent, and the vaccine efficacy in the subgroup of children who had already had one AOM caused by the vaccine serotypes was 48 percent. So the vaccine does provide protection even if a tad failed one.

And this is a summary of the main efficacy results, AOM, vaccine efficacy against AOM due to vaccine serotype, 57 percent; against culture confirmed pneumococcal AOM, 34 percent; against pneumococcal AOM confirmed by either culture or PCR, analyzing PCR or both, 20 percent. These are all statistically significant. Against any AOM, six percent, and recurrent AOM, 16 percent. The latter two failed to reach statistical significance in our study.

And these were the analyses for the protocol analysis, and this is the same for the intention to treat analysis and for the intention to treat follow-up period that started already at two months of age.

And as you can see, the results are very similar to the protocol analysis. What may attract attention in these efficacy results is the different efficacy the vaccine provided against culture confirmed pneumococcal AOM as compared to pneumococcal AOM confirmed by either culture or PCR, and therefore, we have looked at this issue a bit more closely and found that the vaccine does not provide any protection against chemical culture, negative but PCR positive AOM, and this explains the difference between these two entities.

And since the PCR method we used in our study was quantitative or perhaps more precisely semi-quantitative, we have also been able to look at the PCR counts in the pneumococcal culture negative cases of AOM as compared to the Pnc culture positive cases and found that the PCR counts are considerably higher if the pneumococcal culture is positive than if it's negative.

So whatever the significance of PCR positivity in the pneumococcal culture negative cases of AOM is, it certainly does not seem to be a sign of active pneumococcal disease.

The design of the FinOM vaccine trial allowed us to characterize the vaccine efficacy a bit further because we had the culture results from each even of otitis media and we had the serotyping results. And one of the things we were interested in was if the vaccine provided the same kind of efficacy or different kinds of efficacy against AOM caused by individual vaccine serotypes, and this is what we found.

The efficacy against AOM caused by 6B was excellent. The point estimate is 84 percent. It's good against AOM caused by 23F and 14 point estimates, from 60 to 70 percent, but rather modest for AOM caused by Type 19F, point estimate being only 25 percent.

When we designed the trial and decided to have AOM caused by the vaccine serotypes as our primary endpoint, we knew that we could anticipate that the vaccine might protect also against other than vaccine, against AOM caused by other than vaccine serotypes only, and that is the relative serotypes to the vaccine serotypes, and therefore, we have also wanted to look at this and we found, indeed, that there were 41 AOM episodes caused by the vaccine related serotypes in the PncCRM group as compared to 84 episodes in the control group, and this means that the vaccine efficacy against AOM due to the vaccine related serotypes is 51 percent, which is almost as good as the efficacy against AOM caused by the vaccine serotypes themselves.

However, when we come to the other serotypes, the non-vaccine, non-vaccine related serotypes, we see in excess of 30 episodes caused by these serotypes in the PncCRM group as compared to the control group, which translates into a negative efficacy of minus 33 percent in the vaccine group as compared to the control group, and this difference almost reached statistical significance.

However, the bottom line is that the vaccine provides protection against any culture confirmed pneumococcal AOM and reduces it by 34 percent.

And this is now vaccine efficacy against AOM caused by the two most common cross-reactive serotypes, 6A where the point estimate is 57 percent and 19A where the point estimate, 34 percent, actually is even a little higher than for the vaccine serotype 19F itself.

So conclusions from this trial follow-up part are the the PncCRM vaccine is efficacious against culture confirmed vaccine serotype specific, active otitis media, culture confirmed AOM due to the vaccine related serotypes, and culture confirmed pneumococcal AOM.

And now I will move on to the extended follow-up. We have recently collected additional information on the children enrolled in the PncCRM and control groups to assess the long-term effects of the PncCRM vaccine on pneumococcal carriage, antibody persistence, and surgery due to otitis media in the routine practice when those children had completed the trial follow-up.

And I will now present the results for this category as specifically the effect of the vaccine on the incidence of tympanostomy tube placements up to four to five years of age.

I will also briefly present some results for the other two categories.

And this extended follow-up was carried out by inviting the children to a single follow-up visit in spring 2001 when they were at the age of four to five years. And we invited altogether 1,490 children. They represent 90 percent of the original study population, and these were the children who had completed the ITT follow-up and who were still living in the Tampere area.

And 756 of these children followed the invitation and were evaluated at the study clinic in spring 2001, and since these children only represent 45 percent of the original study population, we have also collected information on the tympanostomy tube placement of these children, these 1,490 children to be able to feel comfortable with our tympanostomy tube results.

And I will now show you what kind of data we have available on the tympanostomy tube placements of these children and for which categories of children we have this date.

So, first, the analysis populations. Initially all children were followed from two to 24 months of age in the study clinic. So we had 1,662 children at the beginning and 65 of them dropped out during the trial. So at the end we had 1,597 children, and of these, 107 had moved out or the Tampere area after they completed the follow-up in the trial setting.

So we had, 1,490 children still living in the area, and these children constitute the eligible children, the analysis' population two.

Then we have this subgroup of children, the 756 fully evaluated children, and they constitute the analyst population one, and for this part of children, we have completed tympanostomy data available, and for this part of children, we have the hospital tympanostomy tube data available.

And tympanostomy tube placement in the FinOM follow-up study were ascertained in the following way. For the fully evaluated children, we could ask the parents if the child had had tubes placed after completing the trial follow-up and then confirm the parents' answers by reviewing the hospital records collected from the area hospitals and by reviewing the medical records requested from private physicians.

And it turns out that 78 percent of the tympanostomy tube placement had been performed in public sector hospitals and 22 percent in private medical centers.

For the eligible children we had the hospital records from the area hospitals which are likely to represent approximately 80 percent of the tympanostomy tube placement performed in these children after they completed the trial follow-up.

And before I go to the results, I think I need to explain to you what kind of practices were followed during the vaccine trial and after it when the children returned to normal life, to the real life situation.

During the trial, tube placement, if considered indicated, was included in the study services. They were almost exclusively performed at the Tampere University hospitals. They were free of charge to the patients, and the hospital guaranteed access to treatment within four to five weeks of referral.

When the trial follow-up was over, the children returned to normal life, and in the real life situation in Finland if tube placement is considered indicated, there are two options, two possibilities to have it performed. It can either be done in public hospitals where the charge is nominal, but waiting time can be from three to four months, or it can be performed in private medical centers that charge ten times that of their public sector charge, but there is no waiting time.

And so principally, the indications for tympanostomy tube placement were the same during the vaccine trial and after the trial when the children had returned to the normal life situation, but access to treatment became definitely more difficult when the trial follow-up was over due to the reasons here.

And this makes plain why the incidence of tympanostomy tube placements in the FinOM children during the vaccine trial follow-up was considerably higher than what it is in the children of the same age in Finland in general.

And it also makes plain why this incidence of tympanostomy tube placement dramatically dropped when they returned to a normal life situation. So it appears that milder cases of recurrent AOM and otitis media with effusion were treated with tympanostomy tube placement during the trial and after it, and this makes plain why the effect of the vaccine on the incidence of tube placement was different here from what it was here.

Okay. Now I'll go to the results, and we'll just remind you that I'm going to present them for the full evaluated children analysis population one and for the all eligible children analysis population two.

And these are the tympanostomy tube placements in the fully evaluated children. During the trial follow-up from two months to two years of age, 20.3 percent of the children in the PncCRM group as compared to 23.8 percent of the children in the control group had tympanostomy tubes place, and the incidence rate of events is here.

So the difference between the vaccine group and the control group is 12 percent, and this is not statistically significant.

However, when the normal life situation started during the period from two years to four to five years, only 8.2 percent of the children in the PncCRM group as compared to 13 percent of the children in the control group had a tympanostomy tube placement. The incidence is shown here, and the conclusion is that the vaccine reduced tympanostomy tube placements during this age period of time by 39 percent, and this difference is statistically significant.

And since we have only 45 percent of the original study population in these fully evaluated children, it was, of course, important to see if the results are the same for all eligible children for which we had the public sector hospital data available. And so now only the tympanostomy tube placements performed in the public hospitals are included in this slide.

And here the difference during the trial follow-up is even smaller. It's only four percent, but, again, during the normal life follow-up from two years to four to five years of age we see a reduction of 44 percent in the incidence of tympanostomy tube placements in the PncCRM group as compared to the control group.

And even the lower limit of the 95 percent confidence interval is as high as 19 percent.

Now, this shows the same thing for the fully evaluated children graphically. This is the cumulated hazard for tympanostomy tube placement, and as you can see, there is practically no difference during the trial follow-up up to 24 months of age, but after, as soon as they return to normal life, the curves start to part and continue to do so.

So there is no sign of waning efficacy here. And this is the same thing for all eligible children, and again, the same pattern.

I will now show briefly kinetics of antibody concentrations for three of the most serotypes causing AOM in our study, and I think that these curves are beautifully consistent with the persisting efficacy I have just demonstrated.

This is the antibody concentrations for 23F, and as you can see, the level is the same at the age of 24 months and then at the age of four to five years.

For serotypes 19F and 6B, the antibody levels even seem to increase a little.

And this is data collected at the follow-up visit in spring 2001. We asked the parents if the child has had AOM after 24 months of age, and according to the parents of the children who received the PncCRM vaccines, 67 percent of these children had had AOM after completing the trial follow-up as compared to 72.7 percent of the children in the control group.

At the visit, 11.4 children in the PncCRM group as compared to 12.5 percent in the control group had middle ear abnormalities, and 8.5 percent of the children carried vaccine serotypes as compared to 13.6 percent of the children in the control group, and this last differences is statistically significant.

So these last data is consistent with the conclusions that PncCRM reduces tube placement due to otitis media, and that the vaccine efficacy against otitis media persists for years.

Thank you.

CHAIRMAN DAUM: Thank you very much, Dr. Kilpi.

We have a few moments for clarifying questions. Dr. Griffin.

DR. GRIFFIN: After the study was completed, did the parents and the physicians know who had received vaccine and who hadn't? I mean was the blind broken and they were informed as to whether they had been immunized?

DR. KILPI: Yes. The code was broken in August '99, and the parents were informed about the vaccine their child had received in October '99, and so I guess you are wondering if this fact may have affected the results we received after the completion of the trial, and we looked at this.

DR. GRIFFIN: You just wondered whether physicians say, "Oh, well, they were vaccinated. So they wouldn't need this"?

DR. KILPI: Yes, yes, and that's why we have looked at the incidence.

Yes, because many children completed the trial follow-up long before the code was open, some of them even had two years of follow-up after the code was revealed to the parents. So we looked at the incidence of tube placements after the completion of trial follow-up, but before unblinding, and this is the incidence in the PncCRM group as compared to the control group, and this is the total.

And this is for fully evaluated children and this is for all eligible children. So I think there is no sign that unblinding would have affected the results.

CHAIRMAN DAUM: Thank you.

Dr. Diaz, then Dr. Katz, and Dr. Schwartz.

DR. DIAZ: Dr. Griffin asked my question.

CHAIRMAN DAUM: Dr. Katz, please.

DR. KATZ: On the schedule of both groups, were they also receiving Haemophilus Influenza B conjugate vaccine at the same time? I don't mean necessarily the same visit, but this was part of their routine?

DR. KILPI: Yes, yes. The concomitantly given vaccine was also DTP Hib combination that they received at the age of two, four, six -- of six months, and we used two different DTP Hib combination vaccines.

DR. KATZ: I guess I wondered why you picked Hepatitis B as the control vaccine. What was the motivation for that?

DR. KILPI: Well, it's not included in the routine program in Finland. It's only recommended for risk groups, and it seemed to be the right thing to do to offer something to the control group also, something beneficial.

CHAIRMAN DAUM: Dr. Schwartz and Dr. Overturf and Stephens.

DR. SCHWARTZ: I'm confused or at least I don't understand.

CHAIRMAN DAUM: Turn you mic on. You push that button on the base.

DR. SCHWARTZ: Yes, sorry.

When you did tympanocentesis in that group of patients, whether they were on control or on the study vaccine, was the tympanocentesis 80 percent of all episodes or as close as you could get to every single episode on the study trial or after the first tympanocentesis that yielded a pneumococcal serotype of any serotype, then that child did not have to undergo further tympanocentesis and yet remain on the study?

DR. KILPI: No. It was the first one, in the first way. So whenever they had AOM diagnosed, myringotomy actually was the procedure we used. It made a small hole and suction. So it was performed every time AOM was diagnosed, at every single visit.

Of course, this was not 100 percent. It was saying from 93 percent of the visits when AOM was diagnosed.

DR. SCHWARTZ: So some children could have undergone six procedures or five procedures during this study?

DR. KILPI: I'm afraid so.

CHAIRMAN DAUM: Thank you.

Dr. Overturf and Dr. Stephens.

DR. OVERTURF: I wondered on the organisms that came from both the vaccine related serotypes as well as the organisms from the non-vaccine related serotypes whether you had any antibiotic susceptibility data on either one of those groups as compared perhaps to the serotype from the vaccine.

Do you have that data?

DR. KILPI: We do. We looked at -- what we have in the database is the data on penicillin resistance, but the resistance situation in Finland is very different from that in the U.S. So that almost all of them were susceptible to penicillin.

However, if they were not susceptible they were usually or I think they were exclusively vaccine serotypes.

CHAIRMAN DAUM: Thank you.

Dr. Stephens and then Dr. Decker.

DR. STEPHENS: Two questions. One had to do with the PCR count data. Can you give us a better understanding of that in terms of organisms per mL, presumably in terms of those counts.

DR. KILPI: I'm afraid I can't. As I told you, this method is semi-quantitative. We have now developed also using a better PCR method that allows quantification in a better way. This was just to demonstrate that obviously this huge difference tells us that it is the PCR negative case -- PCR positive, culture negative cases are something different from the culture positive case.

DR. STEPHENS: Okay. Can you also provide any information regarding the serotype replacement issue? That is, is there a difference between non-vaccine serotypes?

You gave us the data that there was a significant difference between vaccine serotypes. Is there an increase in non-vaccine serotypes in terms of carriage?

DR. KILPI: In terms of carriage? Yeah, well, I have some carriage data here.

(Pause in proceedings.)

DR. KILPI: So, well, this is first to show that the vaccine does not have effect on the overall carriage of pneumococcus. This is the other carriage figures at the age of 12 months, 18 months, and four to five years in the PncCRM group as compared to the control group. So always it's approximately the same proportion of children that are carriers.

And then this shows the carriage rates at 12 months of age, and actually there we did not see any statistical differences in these three categories. So there was perhaps a small reduction of the carriage of vaccine serotypes, but this is not statistically significant, and these are also pretty much the same.

This is different from the rate that is obtained in the developing countries. So the effect of the vaccine seems to be different. The effect of the vaccine on carriage seems to be different in developing country situations than in an industrialized country perhaps.

And here we have the carriage rates at the age of 18 months, and there is clear reduction in the carriage of vaccine serotypes. Cross-reactive serotypes are approximately even, and there is replacement by the non-vaccine related serotypes.

And when we come to the age of four to five years, again, we see the reduction in the carriage cell vaccine serotypes, and this time the situation for the other serotypes is even, but there is a small increase of the carriage of the cross-reactive serotypes in the PncCRM group as compared to the control group. However, these differences are not statistically significant.

CHAIRMAN DAUM: Thank you.

Dr. Decker.

DR. DECKER: No questions.

CHAIRMAN DAUM: Dr. Whitley.

DR. WHITLEY: This is an obvious question, and logically antibiotic usage would be lower in the vaccinated compared to the control population. Do you have data to support that logical assumption?

And specifically what I'm trying to get at is was there extraneous antibiotic usage in the vaccinated compared to the non-vaccinated group?

DR. KILPI: I don't have any slides to support that, but the number of antimicrobial prescriptions in the vaccine group was lower than in the control group, and I think it is covered in the FDA presentation.

CHAIRMAN DAUM: Okay. We have time for two more comments. Dr. Faggett.

DR. FAGGETT: Thank you. This is valuable clinical data.

Question number one, do you have national health insurance in Finland?

And Part 2 of my question: what were the criteria for tube placement? It would appear that with decreased costs and increased access that might impact on decisions to have the tube placement.

DR. KILPI: Yes, we do have national health insurance in Finland, and this basically means that the public sector is free of charge or the charge is only nominal, and for the private care, the children get reimbursed for the treatment.

So part of the sum is paid back, but anyway, the cost is considerably more to the parents than what would be in the public sector.

And the indications for tube placement, the recommended indications, I think, are pretty much the same as in the U.S. It's recurrent AOM, three to six episodes per six months or persistent otitis media with effusion.

But of course, as everyone knows, I think, that in a trial situation when it is really followed that this happens, it's different than if parents and doctors make individual decisions based on the waiting list and the financial situation of the family.

CHAIRMAN DAUM: And the age.

DR. KILPI: Yes.

CHAIRMAN DAUM: Dr. Glode.

DR. GLODE: I just wanted to clarify the original entry criteria. I know I read in the briefing materials that the public health nurse gave the vaccine and enrolled the patient in the study initially at two months of age or whatever; is that correct that that was generally done by public health nurses?

DR. KILPI: Well, they are public health nurses by training.

DR. GLODE: Yes.

DR. KILPI: They were trial staff. It's the policy in Finland that nurses vaccinate, and they were hired -- they were part of our staff team. So it was not their normal public health nurses, but it was a vaccinator we had hired for the trial.

DR. GLODE: Okay, and they knew which vaccine they were giving?

DR. KILPI: No. No, they didn't.

DR. GLODE: Okay. They were blinded.

DR. KILPI: Well, the vaccines were letter coded, and there was six letter codes for the three vaccines, and the vaccinator knew naturally which letter code the child received, and they were, therefore, kept separate from the other staff so that the staff didn't even know which letter code was assigned to each child, and this was never recorded anywhere.

DR. GLODE: Okay. Thank you.

CHAIRMAN DAUM: I'm going to take prerogative for the last question.

You showed some antibody data between the Prevnar vaccinees and the hepatitis vaccinees for several of the serotypes. Do you have similar data for Type 19F?

DR. KILPI: I showed for 19F.

CHAIRMAN DAUM: Did I miss it?

DR. KILPI: Yeah, it was --

CHAIRMAN DAUM: I'm sorry.

DR. KILPI: It was increasing also.

There.

CHAIRMAN DAUM: So I guess the question is then in light of the relative poor efficacy against that serotype, what do these kinds of data mean in terms of inferring protection?

DR. KILPI: Well, especially when it comes to 19F, it's very, very difficult to make any conclusions from the antibody concentrations and try to correlate to the efficacy. Obviously, these are antibody concentrations that look rather good anyway. Not a very good efficacy can be reached against otitis media.

CHAIRMAN DAUM: Thank you very much.

I think we now must move on to the next part of the sponsor's presentation, which would be Steve Black again to tell us about the Kaiser trial efficacy.

Thank you very much, Dr. Kilpi.

DR. BLACK: Thank you, Dr. Daum and everyone.

What I'd like to do now is present results on otitis media from the Kaiser Permanente efficacy trial. Otitis media, as well as pneumonia, were also outcomes apart from invasive disease there, and what I will show you are the results from our trial, which I think you'll agree are remarkably consistent with those presented from Finland.

To remind you, the pre-licensure trial was a randomized, double blind, controlled trial with one-to-one allocation, and the control vaccine used in this trial was meningococcal C conjugate vaccine, and as in Finland, immunizations were given at two, four, and six months of age, with a booster dose at 12 to 15 months. And these were given concomitantly with routine childhood vaccines.

The trial began in October of 1995 and was unblinded in April of 1999.

So otitis media outcomes were identified quite differently than in Finland. Diagnoses were made by the patient's regular pediatrician as part of routine care and in both the emergency room and in the clinic, and as part of routine care, optically scannable forms are used which capture out-patient diagnoses, and otitis media is one of these.

There's no cross-training of these observers.

Surgery for ear tube placement was captured as part of our hospital database, and this is exclusively performed in either surgical centers or hospital within our program, and spontaneously draining ears were cultured during the trial as well.

The primary endpoint was all otitis media episodes, and an episode was defined as a visit for otitis media with no prior visit within 21 days. It's important here to realize that if a child has an otitis media visit every 18 days, this can go on for months and still only count as one episode.

And this in retrospect was not such a great idea, but it does blunt the effect that we see for episodes against frequent disease or more severe disease.

The secondary endpoints were first otitis media episode, frequent otitis defined in Finland as three or more episodes within six months, four or more within 12 months; tympanostomy tube placement with spontaneously ruptured ears due to vaccine serotypes; and clinic visits for otitis media.

Just as a frame of reference here, which I find useful, is that for all episodes we estimate that 50 to 60 percent are bacterial. Of those, probably 40 percent in the United States are pneumococcal; 75 to 85 percent, as Dr. Siber showed, are vaccine serotype or cross-reacting. So that the potential overall impact of 100 percent efficacious vaccine against all clinical episodes of otitis media is in the range of eight to 20 percent.

Now, there are two data sets that were submitted vis-a-vis otitis media. Two analyses were performed. One was submitted as part of the PLA, included data through April 30th of 1998.

However, after that time, blighted immunization per protocol continued. The study nurses, the physicians, and the parents, the investigators were unblinded on April 20th of 1999, and there's a second analysis on otitis media there.

These compare these two analysis points just to give you an idea because the population dynamics change pretty dramatically during the year. You can see that the total number enrolled is about 17,000 in each group as of the initial analysis, and that enrollment was stopped in August of 1998. So the enrollment really had not progressed that much by past that point.

However, the number of booster doses is substantially higher in the second analysis, reflecting the fact that the children are now aging rather than just being enrolled, and the number over age two, there was substantial numbers in the initial analysis, but no children over age three. And you can see the number over age two in the second analysis is almost triple and that there are substantial numbers of children over age 3 in the second analysis.

This gives you an idea of the number of events. Otitis media, as has been pointed out, is much more common than invasive disease, which allows us to detect the efficacy that we did, and you can see here that in an intent to treat analysis, there are more than 116,000 visits for otitis media, almost 85,000 episodes, as compared to these numbers in the initial presentation.

Where this becomes especially important is for the less common outcomes here: frequent otitis media or especially for ear tubes where we have much more statistical power in the second analysis.

This is the protocol analysis first, and the two different analysis periods, and first you can see these are very similar between the two for otitis media episodes, the seven percent effect in 1998, 6.6 percent effect in 1999 for visits, 8.9 percent versus 7.9 percent, and for frequent otitis media, apart from the change in the number here, you can see that there's more precision or titer confidence interval as well, an 11.6 percent reduction for frequent otitis media in this population as of the final analysis.

This is the intent to treat analysis, and the numbers are a little bit lower. We were attempting, although the trial was not designed to do myringotomy because we couldn't get our pediatricians or ENT people to do that, we were able to come up with a surrogate outcome to look at vaccine serotype specific effect here, which was spontaneously ruptured tympanic membranes.

It's important to realize this is a different disease really than just acute otitis media, but nonetheless allows us to look at vaccine serotype specific effect, which is shown here. In the initial analysis these numbers were not statistically significant, but are here especially in the intent to treat analysis where we have 66 percent reduction of vaccine serotype spontaneously ruptured eardrums with a much tighter confidence interval, and these results are consistent with a vaccine serotype specific effect in Finland.

A question was brought up by Dr. Whitley regarding antibiotic use, and we did collect data on that. I'm glad we put this slide in sine you asked the question, and this shows -- the way this is groups, these are what was recommended for fist line antibiotic use in our program for otitis media, and you can see that constitutes the majority of the prescriptions.

There was a five percent reduction in that use in our population, really not surprising since otitis media is probably the most common cause of antibiotic use.

For second line drugs, which are shown here, Augmentin and all of these, there's about a ten or 11 percent reduction, basically extremely consistent with the frequent otitis reduction that we saw in the trial.

So the children who were going on to have more complicated or extensive otitis media, our interpretation is here that there is a reduction that's consonant with that in terms of antibiotic use.

There's an overall reduction here of 5.3 percent, and these drugs are still used for prophylaxis, for frequent otitis media, and we see a somewhat higher effect here as well.

So our conclusion from our studies is that Prevnar significantly reduced the risk of otitis media in our trial, and the efficacy was higher for frequent otitis and for ear tube placement.

So thank you, and I'd be happy to answer any questions.

CHAIRMAN DAUM: Thank you, Dr. Black.

We have a few moments for committee questions. Dr. Goldberg.

DR. GOLDBERG: Can I just ask a clarification? The otitis media endpoints are secondary endpoints from the original trial where the vaccine was approved for invasive disease. Can you just clarify for me in the original analysis plans and in the protocol were you thinking about using the same methods that you're using now?

Was that how the data was analyzed? And were there any adjustments made in any of these analyses for the '98 analysis on otitis media compared with the '99 one?

It's just for clarification, please.

DR. BLACK: Sure. The initial protocol specified otitis media episodes as the primary endpoint and also specified endpoints, other secondary endpoints as well.

The interim analysis in '98 was not really a decision point analysis. It was basically conducted at that point in time because we were analyzing the invasive disease at that point and to present that data, but we were not requesting a decision at that point for otitis media. So we did not apply a decision rule correction there.

CHAIRMAN DAUM: Thank you.

Dr. Parsonnet and then Dr. goldberg and Dr. Schwartz.

DR. PARSONNET: Yeah, I have a few questions. Can you just give me a sense of what the overall incidence of otitis was, annualized incidence in the two groups?

DR. BLACK: Yes. The average child in the control group -- and the numbers are very similar because of the differences there -- had about one and a half visits of otitis media per year, which is very consistent with national and published information.

DR. PARSONNET: And just along with that, I was just wondering if you have any sense for what the accuracy of the diagnosis was among your clinicians.

DR. BLACK: Well, you know, there were two approaches here that were taken to this outcome. One is the approach that was taken in Dr. Kilpi's trial, which is, you know, an extremely rigorous validation here.

What we were looking at in our trial was, to use Dr. Kilpi's phrase, the real world impact here, and we really didn't cross-train our observers. We think that that probably reduced the sensitivity of our finding because like in our setting as in others, the individual physicians have different criteria for otitis media, not all of them are assessing the mobility of the tympanic membrane. Some are just looking for redness.

So it's nonspecific, but it, we think, represents the real world picture as pediatricians. We don't think our pediatricians are really different from others for diagnosing otitis media in the rest of the country.

DR. PARSONNET: So then the last follow-up is actually related to that, which is the tympanostomy tubes are placed. Are they usually placed by pediatricians or are they placed by ENT docs.? Is that so you would be more likely to have a real accurate diagnosis in the tympanostomy?

DR. BLACK: Well, the tympanostomy tubes are all done in house in the hospital under general anesthesia by ENT physicians, and the rate of tube placement in our population is relatively low. It's about one percent, which is low even for this country.

But we know that all of those children -- you know, if you look at the average number of visits the children had prior to coming in for tube placement, it's between five and six.

Does that answer your question?

DR. PARSONNET: Yes.

CHAIRMAN DAUM: Dr. Goldberg, and then Dr. Schwartz and Diaz.

DR. GOLDBERG: Can I just get an additional clarification on your answer to the question that I asked?

In your original trial, it was designed for otitis and for invasive disease, correct?

DR. BLACK: Correct.

DR. GOLDBERG: Then you have two major endpoints as I see it. Sorry. I just want to make sure I --

DR. BLACK: Well, three. Actually pneumonia as well, but we're not talking about that.

DR. GOLDBERG: Well, okay. That even takes my question one step further then.

My question really is: did you at any point when you did -- was the original protocol written using the analysis methodology that you're using now?

And if so, was the invasive disease considered as one of those multiple endpoints?

DR. BLACK: Yeah.

DR. GOLDBERG: And what might the impact have been or --

DR. BLACK: Let me ask a statistician to address your questions.

DR. GOLDBERG: Thank you. It would just help clarify my thinking.

Thank you.

DR. BLACK: Bob Kohberger from Wyeth.

DR. KOHBERGER: The pre-specified plan before anything was unblinded, the first stage was invasive disease, which is tested at .05. If that was significant, we went on to the second stage, one of which was otitis media, all episodes.

If that was significant at .025, we then would go on to all those multiple secondary endpoints. So we adjusted for this multiple hypotheses.

In terms of the databases, the official database was 1998. We closed the database, cleaned it up, and that was what was submitted to FDA. The '99 data is primarily confirmatory of what we did in '98.

Does that answer your question?

DR. GOLDBERG: Had your original analysis plan included the one that you're using now?

DR. KOHBERGER: It's exactly the same.

DR. GOLDBERG: That's my question.

Okay. Thanks.

CHAIRMAN DAUM: Let's move on please to Dr. Schwartz.

DR. SCHWARTZ: I'll pass.

CHAIRMAN DAUM: Dr. Overturf or Dr. Diaz ‑- excuse me -- was next.

DR. DIAZ: Thank you.

Just a couple of questions in regards to the tube placement group of children. You commented that tube placement in your practices is lower than generally in other practices, and I was curious about several things.

One is the total numbers of children that we're talking about that went on to tube placement.

Secondly, if you have any data that looks at the timing for tube placement for children, i.e., prior to unblinding of the study.

And, thirdly, if the criteria for tube placement in children in the younger groups -- have you looked at any validation as to the use of criteria for tube placement across age groups?

DR. BLACK: Yeah, okay. There are several questions there. I'll see if I can remember to answer all of them, and if I don't, please remind me.

DR. DIAZ: Sure.

DR. BLACK: I think this gives you an idea as to the total number of events here. This slide shows you the number of children who had tube placement in the intent to treat and the protocol analysis at the two points in time.

So remember there are about 38,000 children in the population. So this is a little bit more than one percent, and this is about two percent here by the time the second year is added in.

You know, the criteria for using this, the pediatricians are free to refer to the ear, nose, and throat people for evaluation really at any time, but the tubes are normally put in if there is documented, persistent effusion with hearing loss or if there are multiple episodes, and the stated criteria are three or more within six months, four or more within a year.

But the average number actually that the children had was higher than that in this trial and in our practice in general.

DR. DIAZ: And also the differences or any data regarding tube placement prior to our after the unblinding of the study.

DR. BLACK: After the unblinding of the study in April of '99, we stopped following these children for tube placement, but we don't really have any reason to -- unlike the trial in Finland where separate study physicians were set up to evaluate the patients and there was a separate clinic, the children really were evaluated in standard care whether they were in the trial or not during the entire time period, and we would presume afterwards.

CHAIRMAN DAUM: Now, Dr. Overturn, thank you for being patient.

DR. OVERTURF: Steve, on your slide on the overall number of oral antibody prescriptions, I assume this was all antibodies or prescriptions, or was it only antibody prescriptions for otitis media?

And if so --

DR. BLACK: No, these are all antibiotics.

DR. OVERTURF: What proportion of oral antibiotic prescriptions are written for the indication for otitis media?

DR. BLACK: Yeah, we've actually not looked in this analysis. Our pharmacy for economic reasons has looked, and it's depending on age of the child. In the younger children, the two year old range, the toddler range, it's about 90 percent.

So the concordance here between antibiotic use and the otitis media effect is probably due to the fact that we're looking at the same thing in two different ways.

DR. OVERTURF: Do you know what proportion of otitis media patients received antibiotics?

DR. BLACK: That's something that's changing over time. Still the majority of them do receive that in the young age groups under two.

Over two the sort of watchful waiting is becoming increasingly more popular. I'm sorry I can't quantitate that for you.

CHAIRMAN DAUM: Ms. Fisher and then Dr. Stephens.

MS. FISHER: I just want to get this straight. In this study, all otitis episodes were reduced by seven percent in the Prevnar group, correct?

DR. BLACK: Correct, yes.

MS. FISHER: Well, your conclusion is that Prevnar significantly reduced the risk for otitis media, and as a parent taking my child in to be vaccinated, I'm trying to reconcile the seven percent reduction with the words "significantly reduce the risk."

DR. BLACK: Okay. You know, "significantly" is a word that has many meanings here, and I guess the statisticians for sure treat that differently than you or I might.

For the individual parent, the effect is not such that it's likely to be noticeable unless the child is one that has frequent or recurrent otitis or goes on to tube placement, in which case, you know, for a family of three or four kids you might expect to notice that.

But on a public health perspective, it is significant. I think as was pointed out, you know, a reduction of a million visits or more per year for otitis media is clearly a significant event as a public health effect, but I think it's fair to say for an individual parent, and it's important that parents realize that the individual parent is not going to notice the difference of an average of .3 otitis media visits over the course of the study, which is what we observed.

CHAIRMAN DAUM: In the strictest sense, you're using the term in the statistical sense, are you not?

DR. BLACK: I think it's to my mind -- I guess it's statistically significant, clearly, and I think from a public health perspective it is as well.

CHAIRMAN DAUM: Thank you.

Dr. Stephens and then Dr. Katz.

DR. STEPHENS: Just to clarify, and I realize the data is meager, but in those failures in the vaccinees, were most of those 19Fs or related serotypes?

DR. BLACK: All of them were 19Fs.

DR. STEPHENS: And what is that number total?

DR. STEPHENS: It's about 20.

CHAIRMAN DAUM: Dr. Katz.

DR. KATZ: These data may have been in the back-up material that I read, but I've forgotten, Steve. Do you have your youngsters broken down who was in day care and who was home dwelling?

DR. BLACK: We have that data from the telephone interviews that were conducted for safety. Day care is not a characteristic that's -- I mean, you can say whether they are -- I guess, rich or poor can change as well, but day care clearly can change. The status can change throughout the trial.

And if we look at the telephone interview data at any point in time, the day care participation rates are similar in the vaccine and control group, but we did not attempt to adjust for that in our analysis since the rates were the same.

We had done a case control study in preparation for this trial that showed that day care was the strongest predictor for risk factor for invasive disease, but that's been done by others as well.

DR. KATZ: But you don't have data to show that the reduction in the day care population was the same or greater or less than --

DR. BLACK: No, actually we've not looked at that, and I think, you know, with the number cases of invasive -- for otitis, you mean, or for --

DR. KATZ: Yes.

DR. BLACK: No, we have not done that. We'd have to really know what their day care status was at each point in time though. I think it's possible, but difficult.

CHAIRMAN DAUM: I'd like to move on at this point and hear from Dr. Siber, who will give a summary medical impact of Prevnar on AOM, and that will conclude the sponsor's presentation.

DR. SIBER: I'll be very brief and hopefully get us or keep us on time.

I think the main point I want to make about what we've just heard is remarkable consistency of two studies that were done in different countries, in different populations, under different epidemiological circumstances, probably differences in day care use, and so forth, and yet at least qualitatively, if not quantitatively, the results are remarkably consistent.

Overall, vaccine serotype OM had a 57 percent reduction in Finland with a reasonably narrow confidence interval. At Kaiser this was not a primary outcome, and a radically different disease, spontaneously draining ears, showed a 69 percent reduction in vaccine type isolates from ear tubes.

Only in Finland did we have data on vaccine related serotype OM and the related serotypes also showed a significant reduction at 51 percent with reasonably narrow confidence interval, and there was an increase with non-vaccine serotypes with a negative efficacy, as you've heard, of minus 33 percent.

Nevertheless, that increase was counterbalanced by the positive effects within that efficacy for the vaccine against all pneumococci, 33 percent or 34 percent with, again, a reasonably narrow confidence band.

For recurrent OM, somewhat different definitions. Kaiser and FinOM had similar efficacy, although only the Kaiser study was powered to have significance with regard to the recurrent OM at 12 percent reduction.

All otitis media, again, similar point estimates, but on the Kaiser study it was powered for significance against all otitis media, and with regard to tube placement, and I show here only the follow-up data for the reasons that I think Dr. Kilpi explained as being more relevant to general practice and a higher threshold for placing tubes.

Again, overlapping confidence intervals with 44 and 24 percent respectively, both significant and both with reasonable confidence intervals.

So let me briefly summarize again the health impact of otitis media and repeat, I think, what's been said before, that one to 1.4 million office visits are prevented each year by Prevnar at the current estimated efficacy rate based on the estimated pneumococcal disease rate or, rather, on the efficacy rate for all otitis media or at least what this is based on. So this is not a trivial public health issue.

We've also heard that there is a measurable decline in antibiotic use that corresponds roughly to the efficacy rate for otitis media, and we would expect in the future perhaps to actually see an impact of that on antibiotic resistance.

And the most important, serious complication of otitis media arguably is ear tube placement, and we calculate an estimated reduction in ear tube placement surgeries of about 60,000 in the United States, extrapolated from Kaiser.

So the otitis media indication, I think, is a rational thing to have as part of this vaccine indication and to be described in the insert. It's now supported by two randomized controlled trials that show statistically significant decreases of OM outcomes, as I've mentioned earlier.

We've seen that there's important medical effects on otitis media disease and its consequences. I think insuring that accurate information is present in the label that informs the significant, but modest effect on otitis media is important so that physicians and parents receive the most accurate possible information.

Let me make one final point that I think is important with regard to otitis media indication for vaccines. We and other manufacturers have programs directed towards other pathogens, bacterial and viral that cause otitis media, and although such vaccines might have high efficacy against their particular pathogen, they nevertheless, if otitis media itself is used as a standard against which they will be measured, will necessarily have a low overall impact on otitis media because so many pathogens are involved.

So to use a traditional standard of 80 percent, 90 percent efficacy that we used to with vaccines with an outcome like otitis media that is probably microbial would pose, I think a very difficult dilemma for the development of Moraxella catarrhalis or non-typeable Haemophilus or some of the viral pathogens that cause otitis media.

So I would want to ask the committee to consider that in their deliberations about this issue of the low efficacy overall for otitis media, and that's it.

CHAIRMAN DAUM: Thank you very much, Dr. Siber.

Are there committee questions or comments that go toward clarification of the sponsor's presentation?

Ms. Fisher.

MS. FISHER: You said that there is a five percent reduction in antibiotic use in these trials with the use of Prevnar, correct?

Okay. So you're saying there's going to be an associated decrease in antibiotic use if this indication is forthcoming. Five percent is not a lot, is it, in terms of decrease antibiotic use?

DR. SIBER: In terms of the total number of prescriptions written, it certainly is a large number. Obviously five percent is five percent.

CHAIRMAN DAUM: Okay. Thank you very much.

That, I think, concludes the sponsor's presentation. I thank all of the speakers and committee questions. I think at this moment we're doing very well time-wise. We will take a ten-minute break, 12-minute break and reassemble at 10:35 Eastern time.

(Whereupon, the foregoing matter went off the record at 10:26 a.m. and went back on the record at 10:41 a.m.)

CHAIRMAN DAUM: We will now begin the FDA presentation regarding acute otitis media and Prevnar, and our first speaker will be Douglas Pratt.

DR. PRATT: Good morning. First I'd like to recognize other members of the review team:

Jingyee Kou from Biostatistics;

Marion Gruber from the Division of Vaccine Applications; and

Carl Frasch from the Division of Bacterial Products.

I also see Pam Getson in the audience. She was a biostatistician with FDA to left us recently. She was involved in many of the early discussions on otitis media.

Well, Prevnar was licensed in the U.S. in February of 2000 for prevention of invasive disease caused by the seven pneumococcal serotypes represented in the vaccine. With this supplement to the license application, Wyeth Lederle seeks to extend the approved application to include prevention of otitis media.

Specifically, regulatory approval has been requested to market Prevnar for active immunization of infants and toddlers against invasive disease and otitis media cause by streptococcus pneumonia due to capsular serotypes included in the vaccine.

Some regulatory background is summarized here. VRBPAC met to deliberate approval recommendations for invasive disease in November of 1999, and at that meeting some data relating to acute otitis media were presented.

However, the committee was not asked to consider approval of an indication for otitis media at that time. The license application for acute otitis media was submitted in June of 2000, and following an FDA review, a letter was sent to the sponsor in May of 2001 requesting additional analyses, clarifications, and other information.

The sponsor responded in October of 2001, and then another FDA letter was sent to the sponsor in March of 2002 after the sponsor had requested that FDA consider additional data from the Finnish follow-up study, which you have seen some of this morning.

And that brings us up to date.

Well, Prevnar is currently the only licensed pneumococcal conjugate vaccine, and Prevnar is recommended for all children under two years of age and for some older children who are at high risk for invasive pneumococcal disease.

Extending the licensed indication to prevention of acute otitis media appears unlikely to impact use of Prevnar in the U.S. in the near future. However, FDA views consideration and discussion of this application by the committee today appropriate for a number of reasons, including those represented here.

Efficacy estimates for acute otitis media outcomes are comparatively low for preventive vaccines. Also, as was mentioned this morning, there's the possibility of increased risk of acute otitis media or negative efficacy for pneumococcal serotypes not included in Prevnar.

And also, concerns have been expressed in the medical community about the potential for unrealistic public expectations. Following the publication of the Finnish otitis media study by Eskola, et al., a number of letters to the editor regarding that article were submitted to the New England Journal of Medicine, and some of those opinions are paraphrased here.

The overall clinical significance of the treatment effect was questioned. Concerns were expressed about the limited overall benefit. The overall benefit may be misunderstood by the public, and there was concern that the existing recommendations for its use may be compromised.

There was also one letter that incorrectly stated that FDA had rejected use of Prevnar for this indication.

Well, given the global issues and the concerns expressed in the medical community, we thought that an open public discussion of these data and these issues was warranted.

Well, data intended to support the intended indication have been provided from two well controlled clinical trials, the Finnish otitis media trial and the Northern California Kaiser Permanente trial. And some additional efficacy data from extended follow-up from each of these trials has also been provided.

This table reviews some of the important differences between the two studies. The Kaiser study was much larger than the Finnish study. The control vaccines differed. The interval separating new episodes differed, 30 days in the Finnish study, 21 days in the Kaiser study.

And the case definition for the primary endpoint in the Finnish study was based on bacterial cultures, and in the Kaiser study it was based on automated data searches for AOM visits.

And there were different primary regulatory objectives for these two studies as well.

Also of note, the pre-licensure formulations of Prevnar were abbreviated differently in the two studies. In the Finnish study it was abbreviated PncCRM and in the Kaiser study 7VPnC. In many of the tables that follow those abbreviations will be used, but for the oral presentation, I'll try to refer to the pre-licensure formulation simply as Prevnar.

Well, with that introduction, I'll move on and again review efficacy data from the Finnish study, including supplementary analysis requested by FDA, as well as some of the data from the follow-up study, the Finnish follow-up study, and then go on to discuss efficacy data from the Kaiser study.

Much of this information will be repetitious from the sponsor's presentation. It's the nature of going second in these meetings, but there will be some FDA comments for emphasis on some of the analyses, and for those of you less familiar with the data, this may be helpful.

There will be a brief discussion of safety data that will be limited to clinical trial data from the Finnish study, and then there will be some considerations for the committee to think about in their deliberations before presenting the questions to the committee.

Primary objective in the Finnish study was to determine protective efficacy of the pneumococcal conjugate vaccines against culture confirmed pneumococcal acute otitis media due to vaccine serotypes.

Secondary objectives were to determine efficacy used in different levels of diagnoses, efficacy in preventing nasopharyngeal carriage, determining the antibody response, as well as the safety and tolerability.

In the Finnish study, subjects were randomized equally to one of three vaccines, Prevnar, PnbcOMP manufactured by Merck, and the Hepatitis B vaccine control.

However, only data related to Prevnar were provided in the application and only data related to Prevnar will be discussed today.

The study was double blind, and eligible subjects were in good health as determined by medical history, exam, and clinical judgment.

Of note, infants born prematurely could be enrolled in the study if they were judged to be in good health.

Children received Prevnar or the control Hepatitis B vaccine at two, four, six, and 12 months of age, and this coincides with the U.S. schedule fir Prevnar. Vaccines administered concurrently with study vaccines were DTP Hib combination vaccines for the first three doses, and these did contain the whole cell pertussis components.

And then IPV, the second dose of IPV was the only concurrently administered vaccine at the 12 month visit.

Dr. Kilpi talked about case surveillance and ascertainment cases were identified through the study clinics which also provided the well child care. Clinics were open every day of the week, and parents were encouraged to bring their children to the study for respiratory infections or symptoms suggesting acute otitis media.

And if Strep. pneumoniae was found -- excuse me -- myringotomy with aspiration of middle ear fluid for culture was done. If clinical acute otitis media was diagnoses and if Strep. pneumoniae was found, then the serotype was determined, and each child was followed until age two years.

The clinical definition that Dr. Kilpi talked about, it included clinical criteria being a visually abnormal tympanic membrane, suggesting an effusion, and at least one sign of disease, including fever, ear pain, irritability, diarrhea, vomiting, and acute otorrhea or other symptoms of respiratory infections, and this definition appears to be consistent with U.S. clinical practice.

The primary endpoint in the study, as was discussed, was AOM episodes due to vaccine serotypes. There was one secondary endpoint pre-specified. That was first and subsequent AOM episodes due to vaccine serotypes and other endpoints were pre-specified, including AOM due to vaccine serotypes by dose; all pneumococcal AOM episodes regardless of serotype, and that included culture over PCR; all AOM episodes with middle ear effusion regardless of etiology; and all AOM episodes regardless of etiology, whether or not middle ear fluid was obtained; and then children with recurrent AOM.

The definition of the primary endpoint is that a new episode was considered to start if at least 30 days had elapsed since the beginning of the previous AOM episode due to the same serotype or any interval for different serotypes, and these had to be culture confirmed.

This screen shows graphically a hypothetical example of the counting process for the primary endpoint. Four numbered episodes of vaccine serotypes are shown. Vaccine 19 -- that should be 19F -- accounts for the first and the second episode because they're separated by 30 days.

Vaccine serotype 23 accounts for the third episode, even though 30 days has not elapsed because it was due to a different serotype.

Then a positive PCR, a non-vaccine serotype 6A, and acute otitis media with middle ear effusion, they did not contribute to the primary endpoint.

And then the fourth episode here was due to vaccine serotype 6B.

Well, it's unusual for preventive vaccine studies that a subject contributes more than one case to the analysis of efficacy. In fact, we could think of no other example of a licensed vaccine for which efficacy was determined using these repeated measures.

A similar analysis was conducted for the primary endpoint in the Kaiser study. The analysis plan was discussed with FDA and did receive FDA concurrence prior to unblinding, but because this statistical approach is somewhat unusual, it's worth describing a little bit, as well as the underlying assumptions.

The analysis used to generalized Cox regression model with Anderson-Gill counting method and risk of acute otitis media was estimated piece-wise, that is, from event to event. The model assumes proportional hazards between groups over time and robust variance estimates were used to compensate for interdependency of events within subjects, and this interdependency was well recognized by all involved. And the analysis is said to provide an average vaccine effect on AOM episodes.

Well, alternatives to these measures would discard some of the data, some or much of the data, but some of these alternative analyses will be shown, will be discussed, and, in fact, FDA looked for multiple checks on the data because of this somewhat unusual approach for a preventive vaccine.

Per protocol follow-up in the Finnish study began two weeks after the third dose. The intent to treat follow-up began at the time of the first dose.

In general, FDA expects to see intent to treat analyses in addition to protocol analyses, and in most of the tables that follow, both per protocol and intent to treat analyses are shown.

Well, getting into the results, information was collected on demographic variables and some characteristics known to be associated with increased risk of acute otitis media. Despite randomization, some imbalances between treatment groups at study entry were observed after unblinding, and here three selected population characteristics are shown, premature gestational age, low birth weight, and prior acute otitis media episodes at the time of enrollment.

All of these showed a slight imbalance with more -- excuse me. I think this backwards. In any case, there were some imbalances that were noted, and because of the direction of some of the imbalances and the fact that multiple events were counted for individuals, there was a potential that these might influence results, influence the efficacy estimates. So we requested some additional analyses of these endpoints.

Birth weight -- I'm sorry. I wonder if the sponsor can help me right here. These are reversed; is that correct? I think that actually the low gestational age, low birth weight, and prior AOM episodes were actually increased in the Prevnar arm. That's why that they were presented.

Okay. Well, this table shows results of the protocol defined primary analysis AOM episodes due to vaccine serotypes. During protocol follow-up the vaccine efficacy estimate was 57 percent; a lower bound of 44 percent.

The intent to treat estimate was 54 percent, with a lower bound of 41 percent. These estimates were statistically significant, and statistical significance at the five percent level can be inferred here and in the subsequent tables if the 95 percent confidence interval excludes zero. P values will not be shown in most of this presentation.

With a contribution of each of the vaccine serotypes to efficacy as measured by the primary endpoint, as shown here, for intent to treat follow-up the most common vaccine serotypes were 23F, 19F, 6B, and 14. Statistical significance was demonstrated for the individual serotype 6B, 14, 18C, and 23F.

The lowest efficacy estimate was for serotype 19F, ten percent, and the intent to treat analysis, but this was not statistically significant.

There were few episodes for serotype four or 9V.

The protocol defined secondary endpoint examined first and subsequent episodes of AOM due to vaccine serotypes. This analysis would count only one -- excuse me -- the first episode analysis would count only one episode per subject and take into account time to first event.

Efficacy estimate for prevention of first episode was 52 percent in protocol, 45 percent in the intent to treat analyses, and these were statistically significant.

Subsequent episodes were also statistically significant, 45 percent per protocol, 49 percent in the intent to treat.

It's also clear from this slide that most of the episodes were first episodes comparing, say, for the Hepatitis B group 177 to 73 or 89 to 18. Most of the episodes were, in fact, first episodes.

The efficacy estimate for culture confirmed to pneumococcal AOM, regardless of serotype, was 34 percent in the protocol analysis, and this was statistically significant. Results of intent to treat analysis were similar.

Although not specified in the protocol as the primary or secondary endpoint, FDA viewed this endpoint as very important in addressing the clinical significance of the vaccine in preventing otitis media.

Analysis of pneumococcal AOM as determined by PCR was specified in the analysis plan. However, the PCR data was not available at the time the amendment was submitted. These data were submitted on FDA request during the review period.

I'll go ahead and talk about some of the exploratory endpoints here. The efficacy for prevention of pneumococcal serotypes belonging to the same sero groups taken collectively was also statistically significant, 51 percent in the protocol analysis, 44 in the intent to treat, both statistically significant.

And when examined by the individual serotypes, serotype 6A, although not a vaccine serotype was associated with a substantial number of cases and, in fact, was statistically significant.

As was mentioned earlier, serotype 19A, related to the vaccine serotype 19F, actually had a slightly higher efficacy estimate, 21 percent, than was observed for 19F.

These are intent to treat analyses. I think this morning the sponsor showed the protocol analyses for these serotypes.

Again, looking at other than vaccine related serotypes, there was a negative vaccine efficacy estimate, minus 34 percent in the protocol, minus 39 percent in the intent to treat. For protocol this was borderline statistically significant, and the intent to treat, in fact, was statistically significant.

Thus, subjects vaccinated with Prevnar were actually at increased risk of getting AOM due to one of the vaccine unrelated pneumococcal serotypes. The most common unrelated serotypes belong to serogroups three, 11, 15, and 35. If this effect were to occur with widespread vaccine use, one might expect to observe replacement vaccine serotypes with non-vaccine serotypes in the general population as causes of otitis media.

Recurrent otitis media was defined as three episodes within six months or four episodes within 12 months. AOM episodes for this endpoint were due to any cause, whether pneumococcal or not.

The efficacy estimate here in the per protocol analysis was 16 percent. It was nine in the intent to treat analyses. Neither of these were statistically significant. However, demonstration of efficacy for this endpoint was not a primary objective, and the study was not powered to demonstrate efficacy for that endpoint.

Other planned analyses included AOM with middle ear effusion and all cause AOM regardless of etiology.

The efficacy estimate for AOM regardless of etiology was six percent in the per protocol analysis, four percent in the intent to treat analyses. Neither of these were statistically significant, but again, the studies were not powered for these outcomes.

It is noteworthy that the six percent estimate is actually quite close to the estimate that was obtained in the Kaiser study for a similar outcome.

Well, nasopharyngeal carriage was assessed as a secondary objective at two time points in the Finnish study, at 12 months and at 18 months. At 12 months the carriage rate of vaccine serotypes was reduced 17 percent. That difference was not statistically significant.

However, at 18 months carriage was reduced 41 percent, and that estimate was statistically significant. But for this table relative risk estimates are shown rather than difference estimates, which were in the application and the study report and also in the briefing document. This is to be more consistent with the other efficacy analyses that have been shown.

Of note, the sponsor has not proposed including efficacy data for carriage in the label with this amendment.

A serology cohort was comprised of 115 children enrolled at one center. The serology cohort for the two treatment groups appear to be well balanced for demographic characteristics. The geometric mean concentration serum antibody to type specific pneumococcal polysaccharides as determined by ELISA are summarized here on this screen. Confidence intervals are omitted for simplicity of presentation.

As can be seen, there's substantial increases in antibody concentrations over control were observed for each type and then going from the third dose to the fourth dose for Prevnar, increases were seen for each of the seven types.

This screen shows serotype specific efficacy estimates from the primary analysis along with the GMCs that were just shown. It's worth noting that although efficacy estimate for serotype 19F was the lowest of the seven serotypes, 25 percent of the per protocol analysis antibody responses were actually comparable to the other serotypes, both after dose three and after dose four.

The highest efficacy estimate was for serotype 6B. However, that had one of the lowest ELISA GMCs after the third dose, though it appeared to have a good boosting response for the fourth dose.

So it appears that antibody levels as determined by ELISA do not appear to provide any insight regarding efficacy by serotype.

There were a few cases of invasive disease due to pneumococcus in the Finnish study. I compiled this table of the four episodes. There was only one in the Prevnar group. This was due to a type not included in the vaccine.

There were two vaccine serotypes, 23F and 19F, in the Hepatitis B control arm.

I'll now discuss some issues that were identified during the review and present some supplementary analyses that were conducted upon FDA request.

As noted earlier, despite randomization there were some imbalances between treatment groups with respect to certain risk factors for otitis media that were observed after unblinding.

To determine whether the imperfect distribution of these risk factors between the two groups would have a major effect on efficacy estimates, covariate adjusted analyses were performed by the sponsor on FDA request.

The supplementary analysis was not part of the pre-unblinding analysis plan. So the effect of gender, AOM history prior to enrollment, and day care attendance on the number of AOM episodes was, in fact, highly significant. However, the interaction between these variables and the vaccine effect was not.

Similarly, no significant interactions were seen between vaccine effect and gestational age, birth weight, breast feeding, or household smoking. And as shown in this table here, all of the adjusted efficacy estimates were similar to the unadjusted estimates. Fifty-four percent, this is the intent to treat analysis. Whether adjusted, they were 54 percent unadjusted, 54 percent, 32 percent, the same.

Actually the adjusted estimate regardless of etiology actually went up a little bit. So these analyses were reassuring in that the observed imbalances for known risk factors were unlikely to affect the outcomes.

Well, it was apparent from examining the culture results from individual subjects that some subjects contributed multiple episodes of the same serotype to the analysis. This screen shows an example of subjects from the control group for whom serotype 23F was isolated on multiple occasions extending over nearly a year.

This subject actually contributed four cases or four episodes to the analysis of the primary endpoint, as the first three episodes here were all within the 30-day window and collectively accounted for one episode.

Here's another example from the Prevnar group, actually two examples. Subject 1450 contributed three episodes of 23F, the vaccine serotype, to the vaccine serotype analysis, and the non-vaccine serotype 15 for subject 2241 contributed three episodes to the analysis of all pneumococcal regardless of serotype.

Well, to assess the effects of these counting multiple episodes per subject on the analysis of the primary endpoint, FDA requested supplementary analyses in which each serotype could be counted no more than once per subject.

This table shows the supplementary analysis for the primary endpoint conducted after unblinding as requested by FDA. The efficacy estimate determined after exclusion of subsequent episodes was 55 percent in the per protocol analysis versus 57 percent in the original analysis plan. The confidence intervals also remained fairly narrow.

So excluding subsequent episodes due to the same serotype from the analysis did not appear to affect the efficacy estimate substantially, and this provided a check, if you will, on the analysis of recurrent events.

Well, a similar analysis was conducted for the endpoint of all AOM episodes due to pneumococcus regardless of serotype, again, excluding the same episode if it occurred more than once in a subject, and here, again, the efficacy estimates were identical, 34 percent, in the per protocol analyses with nearly identical confidence intervals as well.

So, again, these were reassuring with respect to the effect of multiple counting.

Analyses using a case definition based on identification of pneumococci by PCR was specified in the study protocol, but these were not available at the time the study report was written and were not provided with the application.

These were provided during the review period on FDA request.

The PCR assay detects the pneumolysin gene, a gene common to all Strep. pneumo., but it does not distinguish among the serotypes.

In the per protocol analysis of efficacy based on PCR the efficacy estimates were somewhat lower, 20 percent per protocol versus 34 percent, 18 percent intent to treat versus 32 percent by culture, and the efficacy estimates were quite wide.

PCR confirmation contributed actually to a substantial number of new cases. Compare per protocol of Hepatitis B, 414 by culture, 678 by PCR. We saw analyses this morning looking at quantitative PCR. Those data were not in the submission. We had not seen those data before. I think they're interesting. We really can't comment on those data.

But in any case, although the efficacy estimates were lower by PCR, they remain statistically significant.

The clinical significance of the positive PCR and the culture negative is not clear at this time.

A question was asked about antibiotic use this morning, and in fact, we had the same question. Antibiotic usage was not included among the prospectively defined study outcomes, and no analyses were provided with the application.

However, data was recorded on the case report forms during the course of the study. Clearly patterns of antibiotic use could impact on the acute otitis media outcomes.

If use of prophylactic antibiotics were significantly greater in the Prevnar group than in the control group, then some of the apparent treatment effect might be due to the prophylactic antibiotics. So in any case, FDA requested that these data be compiled and analyzed and submitted.

And as shown, the number of subjects receiving antibiotics for treatment was less in the Prevnar group, and this approach reached statistical significance at the .05 level.

The number of subjects receiving antibiotics for prophylaxis and regardless of purpose were also nominally smaller in the Prevnar group.

Taken together, these data relating to antibiotic use during the Finnish study are consistent with a vaccine effect in prevention of acute otitis media.

Information about tympanostomy tube placement during the Finnish study was recorded on the case report forms as well. During the course of the study, these data were not with the initial applications. FDA requested that these data be provided to examine consistency of effect with the Kaiser study for first tympanostomy tube placement.

And we also got information that the recommendations regarding ear tube placement in Finland were similar to U.S. practice.

As shown here, actually the rates of first ear tube placement, number of subjects with events in this table were quite similar and no efficacy estimate was provided. It was suggested that because of the close follow-up during the study that subjects actually sought treatment with ear tubes more often than would ordinarily be the case in Finland, and these rates actually were higher, I think, tenfold higher, nearly tenfold higher than common practice in Finland and also much higher than practice in the Kaiser system.

Well, subsequently long-term follow-up data from the Finnish study became available, and in February of 2002, the sponsor proposed an analysis plan of these follow-up data with inclusion to the licensed application.

In the follow-up study, all eligible children were now four to five years of age at the planned follow-up visit. Parents and investigators were unblinded to treatment assignments at this time.

Now, two populations were evaluated, and the first population included volunteers to the follow-up study. They participated in parental interview for otitis media history, an ear exam, and then records of procedures were verified through the hospital or private physician's records.

The primary analysis of the follow-up data was based on this volunteer population. Then a secondary analysis was performed on the original cohort that remained available for follow-up in the area.

In these analyses, in contrast to what was seen in Kaiser, this is all tympanostomy tube placement, not just the first event.

So in the primary analysis after this follow-up cohort, a total of 756 or about 46 percent of the original 1,662 randomized children enrolled and completed the assessments. The efficacy estimate for ear tube placement for this population during the efficacy study was 12 percent, and this was now statistically significant. That is the efficacy for the period two months to two years during the original trial.

And in the long-term follow-up from two years to four to five years of age, the efficacy estimate was 39 percent. This was statistically significant, though with fairly wide confidence intervals.

In evaluating this result, I think it's important to note that this was a self-selected subgroup of volunteers. Enrollment was not even between the two groups, 353 versus 403.

Also, these children were more otitis prone than the entire study population. That's not easy to see, but the rates were actually increased in this population over the larger cohort. And then, again, the follow-up was not blinded.

It's questionable whether these data actually qualify as an adequate and well controlled trial by the regulatory definition. However, I think they can be viewed as supportive for the other study, for looking at consistency on the ear tube placement effect.

Well, this is the results of the secondary analysis from the follow-up study. Again, here all records were confirmed by checking the hospital records. There was no volunteerism involved here. Everyone that was available that was followed.

Again, this population, during the study itself, two months to two years' follow-up. The efficacy estimate was four percent. That was not statistically significant, but the long-term follow-up, two years to five years, estimate was 44 percent, and this was statistically significant.

I'll move on now and go over the Northern California Kaiser Permanente otitis media efficacy results. I'll go quickly over much of this that has been discussed this morning, and it's probably fairly clear to everyone now.

The study was randomized and double blind. The control vaccine was an investigational meningococcal C conjugate vaccine. Evaluation of AOM was actually a secondary objective, as was discussed this morning. There was no standardized clinical case definition, and tympanocentesis and routine culture of middle ear fluid was not done. Rather, cases were identified through automated database searches to identify the diagnoses.

A diagnosis was based on routine clinical practice using a check-off box on the patient encounter form. An AOM episode, a new episode began if at least 21 days had elapsed. This is somewhat shorter than in the Finnish study. And frequent acute otitis media was defined as three AOM episodes within six months or four episodes within 12 months.

The primary objective was looking at reduction in all AOM episodes. Secondary outcomes that were pre-specified included first episode, frequent AOM, first tympanostomy tube, all AOM clinic visits, and ruptured eardrums.

The primary endpoint is summarized here. Again, per protocol, overall reduction in AOM episodes was seven percent per protocol, and in the intent to treat was 6.4 percent.

The intent to treat analysis here actually includes substantially more subjects. You can see from 16,000 to 25,000, and the reason for that is that there was differential follow-up. Not all of the subjects had received the full three doses at the time that the study code was unbroken.

This is one of the secondary analyses, risk of first episode or at least one episode. Due to the shorter, 21-day interval between new episodes in the Kaiser study, it's possible that some over counting might occur if some episodes were slow to resolve.

Also, using the patient encounter form, a follow-up visit might not be easily distinguished from a visit for a new episode.

Well, one check on the possibility that the definition used might over count or otherwise inflate the outcome would be to look at one episode per subject, and that is captured in this analysis. Here the per protocol analysis of first episode, reduction was 5.4 percent, and in the intent to treat, it was 4.9 percent. Both of these were statistically significant.

For the analysis of frequent acute otitis media, that vaccine efficacy estimates in preventing frequent were 9.5 percent in the per protocol analysis, 9.2 percent in the intent to treat analysis, and these were also statistically significant.

First, tympanostomy tube placement, again reduced in the per protocol analysis by 20 percent, intent to treat analysis by 21 percent. These confidence intervals are fairly wide. Nevertheless, the results were statistically significant.

Thirteen ruptured eardrums, culture positive for vaccine serotypes were observed during per protocol follow-up, four in the Prevnar group, nine in the control group. The efficacy estimate was 56 percent and 57 percent in the intent to treat analyses. Neither of these were statistically significant.

When looking at all pneumococcal serotypes, per protocol estimates for reduction was 62 percent, not significant, but for intent to treat, 61 percent was statistically significant.

Well, vaccine serotype 19F and related serotype 19A accounted for all of the serotypes from the ruptured eardrums in the Prevnar group, and 39 percent of those from the control group.

Taken together, vaccine serotypes accounted for 20 out of the 25 isolates or 80 percent of all of the isolates from ruptured eardrums, all of the pneumococcal isolates.

Extended follow-up data for acute otitis media accumulated after breaking the treatment codes for about another year before parents and clinicians were informed of the treatment assignments and Prevnar was offered to the control group.

This table compares the efficacy estimates at the time of the primary analysis for data where the database was closed on April 30th, 1998, and then the additional follow-up data to April of '99.

All of the efficacy estimates were similar, and the confidence intervals actually became more narrow for many of the outcomes.

I'll now talk a little bit about safety data from the Finnish otitis media study. The Finnish study does contribute new controlled safety data to the safety database for Prevnar. These data were reviewed in the otitis media amendment, and the briefing document contains a more full discussion of the safety data.

The relevance and usefulness of these data to the U.S. population are somewhat limited because, for one reason, the wholesale pertussis containing DTaP Hib combination was administered with the first three doses rather than DTaP vaccine, which is common practice in the U.S. now, and this can complicate some of the assessments of systematic reactions.

Also, the Finnish population does not reflect the heterogeneity of the U.S. population, and also the study was really not large enough to detect uncommon adverse events.

However, parent compliance with report of vaccine reactions was nearly complete in the Finnish study, and so reported data are probably reliable. And these data do confirm an incremental increased risk of fever after each of the three doses, low grade fever after each of the first three doses and also an increased risk in higher grade fever after the third dose.

However, the frequency of high grade fever never exceeded two percent for any of the doses. Also, increased crying was observed after each of the three doses.

This table shows data after the fourth dose. Here the concurrent immunization was the second dose of IPV. Again, low grade fever was statistically more frequent. High grade fever, there was no difference.

So overall the safety data from the Finnish otitis media are consistent with earlier observations regarding the safety of Prevnar. As had been previously observed, Prevnar was associated with increased fever, increased low grade fever, but complications of fever were uncommon.

In fact, there were no febrile seizures temporally associated with administration of either vaccine in this study.

The committee will not be asked to comment on safety at this meeting today. Prevnar is now in wide use. Large post licensure safety studies are underway.

Vaccine labels can be updated at any time with important safety information, and information from the ongoing post marketing study conducted at Kaiser identifying any new safety concerns that better quantify known or suspected adverse events, the label will be updated accordingly.

Well, before presenting the questions, I'd like to show a few screens with things for the committee to consider in their deliberations. First, there really is no guidance from regulations or other published documents which specifically address the minimum level of efficacy required for licensure of a preventive vaccine.

Approval of most drugs and other therapeutic products are based on substantial evidence of efficacy from at least two well controlled clinical studies.

Substantial evidence for drugs and therapeutic products is commonly determined by hypothesis testing using an acceptable level of statistical significance, usually less than .05.

For preventive vaccines, much more emphasis is placed on the efficacy estimates and the 95 percent confidence limits. For most preventive vaccines, efficacy estimates are relatively high, and confidence intervals are substantially above zero.

Well, decisions about the approval of Prevnar for prevention of acute otitis media could have implications for the approval of other pneumococcal vaccines. Current thinking at the Office of Vaccines is that an indication for prevention of otitis media should stand on its own because license applications for other new pneumococcal vaccines for prevention of acute otitis media may not include evidence for prevention of invasive disease.

Approval of an otitis media indication for Prevnar would set a precedent for the level of efficacy, the type of data, and the preferred endpoints for licensure of other pneumococcal vaccines whether or not license applications for these other products include evidence of efficacy for prevention of invasive disease.

One possible scenario that Dr. Siber also discussed this morning is that it's possible that a different vaccine could prevent more acute otitis media episodes, but still have less vaccine serotype specific efficacy. That's a possible scenario.

Well, if evidence of efficacy in preventing acute otitis media is judged adequate, then a new indication that describes the treatment effect of Prevnar regarding acute otitis media will be included in the label indication.

The sponsor proposes an indication for prevention of AOM limited to serotypes represented in Prevnar. Some considerations here include that that indication would reflect the primary endpoint in the Finnish study, but not the primary endpoint in the Kaiser study.

Also, focusing on serotype specific efficacy does not capture the treatment effect for vaccine related serotypes, nor the negative efficacy for unrelated pneumococcal serotypes.

Efficacy estimates were relatively low for some of the outcomes and confidence intervals were relatively wide for some of the outcomes.

Well, in the risk-benefit assessment of product approvals by FDA, substantial evidence of clinical benefit must be provided from well controlled studies. A decrease in office visits for otitis media and potential cost savings of medical care related to otitis media have been cited by the sponsor and in the literature among the benefits of Prevnar related to otitis media.

To the extent that office visits reflect patient disease, they may be considered in the assessment of clinical benefit for the basis of a regulatory decision. However, economic benefit is not considered in the efficacy evaluation by FDA.

Approval of an indication for prevention of otitis media would normally allow the manufacturer to distribute marketing materials promoting use of the product based on information included in the approved labeling. Concerns have been expressed about unrealistic public expectations regarding the effect of Prevnar and acute otitis media.

The potential for misleading public expectations with marketing materials exists for all products and all approved indications. However, FDA is empowered to restrict unrealistic marketing claims. Regulations require that advertisements and promotional labeling be submitted to Siber for review, and any advertisements that are judged false, lacking in fair balance can result in a product being misbranded and then multiple corrective actions are then possible.

That concludes my presentation.

CHAIRMAN DAUM: Are you going to have a statistical presentation also about --

DR. PRATT: There won't be a separate statistical presentation. However, we're happy to take questions relating to the statistics.

CHAIRMAN DAUM: Okay. Then I think the way we'll proceed is that we'll open the floor. We'll thank you, Dr. Pratt. Then we'll open the floor to the committee for questions and comments on your presentation with respect to clarifying what was said.

Ms. Fisher.

MS. FISHER: That was an excellent presentation that we can always expect from you.

I have a question. What is your definition of substantial evidence of clinical benefit? Does that mean clinical benefit to the individual child?

DR. PRATT: I think we mean by clinical benefit to the individual. What is substantial, I think, is open to interpretation, and we'd be interested in hearing the committee's views on that as well.

As I said, for drugs substantial evidence usually means two well controlled studies and statistical significance at the .05 level.

CHAIRMAN DAUM: Dr. Glode.

DR. GLODE: I'm referring to page 22 of your handout. It's about the PE tube placement issue.

DR. PRATT: Right.

DR. GLODE: And this is my question. In the Kaiser study there was a statistically significant overall reduction in all episodes of otitis media. That was not found in Finland, but the study was not powered to show that effect.

DR. PRATT: That's correct. The estimates were actually quite similar.

DR. GLODE: Right. Okay. If we now go to PE tube placement, and so my question has to do with the power of the study to show an effect in Finland. So there is a reduction. It looks to me like approximately 13 percent.

The prevalence or incidence of PE tube placement was higher; I understand that, during the early period than usually occurs in Finland, but was the study powered to show if there had been a 20 percent reduction, would that have been statistically significant?

DR. PRATT: Right. The Finnish otitis media study actually did not look at tympanostomy tube placement as an endpoint. That was something that the FDA requested during the review period, that they go back and look at that. We were looking at consistency of effect with what was observed in the Kaiser study.

So, no, the study was not planned to look at that at all.

DR. GLODE: And then could I ask one quick follow-up?

So, I mean, was your conclusion that it was consistent or inconsistent? In the Kaiser study it showed a significant reduction, and in the Finnish study it didn't. Isn't the early period more comparable between the two studies as opposed to that late follow-up?

DR. PRATT: That's correct. I think that the early period is much more comparable.

Yes, they were not consistent. Again, you know, with the reservation that this study was not really designed to look at that, I think maybe the ascertainment of all of those tympanostomy tube placements might be in question, though they actually found that the rate was quite high. So I guess ascertainment was fairly good.

And, yes, the ascertainment or the number of ear tube's placement was higher than expected, I think, in that study.

CHAIRMAN DAUM: Thank you.

Dr. Markovitz.

DR. MARKOVITZ: Yes. I'm curious. Is there any precedence from the FDA point of view for this committee being asked to approve an indication with such low sort of levels of efficacy? Has this come up before just for historical purposes in helping us think about this?

DR. PRATT: I think for an historical perspective I might turn to Karen Goldenthal or Karen Midthun.

DR. GOLDENTHAL: We have considered some vaccines with lowish efficacy or lower than usual, if you will. One example that comes to mind was a typhoid vaccine, Typh Vi, and as I recall the point estimate of efficacy was about 55 percent and 74 percent for two trials, respectively.

And, of course, that vaccine was approved. I also think that there's some older vaccines that might have lower, you know, efficacy.

Dixie, it looked like you were --

CHAIRMAN DAUM: Thank you, Dr. Goldenthal.

Dr. Snider is next.

DR. SNIDER: Yes, i was just going to mention from my old days BCG vaccine efficacy, of course, would be one where you would be looking at lower efficacy.

Also, there is the issue that we have to confront here around semantics because if you really start looking carefully at the words that you're using, it's sort of analogous to saying, let's say, for BCG vaccine, for example, how much pneumonia do you prevent or with measles vaccine, how much skin rash do you prevent.

Because acute otitis media is due to multiple different organisms, probably mostly viruses. So when you start looking at more ill defined endpoints, then obviously efficacy goes down, and it would go down for other vaccines that we've licensed as well if you use endpoints that are less specific.

So I just wanted to raise that issue because I think we are confronted with the semantic problem as well as the scientific issue.

CHAIRMAN DAUM: Thank you, Dr. Snider.

I'd like to pose a question before I call on Dr. Katz, and it goes to the question that was asked before about precedent perhaps. Here we have a licensed vaccine with some rather stunning efficacy performance data against invasive disease, and now the very nice data that Steve showed us this morning in terms of what happened since it's been out there.

We have at least two committees that I'm aware of, and probably more, that have recommended this vaccine be given to every child born in this country, and that is if we can keep the supply chain going. That, in fact, is happening and will continue to happen.

So I'm not quite clear how the agency, and I'd also like to hear how the sponsor views this request for an additional indication. I'm not clear exactly what it means. We're already using the vaccine in a licensed way with a very acceptable safety profile and a wonderful efficacy trial for everyone. What will change if acute otitis media is added to the package insert?

I'd actually like to hear from both the agency and the sponsor in terms of the implications of what we're talking about here.

DR. PRATT: Well, I think anything that goes into the vaccine label can serve as the basis of promotion.

CHAIRMAN DAUM: Promotion meaning advertisement?

DR. PRATT: Advertisement, yes.

CHAIRMAN DAUM: So that's how the agency interprets this request.

DR. PRATT: Well, I think it's one interpretation.

Karen, do you have something to say?

DR. GOLDENTHAL: I also think that the agency believes, as Doug mentioned in his presentation, that the indication would have to stand on its own merits.

CHAIRMAN DAUM: Would the sponsor care to comment on this before we move on to Dr. Katz, then Dr. Overturf?

DR. SIBER: All right. Let me talk about two aspects of this. From a purely scientific and communications point of view, I think having a package insert that describes the effects of a vaccine that are published now and widely discussed and does so in an accurate way that is carefully reviewed between the sponsor and FDA is in the interest of proper communication about the benefits of the vaccine and its effects.

In the absence of that, physicians and parents are left to draw their own conclusions from media reports or whatever sources they might have, and I don't think that may be as accurate as what would be written in the body of the insert.

Another issue about which I am not the best person to comment is that there are, I understand, constraints on what our staff are allowed to say when there is no mention of an indication in the package insert, which is not in the best interest of accurate communication when you're constrained in that way.

From the vantage of whether it's an indication or not, I think the issue becomes the one at least from my scientific vantage, the one I mentioned, and that is if the standard for licensing a vaccine is some benchmark of efficacy against a syndrome caused by many agents, it would put a real chill on the potential.

Let me give you an example. A non-typeable Hemophilus Moraxella vaccine, which a number of manufacturers are working on, I don't think will have much more of an impact on otitis media than a pneumococcal vaccine. It will probably cause a similar number of otitis medias.

And one would want ultimately a vaccine that covers those pathogens. I don't see a way to license that vaccine. If you all tell us that while ten percent effect or eight percent effect or whatever, it's just not going to cut it as an indication.

The only indication for that vaccine perhaps in children would be otitis media.

CHAIRMAN DAUM: Thank you very much, Dr. Siber, Dr. Pratt.

Dr. Katz and then Dr. Overturf, and Dr. Stephens.

DR. KATZ: I had a comment which maybe has passed by, which was about the efficacy of vaccines and licensure. This is a very major argument, as many people around the table know, when you talk about HIV vaccines.

How much will you invest in vaccine research, development, and licensure of a vaccine to prevent a disease that's 100 percent fatal, but this vaccine may only protect 30 percent of people or it may only prevent clinical disease, but viremia will continue?

There are a lot of these issues that are certainly going to be faced when HIV vaccines come to licensure, but it's apples and oranges, I realize.

I wanted to ask a question maybe of Richard Schwartz. Frequently the statement is made that of all otitis media that occurs, pneumococcal is more aggressive, more virulent, apt to be more overt. So that the question of percentage may also be modified somewhat by that.

Can you corroborate that or deny it?

DR. SCHWARTZ: Based on both experience over a long time, oh, 25 years maybe, and published research that Dr. Rodriguez and I did where we looked at this was culture proven acute otitis media, looking at different pathogens that were obtained from the middle ear fluid, and comparing that with what the appearance of the eardrum and the height of the fever were at the time of presentation, those who had streptococcus pneumonia or pneumococcus had higher fevers, statistically higher fevers, and also a more angry appearance, more red, if you will, but more than red.

The drug, it's not just a red eardrum that's bulging. It has a very thinned out appearance. Often the epithelial cells on the surface of the eardrum that you're looking at begin to peel off. So there's a very noxious organism inside that's causing this drum to undergo some anatomical changes. It's dying. It doesn't die completely, but at least the outer layer peels off just similar to a bad sunburn. In fact, it looks like a bad peeling sunburn.

So somehow pneumococcus is meaner. It's not only meaner by look. Of the three major organisms, it's the only one that has any substantial bearing on separative complications of ear infections, such as acute mastoiditis.

If you look at pneumococcus, Haemophilus influenza and Moraxella, only the pneumococcus is incriminated of those three in acute mastoiditis, sa well as Group A strep. and sometimes staph.

But the other two, pretty much they're middle ear bugs, and they stay in the middle ear, assuming we're talking about middle ear infections. They don't go anyplace. They usually don't do very mischievous things, and they don't have the same import that the pneumococcus does.

CHAIRMAN DAUM: Thank you.

I'm going to try and return the committee to the issue of the FDA presentation primarily and ask Dr. Overturf and Stephens, who are waiting patiently, to direct questions to Dr. Pratt and clarify anything that they wish him to comment on.

Dr. Overturf.

DR. OVERTURF: I'd just like to reiterate again two major agencies have recommended that all children less than two years of age in schedules that were in the original application and were approved. And actually most of this additional data really is embodied in children less than two years of age, for which we already have an indication.

So the issue really is -- one is to put in new language that would include otitis media. The other issue is: is there a different process for including data on otitis media and the outcomes of the vaccine on otitis media without putting an indication in the package labeling?

Does the FDA handle that differently at all?

Dr. Pratt mentioned that we can amend safety data. The issue is: can you amend data or can it be put in the application without a specific recommendation for an indication?

DR. PRATT: Right. Data can be included in the clinical pharmacology section of the label without an indication. That is a possible way to include data.

DR. OVERTURF: But that's not one of the questions you're asking.

DR. PRATT: I see Dr. Midthun wants to comment here.

DR. MIDTHUN: I guess I'd like to speak to that. I think in general -- can you hear me?

Okay. I think that in general we do not like to include data in the clinical study section of the label if we don't believe that those data actually -- especially for efficacy data, if we don't feel that the data really support the efficacy of the particular outcome that's being described.

CHAIRMAN DAUM: Thank you for clarifying, Dr. Midthun.

Dr. Stephens and then Dr. Snider, please.

DR. STEPHENS: I'd like you to comment on two questions that were raised in your briefing. One is this issue that I think was also raised in the New England Journal of Medicine about the less credibility. If we approve this as a recommendation, will this have less credible effect upon the current recommendations, given some of the issues in terms of percentages and so forth of efficacy? That's one question.

And a second has to do with the standards for the licensure of other conjugate pneumococcal vaccines and whether in your mind this will enhance, decrease, change those standards.

DR. PRATT: I think what respect to the first point, the comment in the New England Journal. I really don't have any comment on that. That is a comment from one of the correspondents. I have no comment on that.

With respect to whether this would make it easier or more difficult for a pneumococcal vaccine to be licensed, it's hard to say. As I said in my presentation, I think it would set a precedent for the type of data, the endpoints that would be of interest, and the level of efficacy.

They did provide two well controlled clinical trials. Each one of them met the primary endpoint. I think that would set a level of efficacy for this indication.

CHAIRMAN DAUM: Thank you very much.

Dr. Snider, please, and then Dr. Decker.

DR. SNIDER: I had two issues. One, I would wonder if FDA or the sponsor had any further comments about efficacy against serotype 19F, why there may be less efficacy against that particular serotype in this vaccine.

I noticed, but it's just a subjective impression, that it appeared to me there was less boosting of the serological response on a percentage basis from the third or after the fourth dose with 19F, even though the GMCs were in the range.

So I'd like some comment on that.

Also, I'd be interested in some additional comments about this business about replacement and whether there is replacement with other serotypes and how you view that, if there is replacement.

I'm confused because Northern California Kaiser didn't seem to see replacement in a more general way, I guess, but whether that means replacement with serotypes that cause less severe disease and whether those are serotypes that are also less likely to be drug resistant, and whether those, also, that are less likely to be drug resistant are just less common in the population and, therefore, been less exposed to drugs.

I guess I don't understand the biology here of why some serotypes are more drug resistant than others.

DR. PRATT: Maybe taking the second point first,what was the second point again? I'm sorry.

(Laughter.)

DR. SNIDER: Is there really replacement, and if so, what is the significance of that?

DR. PRATT: Right. Well, I think the antibiotic resistance is clearly most common in those serotypes included in the vaccine. Replacement was observed in the context of the Finnish study. It's not clear whether that will actually occur in the general population. Again, this was from ear tube taps.

There may also be some data about nasopharyngeal carriage and whether there might be replacement there. I think I've seen some data to that effect from Ron Dagan.

But for the time being anyway, to the extent that vaccine serotypes are replaced by non-vaccine serotypes, it appears that one would be replacing more resistant organisms with more susceptible organisms.

Over time there may be antibiotic pressure on these as well, and they could also become resistant over time.

CHAIRMAN DAUM: ?Dixie, does that address your questions or are we leaving one out?

DR. SNIDER: Nineteen F.

DR. PRATT: Nineteen F. We also observed what you saw, that the booster responses were not quite as robust as some of the other serotypes, but I don't have any real insight into the differential efficacy for 19F. The sponsor may have some insight.

DR. SNIDER: What would be the efficacy if you took out 19F? I mean, would it jump up substantially?

DR. PRATT: We didn't analyze it that way. So I can't answer that.

CHAIRMAN DAUM: Let's move on to Dr. Decker, then Dr. Parsonnet.

DR. DECKER: Dr. Midthun mentioned that FDA feels that evidence of efficacy robust enough to be in the clinical pharmacology section -- let me rephrase that -- that evidence of efficacy should not be in the clinical epidemiology section unless it's robust enough to support an indication.

Did I paraphrase that correctly?

DR. MIDTHUN: Yes.

DR. DECKER: Does it then hold by inference that evidence of efficacy robust enough to be long in the clinical pharmacology section should be supporting an indication that should be added?

DR. MIDTHUN: It would depend on how the study was conducted and what the primary endpoint of the study was and what had prospectively been planned to be demonstrated. So I think there would be a number of different factors that one would have to consider.

DR. DECKER: But assuming those were met, if it's good enough for the clinical pharmacology, it's good enough for an indication?

DR. MIDTHUN: If the sponsor were to request this indication and the data supported it, and it had been prospectively indicated as being a major outcome in the study, you know, yes, in general that would be the case.

CHAIRMAN DAUM: Thank you.

Dr. Parsonnet, please.

DR. PARSONNET: It seems to me that one of the obstacles that we're hitting is just this semantic one of having this specific indication listed, this one specific one. And it seems to me just by looking at the proposed indication that when you talk about invasive disease and otitis media, you're basically covering 99.9 percent of all diseases caused by pneumococcus in children.

So I'm wondering if it's within our domain to suggest potentially having a different proposed indication which might be for active immunization of infants and toddlers against disease caused by Streptococcus pneumoniae due to the capsular serotypes; get rid of any specific indication; and then leave that in the text to talk specifically about the data for each of these indications.

CHAIRMAN DAUM: My understanding -- FDA people, feel free to modify -- is that we can make comments about most anything we like, but --

(Laughter.)

CHAIRMAN DAUM: -- but the business at hand is to address what's requested, and to decide how we feel about that, and so that everything you say is recorded. People are paid by various companies and groups to pour over your comments, and please feel free to make them because they will be noted.

On the other hand, we can't directly rewrite what's requested today.

Dr. Katz and then Dr. Schwartz.

DR. KATZ: Again, it may be a little bit apples and oranges, but when you license measles vaccine and you say it prevents measles, you don't say on the front line it prevents measles encephalitis or it prevents subacute sclerosing pan encephalitis.

When you read the small print and you read the papers, you say, "Okay. Measles encephalitis occurs one in 700 or one in 1,000 cases. Maybe after vaccine it occurs one in 100,000 cases."

Measles causing SSPE occurs one in 100,000 individuals. Maybe after vaccine it occurs one in a million or more, but there is nothing in the front line statement that says measles vaccine protects against measles encephalitis or SSPE.

That was comment one. Question two, and this may go down alone with Julie's in a different way. I don't know if either the agency or the producer would consider a modification of the statement. Instead of active immunization of infants and toddlers against invasive disease and otitis media, say against invasive disease and to a lesser extent otitis media.

CHAIRMAN DAUM: Thank you, Dr. Katz.

Dr. Schwartz, please.

DR. SCHWARTZ: I may have a unique position in being in front line practice compared with almost everybody else who's academician, and one of the concerns, potential concerns, that I have because it's already happened is that the pharmaceutical representatives for the company use the stated prevention of acute otitis media as a reason for using one of the antibiotics that the same company manufactures, which is Sufixene (phonetic), saying that since we've already done away with the worry about pneumococcal acute otitis media because of the vaccine, you don't have to worry about using amoxicillin clavulanate or anything. You could just go to Sufixene and hit bacteria number two and three, which is Haemophilus influenza and Moraxella catarrhalis.

And I have a concern that this may intensify with approval for this indication. I don't have a problem with the data. I have a problem with what it means.

CHAIRMAN DAUM: I think at this point we've probably exhausted our need to quiz and query Dr. Pratt, and I'm going to thank him very much for a fine presentation also, as we thank the sponsors for a fine presentation as well.

What I'd like to propose that we do now is to break and go to lunch. It's two minutes to 12. We will resume promptly at one o'clock right here.

Thank you very much.

(Whereupon, at 11:58 a.m., the meeting was recessed for lunch, to reconvene at 1:00 p.m., the same day.)

A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N

(1:03 p.m.)

CHAIRMAN DAUM: While we are settling down, I'll tell you FDA has asked us to reflect on based on this morning's presentations. I'm going to ask Dr. Pratt to please read the questions to us so we can focus, and then Dr. Midthun was going to make a comment briefly, and then we will have general discussion among the committee on issues that you feel need to be addressed in your minds before we vote and address the questions directly.

So Dr. Pratt first.

DR. PRATT: Okay. The first question is a two-part question. Are the data adequate to support efficacy of Prevnar in infants and toddlers for prevention of otitis media caused by Streptococcus pneumoniae due to capsular serotypes included int he vaccine?

If not, would additional analyses derived from the Finnish otitis media study, the Northern California Kaiser Permanente efficacy study, or additional clinical trials be useful in establishing efficacy?

The second question: please discuss the strength of the data with respect to secondary otitis media outcomes.

(a) Acute otitis media episodes cause by Streptococcus pneumoniae regardless of serotype;

(b) Overall reduction in acute otitis media episodes;

(d) Tympanostomy tube placement.

CHAIRMAN DAUM: Always easy questions.

Dr. Midthun, an orienting comment for us, please.

DR. MIDTHUN: I just wanted to state that what we're really interested in is the committee's opinion on whether the data support the efficacy for prevention of serotype pneumococci in the vaccine. Obviously the discussion will be, you know, very helpful to us, and we'll certainly use it as we, you know, consider this application further.

But what we're really looking for is your input on whether you believe that the data are adequate to support the efficacy.

Thank you.

CHAIRMAN DAUM: Thank you.

I think what I'd like to do is allow sort of dealer's choice now. So committee members can feel free to ask clarifying questions or raise issues that they think should be discussed, and we're not going to vote or directly address the questions quite yet until we sort of run out of things to talk about a little while.

Maybe we're further along than I thought we were.

Dr. Decker.

DR. DECKER: All right. I was going to wait for some questions, but I don't see any. So I'll lead off.

This is very interesting because as you pointed out, Bob, and as we all know, basically apart from our limitations of doing what we want to do, every kid in the United States gets this vaccine already, and so from that point of view you could argue that this is largely irrelevant.

But it's not. It's actually a very important question. As was pointed out by both FDA and the sponsor earlier, this is one of our first forays into that complex area where we have to look at a multi-factorial disease for which we have what appears to be solid evidence of excellent control for one of the factors and decide whether that's good enough to grant an indication to that particular vaccine for that, recognizing that it covers no more than at best a slim majority, perhaps depending upon your viewpoint a small minority of the cause of agents of the disease in question, particularly in light of the fact that if everybody already gets the vaccine, it would be easier to take a conservative stance and say, "Well, let's not stick our nose into that. Let's just let things be as they are. Everybody is getting it anyway."

But the consequence of that would be severe. There would be a number of vaccine programs that would be terminated, development programs that would be terminated if the committee did that because a lot of the problems for which solutions are being sought over the next decade or two are problems like this for which sponsors will come in with vaccines proven and well controlled trials to be efficacy for the specific issue, but which clearly in the larger context of the disease are only partial solutions.

If you turn down every partial solution, you never have a complete solution. So that's the problem that faces us.

I note with interest that the FDA clarified in their presentation that the proper consideration for licensure, or in this case for adding the indication to the license, is substantial evidence of clinical benefit, must be provided from adequate or well controlled studies.

And I note with particular interest that that does not say evidence of substantial clinical benefit. The clinical benefit is not for this committee to measure. It is the evidence that is for this committee to measure.

As a practicing physician what I expect of the FDA is that they will insure that the pharmaceuticals available to me have been proven safe and effective for their labeled claims.

Whether or not they're a wise public choice or a wise individual choice for the particular patient in my office at that moment isn't for the FDA to say. Other groups or I myself answer those questions.

I think what we've seen from the evidence presented today is that we've got solid evidence of clear-cut efficacy and safety for the requested label indication. So to me, this is a very simple meeting, and the answer clearly is to approve it for those indications.

CHAIRMAN DAUM: Thank you, Michael.

Your comments raise a question which I might ask Dr. Pratt, Goldenthal, and others to comment on with respect to Question 1. The question is worded to ask if the data are adequate to support efficacy.

So do you want us to reflect on whether we think the data demonstrate efficacy, in which case it may be a simple question, or do you want us to for further and talk about whether we think there should be an indication approval?

DR. MIDTHUN: We want to know whether you believe that the data support efficacy. We don't traditionally ask the Advisory Committee to actually vote on approving or not approving something, but certainly you're welcome to comment in any way.

CHAIRMAN DAUM: Okay. We continue to seek clarity. Other committee members, comments?

Dr. Glode.

DR. GLODE: I stuck on the issue of whether the condition which is going to be prevented ‑- whether the seriousness of that condition should have any bearing on the degree of efficacy. So there is statistically significant efficacy, and so I struggle with if this were, in fact, a presentation about a conjugate pneumococcal vaccine for prevention of invasive disease, and it's better than the control vaccine and reduces invasive disease by seven percent, would I consider that to be of clinical benefit?

And so should I have a different judgment because I'm dealing with otitis media instead of invasive disease? That's my issue.

CHAIRMAN DAUM: Does anyone want to answer that?

Go ahead.

DR. DECKER: I was just sort of going to comment on that because I think that the issue is -- I mean severity may be one of the issues, but I think the other issue is that we're really talking about a syndromic process here, and we're talking about even for the invasive disease indication that was to prevent invasive disease by these serotypes.

And so here we have the same thing, preventing otitis media by the serotypes, and personally, I think the data is very strong in two well controlled clinical trials that it does do that, but what gets people hung up is that it makes a relatively small impact on overall acute otitis media.

But as Sam pointed out and Dixie pointed out earlier, if it were measles vaccine for preventing rashes rather than measles vaccine to prevent measles, you know, then they would have a smaller impact on overall rash disease or overall pneumonia if you're talking about something else.

So to me it helps a lot just to focus down on the actual causes of the otitis media, and the FDA or other people are going to have to deal with this bigger issue of how it's marketed relative to the overall problem of otitis media.

CHAIRMAN DAUM: Thank you, Dr. Griffin.

Dr. Snider and then Dr. Whitley.

DR. SNIDER: I'd like to basically agree with what Diane just said and just emphasize what a bad precedent I think it would set to use the paradigm of a seven percent reduction of acute otitis media because it would really say that ten in the future you have to look at other vaccines according to the same kind of syndrome.

So you would have to look at pneumonia reductions. You'd have to look at rash reductions, that that would be a totally different paradigm that we've used.

And so I'm very supportive of using the clear endpoints related to these specific serotypes of Streptococcus pneumoniae.

Having said that, I have some questions about whether, again, going back to my favorite serotype here, 19F, and whether efficacy really has been demonstrated for that particular serotype adequately from the trials done thus far.

Now, you can argue that in the aggregate, you've shown efficacy for all of the vaccine, all of the serotypes included in the seven-valent vaccine, but have you shown it for 19F?

DR. GRIFFIN: And the only data that's relevant to that really is the Finnish data, right?

DR. SNIDER: Right.

DR. GRIFFIN: Because that's the only data that we have.

DR. SNIDER: Right, and there the 95 percent confidence limits are minus 14 to 51. I forget what the statisticians had to say about that. Maybe it would be useful to remind me.

CHAIRMAN DAUM: Would the statisticians like to clarify this point, please?

DR. GOLDBERG: There really isn't necessarily sufficient power for each of these endpoints, which is really the fundamental issue. It depends on how many -- I mean, I view these kinds of data in the small as descriptive; in the large, the collection of serotypes can be thought of a different way, but each one individually is just describing what your data are.

DR. SNIDER: Thank you, Dr. Goldberg.

CHAIRMAN DAUM: Thank you.

Dr. Whitley next, and then Dr. Decker, please.

DR. WHITLEY: I was going to reiterate a couple of the sentiments that had been made around the table earlier, but just to bring up one fundamental point, and that is many of us bring to this committee our own background, interest in a specific organism, and when we think about disease reduction, we think about how we can impact with a herpes simplex vaccine or an influenza vaccine, and what we're really asked to do here, what we get hung up on is what Mimi mentioned, and that is we look at a global syndromic approach, which is not the question the Food and Drug Administration asked us.

So I think if you look at the first question and just focus on the reduction of disease caused by those serotypes, the answer to the question unequivocally has to be yes.

And then the question is: what other supporting evidence is there to indicate that that's the proper decision or not?

And I think the analyses that Dr. Pratt asked for of the sponsor clarified some of the issues about potential imbalance and potential supporting information that would lead to further iteration of that conclusion.

CHAIRMAN DAUM: Thank you very much.

Dr. Decker, please.

DR. DECKER: You know, committee members have raised a number of really interesting and provocative questions, but I think there's a clear path out of the thicket. Dixie was asking about the 19F, but let's suppose that we knew that 19F completely didn't work. I still think that that would be irrelevant because we're being asked to approve a vaccine, not serotype by serotype.

And the question, although interesting and one that deserved to be addressed in the package material if it's true, if it didn't work, but needs to be addressed in other ways still isn't the question in front of us.

And someone -- I think it was Dr. Overturf -- earlier brought up the question of potentially inappropriate promoting of an antibiotic based on the vaccine indication, but the cure for that lies in FDA oversight and then promotion of that antibiotic and not in inappropriately failing to give the vaccine the indication that it clearly has earned.

DR. SNIDER: Can I just clarify?

With regard to Michael's comment, to me if you didn't include the serotypes, I mean, I don't want to get too nitpicky about this, but if you were talking about collectively four, 6B, 9B, 14, et cetera, then the answer is yes.

My only point is that if you are saying each one, then you don't have the data to say each one. You're talking about collectively.

DR. DECKER: Well, that's interesting because I interpreted it as the collective, but we ought to ask FDA how they want the committee to interpret it.

DR. SNIDER: That's exactly the kind of communication problems you have around this whole issue.

DR. PRATT: Well, I think even for the approval of the invasive disease indication there was not statistically significant evidence for each serotype. We considered it in the aggregate, and in the aggregate it was significant.

CHAIRMAN DAUM: Dr. Stephens and then Dr. Overturf.

DR. STEPHENS: Well, this 19F issue bothers me as well, and I am still not quite sure. The way this question reads to me we're saying that we have efficacy against 19F in this statement. That's the way I interpret this statement to read, and I'm troubled by that because I think there's not data from 19F.

CHAIRMAN DAUM: Dr. Overturf.

DR. OVERTURF: Well, you know, even being generous about the efficacy of this vaccine, if we eliminate the six and seven percent overall effect on otitis media, we're still only talking about a 50 to 57 percent effect against vaccine serotypes.

And when you look at that data specifically, the robustness on the Northern California trial is not as good because you weren't dealing with sero specific disease for most of that. So that the data is more robust for that seven percent.

So when people have alluded to using 55 percent efficacy in typhoid vaccine, which is against a disease which has a 15 to 20 percent untreated mortality rate, I don't think it's comparable to this.

So we're really talking really about a 50 or 55 percent efficacy based only on one trial which has fairly robust data and another trial which is really kind of marginal data to make a recommendation. And I think that needs to be reminded.

The 19F issue aside, we're not talking about huge efficacy rates for what, on the other hand, even if we assume that pneumococcal disease is not a supplement to these others, it's still largely a self-limited process.

CHAIRMAN DAUM: It's been a difficult task to assess pneumococcal vaccine performance for even 60 or 70 years when people have tried to take apart each individual serotype and try and demonstrate efficacy for each serotype, and I can't remember. Maybe someone on the sponsor's team could refresh us, but for the invasive disease indication, whether every single one of the seven serotypes had proven efficacy.

My recollection is not, and yet I think the balance of the data and the prudency of the committee was to advise that the data demonstrated impressive efficacy.

So I think that if you go back even to early trials that were done before most of us and perhaps all of us were born, when you started analyzing by each specific component of the vaccine, the studies weren't powered perhaps is the right way to say it, to show individual serotype efficacy.

The second point with regard to this is that we're not really being asked to address that question when it comes time at least to sum up and say what we think, although in terms of individual comments and exploring issues here, I think it's perfectly appropriate to discuss what that's all about.

Perhaps most interesting is what's the biology of all of this because if 19F was immunogenic and produced decent amounts of antibody, how come it didn't work? That's a separate issue, I think, from deciding the question of are the data adequate to support the efficacy of this seven component vaccine against otitis media.

Other comments?

Dr. Overturf, you did say your --

DR. OVERTURF: Well, I just bring up a point that we're still hung up on the word "efficacy" because we're talking about 50 or 55 percent efficacy, again, sero specific disease, and if you use that definition, the issue is then do you have two controlled trials that really demonstrate that.

CHAIRMAN DAUM: Dr. Markovitz.

DR. MARKOVITZ: Yeah, I also agree with Dr. Overturf about this. Even if you take the generous view, this isn't the world's most efficacious vaccine, and it's also, at least for otitis media, of course, and it's not to be compared with typhoid for the reasons he alluded to.

I'm also concerned that I don't know what the proper purview is for our committee on this, but in terms of the public reaction, I mean, in the vaccine world it seems that when things don't work the way people expect them to, it can actually bode very poorly for people actually taking the vaccine and using the vaccine.

Look at influenza. There's always a lot of people who think it's not worthwhile because it doesn't work all the time, and it has a substantially higher rate of efficacy than this one does.

I don't know if that's in our proper purview to address, but I am quite concerned about that issue, that this could end up being marketed heavily for otitis media and then when everyone is still getting otitis media, there's going to be a backlash. You could end up selling less of the vaccine in the end.

CHAIRMAN DAUM: I'd be happy to hear from anyone in the agency about that, but it seems to me that it's perfectly appropriate to address that, but it's not directly speaking the questions we'll be asked to vote on.

So your comments are noted. They're in the record, and thank you.

Other comments? Dr. Glode.

DR. GLODE: Could someone from the FDA just clarify the issue of one versus two trials that speak directly to Question 1? Is there a requirement for two trials that provide data specific to Question 1?

DR. GOLDENTHAL: There's a document that's available that's called the evidence document in a general way that applies to drugs and biological products, and that document is fairly general, but it has a principle that if you have -- you know, unless you have compelling evidence of efficacy, it's good to have two efficacy trials.

Do the endpoints have to be identical for the two trials? No.

I guess this really boils down to clinical judgment, as to, you k now, your view of these two trials. But the endpoint for the two trials do not have to be identical.

CHAIRMAN DAUM: Dr. Stephens.

DR. STEPHENS: At the pneumococcal meetings, there was some discussion about additional trials that are ongoing, and I just wanted to hear whether there is other data out there that we haven't been a part of. I was made aware of a Netherlands study that dealt with otitis media. I think there's a Navajo study that dealt with otitis media.

Are there other data that are pertinent from the perspective that would at least help us with this discussion?

CHAIRMAN DAUM: Does anyone from the sponsor want to deal with that?

DR. SIBER: Well, the buck stops there, huh? I guess it does.

(Laughter.)

DR. SIBER: There was a study in the Netherlands that was reported as an abstract in Anchorage, which we did not sponsor and which was a high risk group with recurrent otitis media at the beginning, who were given, as I recollect at least conjugate vaccine, seven valent followed by 23 valent polysaccharide vaccine, and in that particular subgroup of individuals immunized generally at older ages with already established disease did not show any benefit of that regimen, and that's all we know about it.

It's been an abstract level presentation at this point.

The Navajo study, I would prefer if -- there's a chart? We'll know more about the details of what was done at the Navajo study.

DR. KOHBERGER: The Navajo study on otitis media was a post hoc study. Basically it's chart reviews, and I think they were just hospitalized kids; is that correct?

We think it's just hospitalized kids. So it's neither a complete ascertainment as Kaiser or a specific ascertainment as in Finland. So there's a little question there.

And just to amplify a little on the Netherlands study, it's a population of children that we'll never see in the U.S. now because they're older kids with recurrent. They were not immunized at two, four, and six.

So I don't think the Netherlands study is relevant here. Does that answer your question?

DR. STEPHENS: I agree that the data isn't in. I just wanted your comments on those two. There are other studies.

The larger question is: is there going to be additional data above and beyond what we currently have?

DR. KOHBERGER: I don't think so.

CHAIRMAN DAUM: Dr. Snider and then Dr. Overturf.

DR. SNIDER: We were shown some serological response data, and I can't remember whether that was the entire population or some subset of the population. I would appreciate some refreshment on that, as well as the comment that the sponsor might have about serological responses among children who did wind up getting acute otitis media by various serotypes.

Is there any such information available?

CHAIRMAN DAUM: A comment to Dr. Snider's question?

DR. SIBER: Could you repeat the first part? Was that directed to the sponsor?

DR. SNIDER: Yes, George. I was just asking about the serological responses by serotype, and if that data -- if we could be refreshed about whether that came from the entire population or was a subset of the population.

And I was also asking about serological responses among those who became the cases of acute otitis media from vaccine serotypes, if there was any data on that.

Jukka has done, the statistician from Finland, some analyses of serologic correlations with otitis media. Again, these were presented as an abstract in Anchorage. They were still early days in terms of the analysis.

I don't know if you want to comment on that. Are you comfortable doing that? I did in Anchorage. So we'll do it here.

DR. JOKINEN: Well, the serological data was from a subgroup analysis of 60 per group.

CHAIRMAN DAUM: Excuse me. Can you speak right into the microphone so we can all --

DR. JOKINEN: Yeah. The serological data was for a subgroup of 60 per group, but they showed similar. There was seven time points for serological data during the FinOM study. They showed very similar responses as for the whole group, which was only sampled once, and os that's the first question, I guess.

And then analyzing serological correlates of protection we can see association with decreasing incidence with increasing antibody concentrations, but results varied between serotypes and not very confirmative, I mean, with regards to number of cases.

CHAIRMAN DAUM: Thank you.

Before you sit down, can you say who you are into the microphone for the transcript?

DR. JOKINEN: Jukka Jokinen, statistician from National Public Health Institute, Finland.

CHAIRMAN DAUM: Thank you very kindly.

Dr. Overturf, please.

DR. OVERTURF: Just one clarification regarding that first question, and there's an assumption here, and I want to make sure the assumption is correct that the efficacy that we want to support is for the use of Prevnar as a four dose regimen in children less than two years old.

So we aren't acknowledging any data that I have seen that really supports the use as a single or in reduced doses; is that correct?

CHAIRMAN DAUM: Someone from the agency? Dr. Pratt.

DR. PRATT: Yes, that's correct.

CHAIRMAN DAUM: Okay. Dr. Parsonnet.

DR. PARSONNET: Yes, I just also want to echo that I agree with Dr. Overturf, but I had a few other comments.

One is that we are talking about an efficacy of in the 50s, but in fact, the efficacy might be lower than that because the 95 percent confidence interval goes substantially lower than that.

So 50 is the middle, but it could be lower and it could be higher, and it would be nice to have supportive data to say which end of that number it really is.

And there is really only one study that directly addresses that question. With that, the second study supports it very strongly. I mean unless we think that this vaccine is having some nonspecific effect, and I think most or at least I would believe that its effects are due to the serotypes that we're seeing.

So I think there is only one study that directly addresses this, but the second study does show pretty strong support that this is likely to be the case.

But I think the problem that I feel like we're saying is that it's the problem with the indication, not with the efficacy again that keeps coming up, and that, you know, I just want to state for the record that I would prefer to address this point with a different indication that was being proposed because I think there are ways to address the ambiguity that we're feeling a little bit better in the overall statement that the FDA is planning to make and ways to address it in the text that follows that indication.

CHAIRMAN DAUM: I'd like to propose that we handle this because I think that it is a concern of a number of committee members and temporary voting members, that we handle this when we do formally vote by addressing the question we've been asked by FDA, but then you're welcome to make the distinction you just made and the comment to go with your vote.

Is that acceptable to FDA as a way to proceed?

DR. GOLDENTHAL: Yes.

CHAIRMAN DAUM: Okay. Additional comments that haven't been raised in discussion so far that people would like to air out a little bit?

(No response.)

CHAIRMAN DAUM: Well, we may be able to vote then. Let's try and do that and see how it goes.

We're going to -- well, you never know -- we're going to consider the first question first, and I don't think we need to have them read again. We had them read by Dr. Pratt, eloquently read after lunch, and, Dr. Stephens, we will call on you first to address question one.

I have two incoming comments here. One is from Dr. Griffin who says yes in two parts. Yes, would you please address both parts in one second?

Oh, when people are finished speaking, would they mind just pressing the little red button? Thank you.

Dr. Stephens.

DR. STEPHENS: The dangers of being at the end of the table.

CHAIRMAN DAUM: It does not happen randomly.

DR. STEPHENS: Yeah, right. I appreciate that.

(Laughter.)

DR. STEPHENS: We'll talk to you off line.

Obviously Prevnar has been very successful as a vaccine preventing basic disease. We've certainly seen that in our population, and it's been remarkable, as was mentioned earlier.

I do tend to view this somewhat as seven vaccines in one, and in general though the vaccine is safe. I think the data, at least the Finnish data, support a 50 percent decrease in culture confirmed pneumococcal otitis media for 6B, for 14, 23F, and 18C, and probably also for 9V and four, and both studies demonstrate an overall efficacy of approximately six percent against otitis media.

And there's also obviously reasonably convincing data that there is significant decrease in tube placement and potentially in complications.

And I think there's also evidence that 6A and 23A and 9N are probably also affected in terms of limited data by the vaccine.

I have concerns about 19F. I'm not quite sure that we have demonstrated that the -- the efficacy is at best very limited for 19F, and I share Dr. Parsonnet's view that maybe this could be handled differently than the statement as worded as the better approach rather than the specific statement.

I am concerned also about the serotype replacement issue.

So with all of those caveats, I will vote a qualified yes to the first part, and it doesn't look like we're going to get any additional data of significance.

CHAIRMAN DAUM: I hate to boomerang this back to you, but I don't think we can have qualified yeses. We need a vote, with the comments being qualifiers as appropriate.

DR. STEPHENS: So moved and seconded.

(Laughter.)

CHAIRMAN DAUM: Thank you, sir.

Dr. Katz.

DR. KATZ: I don't think I have anything different to say than David Stephens has said. The question as it is worded, we're not voting on the wording of what was given to us as the statement to be made on the vaccine insert or the package. We're just voting on are the data adequate to support efficacy, and I would vote yes.

CHAIRMAN DAUM: Thank you very much, Dr. Katz.

Dr. Snider.

DR. SNIDER: With regard to the answer to question one, my answer would be yes. I would have many of the reservations or concerns, I should say, that Dr. Stephens has articulated and would also be with those around the table who have expressed some concerns about whether this would be a wise thing to do to include this indication, and whether that would be good for the vaccine, good for the manufacturer, society at large, at least as it's now articulated, and would encourage further thought in terms of how the indication would be laid out, as well as what kind of physician education and parent education program would be put in place so that people understood what they could expect and not expect from the use of this vaccine.

Because I do worry about many of the secondary and tertiary consequences that people have already mentioned that I don't need to go back over.

CHAIRMAN DAUM: And your vote is yes, Dixie?

DR. SNIDER: My vote is yes. Yeah, I tried to start out and make that clear.

CHAIRMAN DAUM: You probably did. I'm sorry.

Dr. Hamilton.

DR. HAMILTON: I believe the data are sufficient to support efficacy.

CHAIRMAN DAUM: Can I stop you? Just speak right into that microphone so that we can hear. Pull it toward you.

There you go.

DR. HAMILTON: I think the data are sufficient to support efficacy as the question is phrased. I think I'd be interested in looking at the serologic response of the 60 individuals who developed 19F infection. Whether that would be of any use or not I don't know, but that does seem to be an outlier.

CHAIRMAN DAUM: Thank you.

Dr. Schwartz.

DR. SCHWARTZ: Yes.

CHAIRMAN DAUM: Succinct.

Dr. Glode.

DR. GLODE: I'm going to vote no. I think the data support efficacy from the Finnish trial, serotype specific efficacy. I'm simply not willing to presume from the Kaiser trial that that's the explanation for the reduction, and I only have the ruptured eardrum information, which is not statistically significant.

So I think that I'm voting no, not because I don't believe the Finnish trial, but on the basis of just one trial that proves that to me.

CHAIRMAN DAUM: Dr. Glode, we have to ask you part two since you voted no. No deed goes unpunished here.

Would additional analyses derived from the trial or additional clinical trials be useful in establishing efficacy?

DR. GLODE: I don't think any additional analyses would be helpful because it looked like they have been thoroughly done with both trials. Certainly another tympanocentesis trial looking at vaccine specific serotype would be helpful.

CHAIRMAN DAUM: Thank you very much.

Dr. Overturf.

DR. OVERTURF: I also have to vote no because of the same reason. I feel that only one trial, the control trial, has really demonstrated efficacy.

I also am very concerned that a yes vote here sets another precedent which we haven't talked about. One is besides establishing some arbitrary definition of, quote, efficacy, unquote, for a noninvasive disease, it also would, it seems to me, establish a new bar that you must show efficacy for every indication you list, including sinusitis, pneumonia, and other issues.

So actually I still like the recommendation of Dr. Parsonnet, which is that if you're going to do this, if you do it intuitively, you do it based upon all diseases due to pneumococci of the serotypes.

I agree with Dr. Glode that probably the only thing that's going to resolve this is another tympanocentesis trial. It's the only thing that's going to really confirm what I believe is probably a correct observation from the Finnish trial.

So my vote is no.

CHAIRMAN DAUM: Thank you very much.

Dr. Faggett, sir.

DR. FAGGETT: Yes. I do have some concerns that the studies presented are not as inclusive as I would like to see in terms of having the broader, heterogeneous American population represented.

And I do have some reservations about efficacy with 19F.

Having said all of that though, I think in terms of the data adequate to support efficacy of Prevnar in infants and toddlers, I think we do have adequate data to support efficacy because we're not saying just how much.

So I think I'm going to go ahead and vote yes on this, with those reservations.

CHAIRMAN DAUM: Thank you.

Dr. Griffin.

DR. GRIFFIN: I'm going to vote yes, and just state that I think the Finnish trial, as many people have said, I think gave us definitive evidence of efficacy against these serotypes. The fact that the overall effect on otitis media was almost exactly the same as it was in the Kaiser trial, you would have to invoke a thought that it was a totally nonspecific effect and that there wasn't an effect on these same serotypes, that it didn't have the same sort of display of different types of causes of otitis media.

So I think the data is adequate and I think it has been shown in two trials.

CHAIRMAN DAUM: Thank you very much, Dr. Griffin.

Dr. Whitley.

DR. WHITLEY: My vote is yes for virtually identical reasons to Diane's, and that is I think unequivocal evidence of efficacy was established in the Finnish study, and I think it's unreasonable to think that the results from the Kaiser Permanente study could be attributed to anything other than the vaccine.

Having said that, I do think it's important for both the agency and the sponsor to carefully weigh how the package insert reads and whether we go back to what Sam's recommendation was earlier regarding the lesser degree of efficacy in the prevention of otitis media or careful words to indicate that that would be the case, I think, is relevant because we clearly don't want to detract from the impact of this vaccine on prevention of invasive disease.

CHAIRMAN DAUM: Think you, Rich.

Dr. Diaz.

DR. DIAZ: I likewise would vote yes because I am also willing to extrapolate the Finnish data into the Kaiser trial. The data is so similar, and yet I do agree that only another tympanocentesis study would validate that completely for me.

But the extrapolation is there. As Dr. Griffin pointed out, you'd have to imagine something completely unrelated would have played into making the data become or be the same in the Kaiser trial.

That having been said, I think one of the points that Dr. Katz raised earlier, which is that we're not being asked to validate or vote for a package insert as stated in the materials that we received, is extremely important because I, too, have great concerns and reservations about issues surrounding package insert for this indication for all of the reasons that everyone has already elucidated.

CHAIRMAN DAUM: Thank you, Dr. Diaz.

Dr. Goldberg, please.

DR. GOLDBERG: I'm voting yes as well.

That said, I really view the otitis media endpoints here as secondary endpoints to the basic indication for the vaccine, and I urge the agency to find a way to deal with this and not call it an indication in the label; that there needs to be another mechanism for handling this that takes into account the primary purpose of giving the vaccine where the major impact is felt and doesn't detract in the long run from that.

CHAIRMAN DAUM: Thank you.

Dr. Markovitz.

DR. MARKOVITZ: Yeah, I think the way the question is phrased I have to vote yes because I think there is efficacy within this narrow spectrum.

I have reservations that have already been expressed by me and others here and I might return to on question two, but for question one I would say yes.

CHAIRMAN DAUM: Thank you.

Dr. Parsonnet.

DR. PARSONNET: I say yes, and I just also want to point out that I don't think there's much experience with what the efficacy should be for a disease that's not invasive and not produced by a toxin. So it may be that 50 percent is really fantastic for a noninvasive disease and we just don't know from our experience yet what noninvasive disease efficacies are ever going to be.

And with that I just want to reiterate my thoughts about the labeling, that I'm concerned about the labeling as we talked about before.

CHAIRMAN DAUM: Thank you.

Ms. Fisher.

MS. FISHER: As I understand it, the function of the FDA standards for including a use indication in a vaccine manufacturer's label is to insure truth in advertising. The public looks to the FDA and trusts that CBER protects their right to informed consent when using biological products.

We've been asked by Prevnar's manufacturer to agree they have proven efficacy for prevention of otitis media, which would then allow the advertising of Prevnar as a vaccine to prevent otitis media even though a large U.S. study has shown only a seven percent reduction in all otitis media cases.

Practically, that means that 93 percent of the children whose mothers believe Prevnar is an ear infection vaccine because the label says so, 93 percent of these children may be susceptible to otitis media even though they've been vaccinated.

Seven percent is not a scientific standard for vaccine efficacy for a specific condition that engenders a lot of trust. Because apparently there is no room under the regulations to simply state that Prevnar efficacy with respect to presenting otitis media has been shown to be seven percent, which would be the most truthful and accurate labeling.

I must vote that efficacy has not been established to justify the labeling change, and I don't think the manufacturer should spend more time and money trying to prove efficacy for otitis media in light of these studies and should be well satisfied with the use of Prevnar to protect against severe invasive disease.

CHAIRMAN DAUM: Thank you, Ms. Fisher.

I guess I'm probably last and possibly least, but a couple of comments.

I think that I'm going to vote yes for the question the way it's worded. I think this vaccine represents a real triumph for children in this country. It's had a wonderful track record in terms of its safety profile, and as important, a wonderful impact on invasive pneumococcal disease.

I believe that the two trials that were done, although I agree with Dr. Glode's comments about the differences in methodology and conclusions that can be draw from each one of them are internally consistent, and they do show that there is efficacy against otitis.

A plausible secondary part of that is for me the more severe the otitis was, the better the vaccine seemed to perform. I'm not troubled globally because of a lot of previous work that's gone on in this field that there isn't efficacy shown for every single serotype. I think that the problem of powering the study to do that would be formidable.

I do have some concerns though. One of them is that by saying yes to the demonstration of efficacy that we somehow have established or influenced the agency to establish a candle for equivalency or noninferiority of other vaccines to these numbers, and I have some concerns about that.

I'm disappointed that the vaccine isn't more efficacious. I think my answer is, yes, it is, but I wish it were higher, and so that I share some of the concerns by Dr. Parsonnet, Ms. Fisher, and many others that this number be translated into an important clinical message.

I'm concerned about the point of Dr. Overturf, that even when you put the best possible light on things and take only culture proven otitis due to serotypes contained in the vaccine, that the efficacy was still only in the 50 percent range, and I think that raises some points that we haven't said much about, and that is that we don't completely understand the biology of preventing otitis media.

The 19F story is instructive to me only in that it says that there isn't a simple relationship between the production of antibody and efficacy in this case. I don't know what the other issues are. They could have something to do with cytokines or something to do with individual eustachian tube kinetics or lots of other issues, but I think this is more than just measuring antibody and looking at efficacy, and that's why I think in the best case we only got to in the mid-50s.

Having said that and voted, and to emphasize I do vote yes about efficacy, I would offer the extraneous advice or probably unwelcome advice that this not be used to translate into promotion, sales marketing, or pressure on the public to accept the vaccine based on this performance alone.

I think there are many other reasons to promote and use this vaccine, and I think that it has been said repeatedly, and I think that it's recommended for all U.S. children as it should be, and I think that should continue to go on.

But I personally would not like the efficacy that I believe has been established to be translated into detailing or advertisements or direct marketing to the public about the impact of this vaccine on otitis media.

And with that, I think I'm done.

I'd like to move on at this point to announce the vote for question one. Sorry. One second.

Dr. Decker, you need to state your opinion, but I think you already have. Did you want to say something else about question one? I'm pretty sure you've done it.

The outcome on the vote on question one is 13 voting in favor of the data being adequate to support efficacy, three opposed, 13 to three.

The second question isn't really a question, but is a discussion point that we would like to hear or the agency would like to hear committee members weigh in on specifically.

I normally start off with some general discussion before we go to the specifics, but my sense is we've had the general discussion. If anyone disagrees with me, please let me know, but I'd like to start with Dr. Stephens again and just address -- perhaps we can do this fairly briefly -- the issues in question two, but please feel free to make whatever points you wish.

DR. STEPHENS: I think a lot of this has already been stated. I'd like to make two points.

One is the issue of 19F and an urge. Continue work on trying to understand why 19F is, in fact, not working very well in current vaccine.

Secondly, I think pneumococcal disease is a moving target, and what is efficacy now with the serotypes in terms of prevention of otitis media may not in five years, given serotype replacement and other issues be efficacy.

So there needs to be continued monitoring of post marketing, and a continued look at otitis media.

CHAIRMAN DAUM: Thank you very much, David.

Dr. Katz.

DR. KATZ: I would like to emphasize what Dr. Parsonnet had, and that is otitis media, even due to a single bacterium, is not a single etiologic relationship. We know that preceding viral infection; we know that allergy; we know that smoke inhalation; there are many factors that have to do with this, and to expect that a vaccine against a bacterium is going to change things 80 percent is naive.

And I think that we need to continue to support research in all the areas relating to the etiology, not just the pneumococcus in the reduction of acute otitis media episodes, and therefore, I'm comfortable with saying that what we voted, with question number one.

I would be uncomfortable, as has been repeatedly stated to say that, "Okay, folks. Now we've got the panacea for otitis media. Have your children immunized not to prevent invasive disease, but to prevent otitis media."

And I think a lot needs to be done in research. You used the term or Diane did "syndromic." And I think we would be less than a good Advisory Committee if we let it go as just we've solved an issue.

We need a great deal more to be studied. I think one of the side benefits that I hope might come from this, of course, and this might be studied particularly in the Kaiser population, is reduction in the use of antibiotics and antibiotic resistance because I think those are probably more major questions than otitis media.

CHAIRMAN DAUM: Thank you very much.

Dr. Katz raised many good points.

Dr. Snider.

DR. SNIDER: Well, with regard to acute otitis media episodes caused by Streptococcus pneumoniae, regardless of serotype, I mean, obviously efficacy drops. I mean, that's pretty clear, although, you know, some of the related serotypes, that data, although as Dr. Goldberg has pointed out are only anecdotal in the sense that they're not definitive and not statistically significant. They're encouraging.

But overall, as one would expect, there's not a huge amount of efficacy against serotypes that are not included in the vaccine, nor should we expect there to be.

Like others, I think that we -- I think David mentioned this -- we need to be vigilant because doing this trial in one setting, in one country, at one particular point in time doesn't necessarily give us a clear indication what might happen when the vaccine is used in larger populations in different places, at different times, and so forth.

So I think we need to be vigilant around this issue of potential replacement. I don't think it's completely off the table that it will occur or won't occur, and the issue of drug resistance, the story seems promising right now in terms of potential for using this vaccine to reduce the number of drug resistant cases of pneumococcal acute otitis media.

But we need to monitor over time to make sure that that's not just a phenomenon that is a result of particular circumstances that exist now or exist -- but won't exist in the future.

I was a little surprised with regard to tube placement not being different early in the trial and showing a difference only later on, but I think we had some explanation given to that that sounded reasonable.

So I think there is a lot more work that needs to be done to try to understand this disease, to try to understand this vaccine, to try to understand why certain -- it seems to protect against certain serotypes better than others. One would hope that the whole pneumococcal conjugate story would continue to evolve with more serotypes in subsequent vaccines down the road, and that we'll continue to be discussing this as new data become available, as newer pneumococcal vaccines become available.

CHAIRMAN DAUM: Thank you, Dixie.

We'll move on, please, to Dr. Hamilton.

DR. HAMILTON: You requested the strength of the data, but not in which direction.

CHAIRMAN DAUM: Dr. Hamilton, will you speak right into the microphone so that we can be sure we hear you?

DR. HAMILTON: Okay.

CHAIRMAN DAUM: Thank you.

DR. HAMILTON: You've requested the strength of the data, but the question does not refer to which direction you're interested in.

CHAIRMAN DAUM: Your choice.

DR. HAMILTON: Okay. Acute otitis media caused by S. pneumonia regardless of the serotype, certainly the Finnish study suggests that the serotype related strains, there's some efficacy going from 56 to 34, but for those that are not related at all, there doesn't appear to be any efficacy.

And then it's kind of like the top of a pyramid that branches out. As you get to the syndrome at the bottom, you get less and less effective, and one of my queries is actually whether that 56 percent and 34 percent reduction directly correlates into the six percent reduction or you're seeing replacement with other radiologic agents there.

And no other microbiologic data was mentioned, but I'd just be curious to know.

Overall reduction, acute otitis media episodes because it was supported by two trials, although somewhat differently done. I think the strength of the data is pretty good.

Frequent otitis media, I would say no, and tympanostomy tube placement I would say no. I agree that the tympanostomy tube replacement was not adequate and well controlled, and it's impossible to determine how one would relate one to the other, given the different time frames and the different treatment methods.

The other thing that you haven't mentioned here but I think is provocative is the rupture issue out of Kaiser, and I would like to know a little more about the microbiology there, whether they were doing microbiology for, say, Group A strep. with those who ruptured.

CHAIRMAN DAUM: Thank you very much.

We're going to go out of sequence here and ask Dr. Decker to comment next.

DR. DECKER: Thank you, Bob.

CHAIRMAN DAUM: Because he has to leave.

DR. DECKER: You know what I was struck by more than anything else, I think, in looking at these data was the remarkable consistency of the data when you recognize the usual statistical limitations and variations, particularly with declining sample size. What you see is a remarkably consistent picture, I think, of efficacy.

Let me put it this way. Were it known without question that the vaccine were efficacious against the serotypes that cause otitis media and it was reducing otitis media, then you would see the pattern of results that we have here, and apart from the limitations caused by shrinking sample size as questions become more narrow, the variability caused in each statistical sampling and the fact that not all serotypes are equally prevented, with those thoughts in mind, the data are remarkably consistent.

And i think the questions that the data really leave for us aren't necessarily the ones listed here, but rather some of the questions that have already been brought up, the 19F question and what's going on with that. Why is it behaving differently? The serotype replacement question begs further attention.

To me most of the questions that are asked in item two, although perhaps not proven statistically with these data, clearly are addressed so uniformly and so consistently that I'm not concerned by them.

CHAIRMAN DAUM: Thank you.

We'll continue going in sequence now. Dr. Schwartz, I don't think you can be quite as succinct this time as last time.

DR. SCHWARTZ: No. Everything that I can think of that's been said has been said with one exception. There are actually two different vaccines, and I know we're not discussing the other vaccine that was tested in Finland, and to me that means that there are, in fact, two trials. Even though the vaccine is a different manufacturer and a different carrying protein, we don't have any information at all as to how effective was the Merck vaccine, the septavalent vaccine.

Just for curiosity, did that do the same thing? And if so that would bolster my confidence that I voted the right way.

CHAIRMAN DAUM: Does anybody from the agency or the manufacturer wish to comment on that question? You have the right to remain silent.

DR. GOLDENTHAL: I don't think it would be appropriate for the agency to comment, but I do see a representative from Merck.

CHAIRMAN DAUM: We also have an investigator at the microphone.

DR. KILPI: All right. The other vaccine, we have presented the results for the other vaccine at several conferences, and it provided almost equal protection against AOM due to vaccine serotypes. The point estimate was 56 percent against the overall vaccine serotypes.

It also had a problem with 19F. I think the point estimate was 34 or 37. I'm not quite sure, but overall the efficacy against the vaccine serotypes was pretty much the same.

However, the point estimate in the prevention of any pneumococcal AOM was slightly lower. The point estimates was only 25 percent.

CHAIRMAN DAUM: Thank you very much. That's very helpful and very interesting, and I'd like to just remind the committee though that it's not part of what we're considering at the table today.

Dr. Glode.

DR. GLODE: I think that the table on page 29 of Dr. Pratt's handout summarizes this information very nicely, and it pulls it all together.

So acute otitis media caused by Strep. pneumoniae, regardless of serotype, again, the single study that answers that is the Finnish study with 32 percent reduction, lower limits of confidence, 20 percent, efficacy. Sorry.

Overall reduction in acute otitis media, again, as already mentioned, statistically significant in the Kaiser trial, in the same range in the Finnish trial, but not statistically significant due to lack of power to detect that apparently.

And then effect on frequent otitis media, again, summarized in the table. The most relevant data there would come from the Kaiser study.

In tympanostomy tube placement, I remain confused about whether or not there was adequate power in the Finnish study to show it, and if there was why it didn't confirm and validate the Kaiser study.

So I think that at least with tympanostomy tube, I'm still a little confused about that issue, and overall reduction in acute otitis media I can explain that by a power of the study.

I did think between these times of one other analysis that would be of interest to me and was probably done, and that addresses the issue of should you presume that the only explanation for the reduction of overall otitis media in the Kaiser trial has to be related to the effect of the vaccine on vaccine specific serotypes because I have been impressed for a number of years with the biologic cross-reactions between organisms, in particular, pneumococcus and Haemophilus.

So it would be of interest, and I'm sure people have already done this, to just look at the efficacy or lack of efficacy against Haemophilus otitis and Moraxella otitis in the two groups.

CHAIRMAN DAUM: That was a provocative comment. Thank you, Mimi.

Dr. Overturf.

DR. OVERTURF: I think I agree with most of what's been said. In particular, regarding episodes caused by Strep. pneumo. regardless of serotype, I think it's clear that there is very little, if any, efficacy, and this is actually supportive of the vaccine against serotypes not contained within the vaccine.

And I think the issue of replacement serotypes is an important issue for this disease, and it's probably going to be the same problem for Haemophilus and other otitis media vaccines in the future.

And one of the concerns I have is once the vaccine is approved for this indication, how demanding is the post marketing surveillance going to be in terms of watching for this?

Today we have a six or seven percent reduction in otitis media if you believe the data now. Will it be three and a half percent next year and will it be less than that thereafter?

Because we will be setting a new baseline here, which will be established by the routine immunization of all children less than two years of age in this country with this vaccine.

So I think that's going to be an issue about future overall reduction in acute otitis media episodes. And eventually if frequent otitis media is really more due to pneumococcal disease, that also will be an issue.

I have the same concerns about the tympanostomy tube placement. I don't think I'm clear because I think this is a disease which probably is related to a lot of other issues that can't be directly identified in the study, particularly issues of individual anatomy genetics, other risk factors, and also just the indications and use of tympanostomy tubes.

So I think it may be at some point appropriate when you do further studies to move this particular indication up the ladder a little bit so that one moves it up to a primary outcome rather than and with more restrictive indications because I think right now it's not clear to me that the way it was examined in the current studies really makes it clear that the outcome was affected by the vaccine or affected by some other factors that we couldn't really identify.

CHAIRMAN DAUM: Thank you very much.

Dr. Faggett.

DR. FAGGETT: I pretty much agree with my distinguished colleagues and their comments. Specifically the acute otitis media episodes, question A, the Finnish study does have good evidence of efficacy. Kaiser is equivocal.

Question B, overall reduction in acute otitis media, there is some evidence of efficacy.

Question C, frequent otitis media, still equivocal. Some evidence of efficacy.

And, again, pretty much the same with the PE tube placement.

I would hope that further studies would allow us to have the same kind of eloquent data in terms of looking at the broader population so that we can, indeed, draw more conclusions for the population that is not included in this study at the present.

CHAIRMAN DAUM: Thank you.

Dr. Griffin.

DR. GRIFFIN: Okay. I think that the two studies, because of their very different design, shed light on different ones of these questions, more or less light on different ones of these questions.

In the first, you're talking about acute otitis media caused by all serotypes of S. pneumoniae, and in a way that's just making it less specific and so diluting out the effect somehow by adding in all of the serotypes.

And I think there is statistically significant evidence from the Finnish study that that happens, that you still have a significant effect even with that dilution.

I think the issue of replacement serotypes is an important one, and I think it's also an interesting one because it isn't clear to me. I mean, these serotypes were chosen to make a vaccine partly because they cause most of the disease.

And so are those replacement serotypes intrinsically less virulent and, therefore, even though they may appear more, they still won't cause more severe disease or not, and I think time is only going to tell what that answer is.

The second on the Part B, those studies showed a very similar decrease in overall reduction, one, because the power of the study is not statistically significant while the other of the Kaiser study is.

Again, with frequent otitis media, because that was an outcome where there were larger populations in the Kaiser study, there was a significant difference there.

Like everyone else, the tympanostomy data is somewhat confusing, although I thought the Kaiser data for an American population using what's pretty much a real world situation did indicate that there will be a reduction in tympanostomy tube placement.

CHAIRMAN DAUM: Thank you very much.

Dr. Whitley.

DR. WHITLEY: I really don't have anything to add to the discussion that has taken place. I just reiterate four points:

That if the FDA has a chance to do Phase IV studies with the sponsor, they would be important, and clearly replacement serotypes and the response to 19F are critical, but also the whole issue of understanding better antibiotic usage in this targeted patient population, the propensity to develop resistance or not develop resistance becomes essentially important.

CHAIRMAN DAUM: Thanks.

Dr. Diaz.

DR. DIAZ: Thank you.

Just a couple of comments in general. I'm not surprised with the delusional effect as has been described as one moves from specific serotypes to, you know, related serotypes, to nonrelated serotypes, to acute otitis media as a whole. It follows logically, and I think the data followed logically.

It is a specific vaccine in the sense that it has certain serotypes within that vaccine.

I do think though that this issue that's been raised about replacement serotypes is an important question, but I might even consider broadening that in wondering about not only replacement serotypes, but perhaps replacement of other organisms shift because this really is not a selective ecosystem in the middle ear for the pneumococcus per se.

And I think a lot of my colleagues have commented upon sort of the complexity of acute otitis media. I mean, it's such a common diagnosis, such a common disease, and yet our ability at diagnosing it remains at about the same point it was years and years and years ago.

That having been said though, if all otitis media is not the same, meaning that there are more pathogenic and virulent organisms, as appears to be the case, then we should be able to see some differences or changes over time as the use of the vaccine increases dramatically in this country.

And one would hope that even if we do not see truly a decrease in cute otitis media, the seven percent versus three percent next year, et cetera, et cetera, that we might see some changes in some of the things that others have commented upon in terms of monitoring, and that might be the use of antibiotics overall, changing patterns in use of antibiotics, changes in diagnoses for febrile illnesses, more virtual syndrome versus acute otitis media per se.

And likewise perhaps tympanostomy tube placement. Some of those things can be looked at. Tympanostomy tube placements can be monitored by ICD-9 codes and a variety of other mechanisms sine they tend to be in hospital procedures, and I would encourage the utility of looking at some of these other perhaps not specific, but yet indicators of morbidity due to acute otitis media in general in the future.

CHAIRMAN DAUM: Thank you.

Dr. Goldberg.

DR. GOLDBERG: I think that one comment is that I think the trials are remarkably consistent. The places where they're not -- the only difference that I would call possibly nonconsistent is the tube placement issue, which I think various reasons for that have been alluded to throughout the morning.

All of that said, you can look at the two trials as complementary and quite supportive of one another with regard to the results. The real issue is: is this efficacy sufficient, particularly with regard to the acute otitis media overall endpoint?

In fact, would I be happier if four percent were really significantly different from zero in the intent to treat in the Finnish study? Not really. It wouldn't matter to me.

I mean, I would urge the agency to really give careful consideration to what we mean by efficacy and what is meaningful efficacy and how you deal with reporting results like this in a label without calling it an indication that can detract from the primary purpose of the vaccine.

I mean, I think that has to be considered. I mean, I think that they basically in one way or another -- some efficacy, in quotes, or activity with regard to each of these endpoints has been demonstrated in one or the other or both of these trials. The issue is what to do with it.

CHAIRMAN DAUM: Thank you.

We are going to address the issue of how to communicate information based on your beliefs today when we finish addressing question two. So I'm glad you raised that point, and we'll come back to it.

Dr. Markovitz.

DR. MARKOVITZ: Yeah. I don't have too much to add, except that if we could make question two a yes or no, I'd have to say no. I'm not very impressed with any of the other efficacy of this vaccine beyond what was asked in question one.

And I'd like to echo the comments just very quickly of several of my colleagues. Actually Dr. Goldberg, who I know is a statistician, has correctly pointed out, I think, that just having a good P value doesn't make something really clinically useful and worth marketing.

And then I'd like to also agree with Ms. Fisher in terms of the possibility that by marketing this for otitis media we'll squander the good name of the vaccine and hence do much more harm than good for both the public and, frankly, for the company, too

And then lastly, I'd like to agree with Drs. Parsonnet and Katz that some rewording would be very much in order when it comes to the indication, and that might solve the problem.

Thanks.

CHAIRMAN DAUM: Thank you.

Dr. Parsonnet.

DR. PARSONNET: Yeah. I don't have very much to add to the things that have already been presented, that these particular secondary outcomes have not been as well demonstrated, but I think one of the things that I think from a philosophical perspective it's important to think about is, again, that this is a vaccine. We don't typically have vaccines against such a common resident of our normal flora. This is something that people carry a lot and very frequently, and I don't think that it's possible to really predict what the long-term consequences of trying to address that with a vaccine is going to be in terms of the disease outcomes that might be associated with it that may be replaced by other pneumococci, may be replaced by other organisms, may have effects that we just don't really know.

So I would encourage since we are changing the human ecology here, that people really look at it very carefully and address this after marketing.

CHAIRMAN DAUM: Thank you.

And Ms. Fisher.

MS. FISHER: Well, a five to 21 percent reduction of these secondary otitis media outcomes is not strong evidence for proof of efficacy for this condition, and I do think that more data needs to be generated regarding the possible future increases of otitis media due to serotypes not covered in Prevnar as a result of mass use of Prevnar.

CHAIRMAN DAUM: Thanks a lot.

Dr. Midthun, you wanted to ask the committee to address an additional issue informally. Could you tell us what it is?

DR. MIDTHUN: Yes. I'd really appreciate input from the committee members on how they would suggest communicating this information to the prescriber because I've heard a lot of concerns about how this might be done, and we're really appreciate input on that.

The other thing I would like to speak to, Dr. Daum, is you made a comment that you had concerns about how perhaps some of this information might be used for marketing purposes, and I just wanted to come back to say that information that is included in the package insert can be used for marketing, promotional labeling, but that is something that is reviewed by FDA and has to be approved by us, but that is something that is done.

CHAIRMAN DAUM: Okay. So I would be happy to have comment on this question from people who -- we don't need to hear from every single person, but people who want to address this question, fine.

Maybe, Ms. Fisher, I'll bet you have a comment on this. Would you like to start?

MS. FISHER: Well, I think my comments probably have covered that territory.

CHAIRMAN DAUM: Okay.

MS. FISHER: I think we have to be extremely careful about how this vaccine is marketed because I do think it will result -- it could possibly result -- in the compromising of trust in the labeling that the FDA puts on the vaccine.

CHAIRMAN DAUM: Anybody else?

Dr. Parsonnet, we'll just go around. Welcome to pass or comment, as you wish.

DR. PARSONNET: No, I don't have anything to add. I think, again, being less specific about the indications, about what specific things that Strep. pneumoniae causes or saying to a lesser extent this affects otitis media, either of those or some other alternative of that would be fine.

DR. MARKOVITZ: Yeah, I'd agree with dr. Parsonnet.

DR. GOLDBERG: I would agree with the comments that were just made.

DR. DIAZ: The same.

DR. FAGGETT: Yeah, I kind of agree with previous comments. I think some wording like the vaccine with its proven efficacy against invasive disease is also useful against otitis media. I think something like that would be a possibility.

DR. OVERTURF: It would seem to me that there's a requirement here for the manufacturers who are going to accept the responsibility for disseminating this information also to disseminate effective education.

And because this is going to be a requirement, patients will come in and ask for the otitis media vaccine, and I think that's going to have to be explained.

I think the efficacy that we have said we'd demonstrate is going to have to be explained, and people are going to have to have realistic interpretation of this.

So I think in addition to how you label the vaccine, the educational component is going to be extremely important here. Part of that starts with the labeling of the vaccine, but it's going to require one step further as well.

I think also certain kinds of venues for marketing of this vaccine with this indication are probably not appropriate, and I don't think they should be approved, such as direct marketing. I think that would be a mistake because it doesn't give the practitioner the opportunity for the educational piece.

CHAIRMAN DAUM: Mimi, comments?

DR. GLODE: I could see things going wrong in two directions. If a parent -- again, if the educational system wasn't as it should be and someone said, "I really don't care about this vaccine for my child. I barely know what meningitis is, but I sure want it for ear infections," that's a disservice to the country and to the children in the country.

Similarly, if it's marketed in any way as an otitis vaccine and then loses credibility because it doesn't prevent otitis in the majority of children, that's a disservice. If in any way it inhibits one single parent from getting it to prevent meningitis and serious invasive infection. So I'm worried about that.

DR. SCHWARTZ: I would like to recommend that as part of post marketing, if it's at all possible, when any company does studies that involves tympanocentesis, to have a central repository for pneumococcal isolates to have serotyping done because that's not usually done for studies of acute otitis media; only identification of the organism and not serotyping.

But the only way we're going to know is by having those investigators who are adept at and perform usually for study purposes tympanocentesis to have such a central repository to keep tracks on what is happening with the serotyping of the isolates that we get from the middle ear.

DR. HAMILTON: I think one of the difficulties with the claim is that it's not in gray, but it's totally in black and white, protective against otitis media caused by S. pneumoniae. If that could be accompanied at all points by a statement that describes the amount of otitis media that's caused by S. pneumoniae against the vaccine would be effective. For instance, all pneumococcal isolates, vaccine related pneumococcal -- this is too complex for a label, but I think the public could understand a five percent reduction in otitis media because that is what is caused in the studies by capsular serotypes, represented in the vaccine.

DR. SNIDER: Dixie Snider.

The scientist part of me says that it is appropriate to communicate to physicians certainly that the vaccine has demonstrated efficacy in the range of, well, 57 percent efficacy against the serotypes that are included in the vaccine, and that overall reduction of six to seven percent in the incidence of acute otitis media have been observed in two trials or something to that effect.

I guess there's another part of me that is a public health policy type who wonders if going beyond that is really wise, and I have concerns that others have expressed certainly about direct marketing, but about really promoting this even to physicians who are overwhelmed, and certainly to parents to give the impression that it's some sort of panacea for acute otitis media.

So I'm very much in favor of communicating accurately the scientific information to physicians that wouldn't object to communicating it to parents along the lines that Dr. Hamilton just mentioned. But I have some real concerns about going much further with this information.

CHAIRMAN DAUM: Dr. Katz.

DR. KATZ: I'm surprised Dr. Snider didn't say what I was going to say, which is that there are three groups who spread most of the information about vaccines in this country. One is the Committee on Infectious Diseases of the American Academy of Pediatrics. One is the Advisory Committee on Immunization Practices of the Centers for Disease Control, of which Dr. Snider is the Executive Secretary or something like that. And the third is the Committee of the American Academy of Family Physicians.

I think that they are very knowledgeable, very judicious individuals, the members. I feel quite confident that they will not alter their recommendation statements to say that this is a vaccine for otitis media.

In fact, I hope that in their next statements about the vaccine, they may say there may be a fringe benefit of this vaccine which we recommend universally for children to prevent invasive disease, that maybe it reduces otitis media by six or seven percent, but that its primary justification for use is to prevent invasive disease and not to be fooled by thinking of it as an otitis media vaccine.

So I am not worried about the educational aspects because I think the groups, the responsible groups who promulgate vaccine recommendations and to which 55,000 pediatricians and a 100-plus thousand family physicians listen very carefully, will not push this as an otitis vaccine.

CHAIRMAN DAUM: Thank you, Dr. Katz.

Dr. Stephens.

DR. STEPHENS: Being last this time, I --

CHAIRMAN DAUM: I'm last.

(Laughter)

DR. STEPHENS: Almost last, that's right.

I agree with almost every comment. I think it is a mistake on the part of the company and the FDA to market this as an otitis vaccine, otitis media prevention vaccine.

I think that the package insert should try to deal with this issue in a way, separate from a clear indication that this prevents otitis media.

Whether that be the epidemiology section or another section, I appreciate the comment that you don't like to do that, but I think in this instance it might be worthwhile reconsidering.

CHAIRMAN DAUM: And as the last person, I would just like to emphasize and elaborate slightly on some of the points that have been made.

In addition to the groups Dr. Katz mentioned as formulating policy and reaching consumers and providers, there is, of course, a fourth group and that is the media. And I think that what this vaccine, that road was already started down when it was first licensed. There are a number of media contacts by myself, and I'm sure many others at this table, about the new vaccine that has just been licensed that prevents otitis media.

I think we are coming full circle now and visiting it with actual data, data that have been published and are already available in the medical literature, by and large.

I am very worried that the notion go forward from this meeting that we have established today that this vaccine prevents otitis media to an important clinical degree. I do believe that the vote for efficacy is correct, but the overall impact on otitis media, particularly at the individual consumer level, is going to be very small.

We are already immunizing every child in the United States, at least with intent to treat in statements of committees. I just don't think that there is a reason why a child should come forward and say, "I want this vaccine because it is the otitis media vaccine."

And I would urge the company and urge the agency. I don't completely understand the process by which things are added to labels or given indications, but I would almost not push for an indication for this and I certainly wouldn't push for a marketing and promotional blitz based on what we have heard today.

I think this is important information. I agree with Dr. Katz and others. This is information to build on, understand otitis media biology better, but to be very careful about how it is represented to the public.

I don't think that we have an otitis media prevention vaccine yet.

I think we are done with the otitis media portion of today's meeting. I would like to propose that we take a 15-minute break, but before we do that let's have the open public hearing on acute otitis media, could we?

Is there anyone in the audience that wishes to speak to the committee about acute otitis media?

(No response.)

CHAIRMAN DAUM: That being the case, we will take a 15-minute break and reconvene at exactly 2:45.

(Whereupon, the foregoing matter went off the record at 2:37 p.m. and went back on the record at 2:48 p.m.)

CHAIRMAN DAUM: Okay. We would like to call everyone to order, please.

The next item on the agenda today is a committee update with respect to the GSK, Lyme disease vaccine, LYMErix, followed by an open public hearing in which we are aware of nine individuals or organizations that wish to be represented there.

In the interest of expediting the sequence and allowing everyone to be heard, I'd like to ask the nine individuals, who are on our agenda as scheduled to speak, to come down into this area on the side, if they would, so that they can come up to the microphone at the time we announce them.

They are currently Karen Vanderhoof-Forschner, Norman Latov -- I hope that I'm not butchering anyone's name. I apologize if I am -- Mark Geier, David Geier, Stephen Sheller, Lonnie Skall -- Lonnie Skall has canceled -- Kathy Shepanski, Pat Smith, and Jenny Marra.

So thank you very much to those individuals for accommodating us.

I would like to now call on Patricia Rohan, who is already at the microphone -- thank you ‑- for a committee update on the GSK Lyme disease vaccine.

Dr. Rohan.

DR. ROHAN: Good afternoon and as Dr. Daum mentioned, I would like to briefly update the committee on the status of LYMErix, Lyme disease vaccine. You may be aware that LYMErix was voluntarily withdrawn from sale earlier this year. The sponsor, GlaxoSmithKline, halted distribution and made their announcement in February 2002, citing poor sales as the reason for this decision.

GlaxoSmithKline further recommended that no additional vaccinations be administered, particularly for those considering initiation of the three-dose vaccination series.

Clinical trial vaccination was ended. The information was disseminated in a series of letters to doctors, to investigators, and to distributors. There was a mechanism provided for refund for returned vaccine.

I'd like to turn now, for a moment, to update you on the status of the Phase IV safety study for LYMErix. This is based on an interim report, not an interim analysis.

As you may recall, may or may not recall, the overall goal was to detect rare, but significant adverse events that are associated with product use that may not be recognized in studies of the sizes typical for pre-licensure studies.

In the original plan, there were 25,000 adults who would be aged and gender matched to 75,000 unexposed controls accrued over a two year period at the Harvard Pilgrim Health Care HMO.

Events are identified using ICD-9 billing codes and include both ambulatory and in-patient claims data.

Outcomes are confirmed by blinded review. Record review is completed by the appropriate sub-specialist and established diagnostic criteria are used where applicable.

The incidence of predefined adverse events in the exposed cohort are compared to the incidence in the unexposed cohort.

The primary endpoint for the study is new onset inflammatory arthropathy, and other endpoints include selected neurologic disorders, Lyme disease, rheumatoid arthritis, allergic events, hospitalization, and death.

Due to low accrual rate, additional HMO sites were added late in 2001. These include the Tufts HMO System in New England and Health Partners located in the upper Midwest.

In addition, the accrual period was extended to three years.

This is a recap of the accrual at various time points. The initial planned accrual of 25,000 vaccinees. After nearly two years, 2,568 vaccinees, just a little over ten percent of what had been anticipated, and with the additional activities that I have just mentioned, there are now 7,643 vaccinees with unexposed matched controls.

This is a breakdown of those 7,643 vaccinees and their controls by site. So you can see that the Harvard Pilgrim Health Care System has continued to accrue over a thousand subjects in a little over the last year, year-and-a-half. The Tufts System and Health Partners are each now contributing close to 2,000 vaccinees.

Now for those subjects accrued to date, we have the ICD-9 codes, the events that have occurred since the time of their first LYMErix vaccine onward. There are 847 musculoskeletal events that have been reported in the vaccinees and 2,063 in the unexposed.

There are various levels of review that these events then go through. The first level by registered nurse takes out many events, primarily trauma and injury, which is the bulk of musculoskeletal events, and then goes to a second level of review by a rheumatology fellow.

After that level, possible inflammatory arthropathy is then sent on to a rheumatologist, who uses an adjudication form to determine new onset inflammatory arthropathy.

As you can see, there are seven new onset inflammatory arthropathy cases in vaccinees and 15 in the unexposed as of the eighth quarterly report, which was submitted to us early this year.

I also wanted to point out that the other numbers that you see on the table, except for the bottom row, these are events not people. The bottom row is how many people with those events. So it is hard to make a direct comparison at this point.

In conclusion, GlaxoSmithKline has committed to full safety follow-up for all ongoing adults and pediatric clinical trials. They will also complete the Phase IV post-marketing study with full four year post-vaccination follow-up. That is slated to be completed in the year 2006.

In the meantime, we will continue to monitor IND and VAERS reports for safety issues.

Thank you for your attention.

CHAIRMAN DAUM: Thank you.

Are there committee questions for Dr. Rohan? Clarification issues?

(No response.)

CHAIRMAN DAUM: Dr. Rohan, we thank you very kindly.

We will now turn to the open public hearing portion of this session. There are now, as I understand things, eight individuals who have requested time to speak. I'd like to ask them each to limit their comments to five minutes. We will time them and provide input for you by this little traffic light device sitting on top of the projector which will turn green when you start, orange after you have spoken for four minutes, and red after you have spoken for five minutes.

And I thank you very much for cooperating. We look forward to hearing your comments. In advance, collectively, we thank you very much for taking the time to come today.

Ms. Karen Vanderhoof-Forschner. I hope I'm pronouncing your name correctly.

MS. KAREN VANDERHOOF-FORSCHNER: Yes, that's fine.

CHAIRMAN DAUM: I apologize if I'm not.

MS. KAREN VANDERHOOF-FORSCHNER: Thank you.

Okay. I'm Karen Vanderhoof-Forschner, President and Chairman of the Board of Directors of the Lyme Disease Foundation, or LDF, established in 1988.

The LDF is the only nonprofit meeting federal standards as a national nonprofit. We represent millions of people across the country and have a database of 85,000 supporters, that includes family researchers, business people, and government employees.

We have held 16 international scientific conferences; have state and federal public policy programs; coordinate a network of task forces and support groups; and have funded research programs that resulted in a 130 publications.

All members of my family, including my pets, are fully current in our vaccinations. Every year, I voluntarily take the flu vaccine and I have recently taken the pneumonia vaccine.

I am keenly aware that this committee takes seriously its duty to weigh the risks and benefits of each vaccine based on scientific data and the public need.

Today I am here in a continuing role to keep you informed of additional science relating to the safety and efficacy of the Lyme vaccine. I have given you a packet today, like this one, because I have no doubt that you will see the LYMErix vaccine back in the marketplace with the current or a different manufacturer for adults and pediatrics.

In this packet you will see, on the top, a patent for a safer Lyme vaccine. This is the United States version of the safer vaccine as compared to the current OspA vaccine filed in March of 2000, ten months before this committee held its special Lyme disease vaccines last January.

I have discussed these patents with the Advisory Committee members who felt that their recommendations would have been significantly different if they were aware of this and other material that I had given your committee last November.

Also in this packet is a 1997 memo to the Lyme Disease Foundation from SmithKline Beecham guaranteeing us that anyone who tested positive for Lyme disease would be excluded from the trial, and nobody would be included in the trial unless they had had their test run beforehand. This was not true at that time according to FDA documents and corporate documents.

Why will the public and scientists not necessarily believe the VAERS data that has been presented here today? Probably because science indicating opposite scientific conclusions are not presented at the same time or at other governmental forums, yet have been presented.

Because of the personnel from the CDC, which is Ned Hayes and Dave Dennis, have a perceived conflict of interest because they sat on SmithKline Beecham's private data safety monitoring committee and were at the same time in charge of the CDC's working advisory committee, working group on the ACIP's recommendations for use of the vaccine, people will wonder if this data is tainted.

Indeed, in this packet, you will notice five of the nine people on the ACIP working group had conflicts of interest.

One was an employee of one of the Lyme disease vaccine manufacturers, and the other was a private consultant to a second Lyme vaccine manufacturer.

And at least one of the CDC individuals in this that was a consultant to SmithKline Beecham was also involved in the VAERS analysis, leading to a public perception of a conflict of interest.

After my presentation, you will hear Lyme disease adverse event data from the LYMErix vaccine by Dr. Mark Geier, a world renowned VAERS data analysis expert. He will be followed by Dr. Norman Latov, a world renowned expert in peripheral neuropathy from Cornell, New York. And then, David Geier, regarding VAERS data analysis.

At a future date, we will be here to present research that directly questions the validity of the Western blot data used to determine which vaccinees did or did not get protection from the Lyme vaccine.

Thank you very much.

CHAIRMAN DAUM: Thank you very much, Ms. Vanderhoof-Forschner.

I show next Dr. Norman Latov; is that correct sequence? Okay.

Thank you, Dr. Latov, and welcome.

DR. LATOV: I'd like to bring to the committee's attention the occurrence of neurological sequelae following Lyme vaccination.

In the past several months, we have seen several patient who developed neurological impairment after the vaccine. Four patients had a demyelinating peripheral neuropathy documented by EMG and nerve conduction studies. Three had cognitive impairment with leukoencephalopathy and multiple white matter lesions in their MRI's. One had both neuropathy and leukoencephalopathy.

The syndromes are strikingly similar to those seen in patients with chronic Lyme disease who have had active infection in the past and treated.

In most of the patients the symptoms were presumed to be due to arthritis. Prior to a neurological evaluation the correct diagnosis was initially missed. So I think there are more patients such as these, but they really have not been evaluated properly.

By sequence analysis, the OspA protein in the LYMErix vaccine has three regions of homology, each consisting of six amino acids corresponding to brain cDNA sequences in the GENBANK database.

In addition, a human genomic database search revealed 16 additional regions of homology of six amino acids or more corresponding to genomic sequences. The observation suggests that the LYMErix vaccine may have induced an autoimmune reaction to a cross-reactive neural protein in the central or peripheral nervous system, resulting in neurological disease.

A similar autoimmune reactivity induced by infection might be responsible, in part, for the neurological manifestation of chronic Lyme disease.

Patients administered the LYMErix vaccine and their physicians need to be informed as to the possible development of neurological sequelae and be properly evaluated if the symptoms are present. It would also be important to examine patients vaccinated with the vaccine, who develop neurological disease, to determine whether they have T or B cell reactivity to the cross-reactive epitopes.

In addition, there is a need for studies to determine how to best treat these patients as they might have both an autoimmune disease and ongoing infection in some cases.

Thank you.

CHAIRMAN DAUM: Thank you very much, Dr. Latov.

DR. KATZ: Could we have a question?

CHAIRMAN DAUM: We don't usually do that, but we have time for one quick question sure.

DR. KATZ: Can you tell us did these patients developed their symptoms after the second dose, the third dose, or how long afterwards? Can you give us any time sequences at all?

DR. LATOV: Three of the patients developed symptoms acutely after the third vaccine which is a year after the first, with a severe flu-like syndrome, followed by weakness, paraesthesia, et cetera.

A couple developed after the second dose, never received a third dose. One patient started noticing mild symptoms after the first dose and then progressed thereafter.

So it is variable, but the ones with the most striking onset were after the third dose.

DR. SAMUEL KATZ: Thank you, very much.

DR. LATOV: Yeah.

CHAIRMAN DAUM: Thank you, Dr. Latov.

Dr. Mark Geier. Again I hope I am saying names correctly.

DR. MARK GEIER: I'm Dr. Mark Geier of the Genetic Centers of America.

I have spent the last 15 or 20 years working on adverse events in vaccines, and I would like to present a little data from a paper that we have recently had accepted for publication in a peer-reviewed journal.

Basically, we studied in the VAERS database. We compared the adverse reactions to those receiving TD vaccine, which is a vaccine that has been found to be causally associated with peripheral neuropathies by the National Academy of Sciences. And we also compared adverse reactions to LYMErix with the MMR vaccine, with the Rubella vaccine having been found, again by the National Academy of Sciences, to cause acute and chronic arthritis.

Next slide, please.

We did this to control for various things that need to be controlled for in the study of the VAERS. We found that when you compared TD vaccine to Lyme vaccine, there was a tremendously significant, and clinically significant, increase in the rate of total reactions, ER visits, life threatening reactions, hospitalizations, and disabilities, and an increase in death, but it wasn't large enough to be statistically significant.

Next slide.

When we looked at the severe adverse reactions comparing with TD and LYMErix, we found that there was a statistical increase in arthritis; chronic arthritis as defined by arthritis still around one year later since VAERS follows up at a year; neuropathy, chronic neuropathy; convulsions which were not significant; thrombocytopenia; lymphadenopathy; hair loss; and the whole list that is up here. Because of time, I will just show it you.

But basically, arthritis and neurological disorders were clearly compared to a vaccine that everybody admits causes some neurological disorders.

When we compared LYMErix to the rubella vaccine, looking specifically at arthritis, both chronic and acute, we found a very large increase of arthritis of both kinds compared to the rubella. And remember, rubella is a vaccine that again is widely accepted as causing in itself chronic and permanent arthritis.

So we found that it is very remarkable that this vaccine caused significantly more than a vaccine that we know causes arthritis.

We looked at a paper, this paper that is listed here. These authors looked at and found similar numbers to what we found, but they concluded that the LYMErix was generally well tolerated and that there were no, or very few, unexpected reactions. It all depends on what you expect.

I mean, here is a vaccine that causes tremendously higher rates of neurological reactions and chronic and acute arthritis than vaccines that are admitted to cause those things. And yet these were considered to be generally well tolerated.

Our data and our analysis of our data does not show them to be well tolerated at all, but rather very poorly tolerated.

The question to ask is how could they find that they were generally well-tolerated. I think that is just a point of view. I mean, they didn't cause anything that hadn't been seen in the studies, I guess.

But I think that our recommendation for current LYMErix vaccine is that either it shouldn't be reintroduced or if it is reintroduced it should be recommended only in cases where there is a very high rate of Lyme in the area and then only with informed consent. That is, the patients and the doctors have to be aware that there are high rates of adverse reactions, some of them very severe.

In addition, they have to be aware that there is treatment for Lyme. Lyme, although it is a bad disease, responds very well to antibiotics if they are given in a timely fashion and given in the correct amounts.

Our better recommendation is that we wait and introduce a better vaccine that doesn't have so many adverse reactions.

As a final statement, I would like to point out that I am strongly pro-vaccine. I just am interested in the improvement and in full disclosure of vaccine problems.

Thank you.

CHAIRMAN DAUM: We thank you.

David Geier, is our next speaker.

MR. DAVID GEIER: My name is David Geier and I'm President of Medcon, which is a company that analyzes adverse reactions to vaccines.

I don't have a conflict of interest in this.

What I am going to present to you briefly here is about epidemiology of the Vaccine Adverse Event Reporting System, or VAERS database. This is in light of what my dad just presented before.

As what you know, CDC has maintained VAERS since 1990. Adverse reactions are required to be reported to this database as commanded by U.S. law despite claims by other people.

Additionally, the CDC requires written telephonic communication of these reactions. The CDC additionally follows up these reactions to determine whether patients recovered from their reactions or not.

This is what we classify as chronic reaction, those patients who haven't recovered at one year following vaccination from their adverse reaction.

Additionally, the VAERS working group analyzes and publishes epidemiological studies based on VAERS. And Mark Geier and myself have published more than 20 articles in peer-reviewed medical journals analyzing VAERS for the types of adverse reactions he described following Lyme vaccine.

Additionally, VAERS working group reported how useful VAERS is.

Next slide.

VAERS database includes important information. It is listed up there. Of particular interests are the co-starts. These list the adverse reactions that were reported following vaccinations.

Additionally, it gives information about which vaccine was attributed to the adverse reaction that was reported and what year.

What we have done that is new is that we have used Microsoft Access, a relational database, to assemble the whole VAERS so we can analyze it at one time with one search. So we can analyze any of the fields that are found in VAERS.

Additionally, we used biological surveillance summaries compiled by the CDC to calculated the incidence rates of adverse reactions.

When doing that the question arises what does that mean because VAERS is complicated by under-reporting and erroneous reporting. So we have used vaccine control groups in order to alleviate this.

A vaccine control group is a vaccine administered to a similar age population as the vaccine under study.

So our hypothesis is that an unbiased search of VAERS database should yield non-statistically significant differences in the incident rates of adverse reactions administered to a similar age population because the inherent limitations in VAERS should apply equally to both vaccines as well as the biological surveillance summaries. Their limitations should apply equally to each vaccine.

Some of the terms that you saw in the slides previously presented on VAERS: we use relative risk. That is the vaccine under study by the control vaccine.

Trivial risk is just subtracting one from the relative risk.

Percent association is dividing the relative risk by the relative risk plus one.

The overall importance of these statistical kind of calculations is that we believe that if you have an adverse reaction, following a vaccine, in comparison to a control group with the percent association greater than or equal to 67 percent or relative risk greater than or equal to two or trivial risk greater than or equal to one, and that additionally these criteria are listed, that it's medically plausible for a component of the vaccine to cause the injury alleged; that the association between the vaccine and the alleged injuries was reported in the peer-reviewed literature; and the vaccinee suffered an injury which is medically accepted as a possible reaction; and that the injury occurred within a medically accepted time period; and the alternate causes were considered but otherwise limited; then more likely than not you can say the vaccine caused the alleged injury.

And as with the case with Lyme, we believe that each of these criteria has now been met.

Thanks.

CHAIRMAN DAUM: Well, thank you very much.

Stephen Sheller, please.

MR. BROOKS: Thank you, my name is Albert Brooks. I'm an associate of Mr. Sheller. Mr. Sheller has become unavailable and asked me to appear in his stead.

Chairman Daum, members of the committee, I want to thank you for the opportunity to speak here today.

I must express a bit of puzzlement at why we're here. It is so late in the afternoon at this time and this place given no questions being posed to the Advisory Committee and given the FDA's presentation.

But I do want to echo what has been said earlier and point out that this vaccine, from what I can see -- and we have been contacted. We are attorneys in Philadelphia -- we have been contacted by well over 500 people now who have experienced arthritis, general Lyme disease-like symptoms, and very, very serious neurological conditions, such as those described by Dr. Latov, including one patient with acute transverse myelitis who is now on a trach tube and a feeding tube and will in all likelihood be dead soon after vaccination with LYMErix, and that occurred within a week of her second vaccination.

This vaccine is really a cautionary tale about what happens when qualified recommendations for approval are made.

In 1998, there were a number of safety concerns that were expressed by the committee, albeit with a recommendation for approval.

In January of 2001, many of those concerns were revisited and the committee stated, by and large, that many of those safety questions had not yet been resolved, given two years of marketing.

The committee did make several recommendations about what should be done, many involving the dissemination of information to the public and to doctors about the ongoing safety questions.

In that past year, as far as I can tell, none of that has been done. Information has still not gotten to the public. And there is more information that has come out since then.

We have the information from Dr. Latov. We have the development in the VAERS reporting system. And we also know that the FDA is looking at the VAERS Reports, albeit in a very limited way, specifically with arthritis.

They have, as of November when they presented an abstract to the Rheumatology Convention, gathered records on 31 people who had full sets of records and arthritis complaints. Fourteen of those had physician-diagnosed onset of arthritis and that is a very strict definition.

Seven of those 14 could not otherwise be explained by preexisting conditions, family history, predisposition to autoimmune-related conditions and these are of quite a bit of concern to us.

Most importantly we have the vaccine being withdrawn from the market, suddenly, and at the very beginning of what is believed to be the tick season in most of the Lyme endemic areas, which raises substantial questions about the safety profile of this vaccine.

I think it is strains credibility for the manufacturer to maintain that it is being withdrawn because of poor sales due to lack of demand, especially in light of the CDC's recent analysis that the incidence of Lyme disease is higher than ever.

The reason there are poor sales -- and I think that poor uptake into the Phase IV study demonstrates that there have been poor sales -- is because this vaccine is not a good vaccine and it is hurting people.

And many people are reporting adverse reactions. The numbers reported to VAERS are remarkable, and they are only the tip of the iceberg.

I have talked to people who have gone for months after vaccination not realizing that arthritis is possibly related. Therefore, they never associated their arthritis with the vaccine.

I have talked to people who have said they are 40 and they feel like they are going on 80. People who say they felt like they were just getting old; that's a 35 year-old. I have talked to a client, who was a vice president of a major manufacturing company of clothing, a multi-million dollar salary per year, who is no longer able to work; would hold meetings in the afternoon and the next morning forget that the meetings occurred. She has now a lupus-like condition. I have talked to several people like that.

It is not enough for the vaccine to have been voluntarily withdrawn and the Phase IV study is not enough. We are calling on the FDA to actively solicit information from doctors and the public regarding arthritis and the more generalized Lyme disease conditions, as well as neurological symptoms, to really build a meaningful database.

It's not acceptable that this vaccine can come on the market for two years, be withdrawn, and leave injured people in its wake with really no answers.

The public health demands an answer. People who have been injured demand an answer. The treatment of people who have been injured and the treating physician's ability to treat those people demand an answer.

That is what we are requesting at this point. And we certainly hope that the disappearance of the vaccine from the market will not also bring with it a disappearance of an investigation of these conditions.

Thank you.

CHAIRMAN DAUM: Thank you very much.

Kathy Shepanski.

MS. SHEPANSKI: I'm Kathy Shepanski and I am a LYMErix victim. I received --

CHAIRMAN DAUM: Ms. Shepanski, excuse me. Can you speak right into the mic?

MS. KATHY SHEPANKSI: Oh, I'm sorry.

CHAIRMAN DAUM: Thank you. We want to be able to hear you.

MS. KATHY SHEPANSKI: All right. I'm a LYMErix victim. I received one shot in May of 1999, May 4th. By May 6, I was starting to feel like I had the flu and that was the only reaction that anyone told me that I would get.

I became very ill. I did not receive the second shot because they were deciding whether they were going to put me in the hospital on the day of my second shot.

I have been to many doctors that they don't even want to hear what you have to say.

Finally, I have gotten to a doctor that has diagnosed me with Epstein-Barr and rheumatoid arthritis. That's where I am.

I was perfectly healthy before LYMErix shot and now I'm not.

Thank you.

CHAIRMAN DAUM: Thank you very much, Ms. Shepanski. We hope you recover as quickly as possible.

Is Pat Smith here, please?

MS. SMITH: Thank you, Mr. Chairman and members of the committee.

My name is Pat Smith. I'm President of the Lyme Disease Association, an all volunteer association with five nationwide affiliates. It consists of patients and families of patients.

The LDA has provided funding for research from coast to coast, some published in peer-reviewed journals including JAMA.

Along with our Greenwich affiliate, we were recently honored at a luncheon by Columbia University for partnering with them in the establishment of an endowed chronic Lyme disease research center at Columbia.

We also co-sponsored a fully accredited medical conference for physicians with Columbia.

Working with legislators, we have developed a bill in Congress, H.R. 1254, which will provide $125 million for Lyme disease research, prevention, and physician education.

The association provided testimony to this committee in January of 2001 seeking a moratorium on the vaccine, but we felt that no action was taken by the FDA. And to that end, in January 2002, the LDA had a private meeting with the FDA's Center for Biologics Evaluation and Research, CBER. We brought along several experts to discuss the vaccine issue with FDA officials, including Karen Midthun, Susan Ellenberg, Peter Beckerman, Norman Baylor, Miles Braun, and Robert Ball.

It is my understanding that the committee has not received an update on that meeting, and I would like to present a quick update.

We presented Dr. Donald Marks, M.D., Ph.D., former Lab Director of Connaught, 14 years of clinical research and regulatory affairs experience in the pharmaceutical industry, including Director of Clinical Research in charge of the Lyme disease vaccine program at Aventis Pasteur; presented to the FDA and he was the leader of a competitive effort to manufacture a virtually identical vaccine.

Currently, his focus is diagnosis of adverse events from medications, vaccines, biologicals and medical devices. The LYMErix associated cases he reviewed included: arthralgias and arthritis, as well as complicated neurological problems, and include adverse events that are long lasting.

A summary of Dr. Marks' presentation follows.

Why more adverse events were seen after the vaccine reached the market? People receiving LYMErix after product launch lived in Lyme endemic areas. Many people may have had prior exposure and clinical or subclinical infection. In these cases, LYMErix could be triggering or reactivating the damage caused by old and presumably cured Lyme disease.

Pattern of symptoms experienced after LYMErix mimic patterns of prior infections in many individuals. In these patients, LYMErix-related symptoms seem to respond to antibiotics as did the initial infection, bolstering the theory of disease reactivation.

Issues which confuse the vaccine picture. As proof of safety, the company inoculated arthritis-prone mice with OspA. But since the mice did not possess the HLA marker known to interact with OspA in humans, this rendered the experiment meaningless.

The company masked serious causally-related adverse events behind qualifiers, such as, quote, and which may have no causal relationship with the vaccine, unquote, and, quote, cannot be distinguished from the natural history of the underlying disease, unquote.

The company says that the possibility of severe rheumatological neurologic autoimmune adverse events is inherent in Lyme disease, attempting to shift the blame onto the patient and their illness, but does not inform physicians that the same adverse events can be separately caused by the vaccine, in addition to the symptoms of an underlying disease.

As a result of these actions, general practitioners in the U.S. were kept in the dark about the life-threatening side effects of LYMErix. Some basic problems: non-specific hyperactivation of the immune system, often evidenced through swollen hands or arthritis is an adverse event associated with LYMErix. This may be due to the presence of adjuvant.

This hyperactivation creates "dirty" Western blots in which multiple Lyme disease bands appear whether the individual has Lyme disease or not. The dirty banding makes it impossible for physicians to differentiate between LYMErix vaccination, new infection, or reactivation.

The net result is that cases of Lyme disease will go undiagnosed and untreated. Adverse reactions to LYMErix will be misdiagnosed as Lyme disease and people will be unnecessarily treated with antibiotics.

The vaccine manufacturer provides no warning to these possibilities.

The intention of FDA regulations is to provide a vaccine that is safe and effective. The intention of prescribing regulations is to provide sufficient information to prescribing physicians to enable safe and effective use of the vaccine. In both regards, SKB's actions appear to be contrary to FDA regulation and intentions and contrary to accepted standards within the vaccine industry.

The cases he examined, four of four neurologicals that he felt were related; 15 of seven rheumatologicals.

And just in conclusion, I would just say that I do have a copy of the remainder of my talk and I would ask that the committee does not drop this. We do not want this vaccine to be re-marketed or a similar vaccine without studies being finished.

Thank you.

CHAIRMAN DAUM: Thank you very much, Ms. Marra or Ms. Smith. Excuse me. I get confused.

Ms. Marra is the next and last scheduled speaker. Ms. Marra.

MS. JENNY MARRA: My name is Jenny Marra and I'm a Hospice nurse from New Jersey and I am a vaccine victim.

Some may remember me from January 31st meeting of 2001. I was here with a lot of different people that I have met that were also hurt by this vaccine.

I don't have the gumption; I don't have the health that I had then.

I want to put a face on these numbers that you are looking at. My life has been destroyed, completely. I have an elbow as big as a knee.

The symptoms came on a week after the first injection. The doctors, I have seen 17 of them. Three know about LYMErix and the problems it is causing. The others don't.

I have been diagnosed with fibromyalgia and depression. My husband is the same way. We both have fibromyalgia and depression at the same time. According to him, this is possible.

I don't understand. I am hearing from the VAERS investigation. I did the telephone investigation from the FDA. I don't understand the numbers, that they're not finding a connection.

When I know personally 133 people that have had this vaccine and that are hurt and out of them 121 of them are just like me, and we can't get help.

So if you can't find this connection, we're never going to have doctors to be able to find ‑- I'm not even looking for a cure. I just want treatment. I need -- I need help. I am in pain and I need help.

And I want to know if you people care. Do you?

You are the FDA. At times you could trust what you put on the market. I tell you what. I wouldn't touch anything that you put on the market anymore, not from what I've learned since taking this vaccine.

And I'm a nurse. I'm a nurse. I'm 43 years old and I'm going on 80 and I'm not alone. There's thousands of people just like me. Why can't you find the connection? I don't understand.

I'm not allowed to ask questions, am I?

CHAIRMAN DAUM: You're welcome to ask them. And we'll --

MS. JENNY MARRA: Can you answer me?

CHAIRMAN DAUM: I don't know.

MS. JENNY MARRA: Yeah, I did not think so.

I am in touch with several people in the FDA that are -- you have non-medical people doing these investigations on the phone. Have you once physically examined any of us?

We have doctors. They have no idea what's going on with us. All of us. I mean, I know 121 people just like me, with the same symptoms, the same problems.

You know, I have memory loss. I can't even use my right hand right now because I drove here yesterday from New Jersey. You know, there's so many people like me. You can't, you know, you can't -- for one, I keep hearing people talk about them putting it back on the market, God forbid. You know, enough people have been hurt.

I also heard that SmithKline is following up on the people that were in the studies. I know 11 people that were in those studies and they have been trying and trying to get SmithKline to help them, as they promised, and they are doing nothing. They are shutting them out. They are shutting the doors. It's over. It's over.

I just ask -- you know, I don't even know what I'm asking, except for help. I need help. I can't live on steroids. It is the only thing that does give us some relief when it flares up.

But I can barely walk mornings. Today I'm lucky. I only have loss use in my hand. Half the time it's my arms. I have memory loss. My husband is worse than I am. And I have a son that I can't play with anymore.

CHAIRMAN DAUM: Thank you, Ms. Marra, for sharing your touching story with us. We wish you a speedy recovery, as well.

I think we want to reassure you that the committee remains concerned about the safety and the efficacy of all the vaccines we discuss here and will continue to do so.

Is there anyone else that would like to address the committee at this time as part of the open public hearing?

(No response.)

CHAIRMAN DAUM: Ms. Fisher, did you want to make a comment before we close?

MS. FISHER: Yes. I had wanted to ask a question to the FDA, but didn't get my hand up in time when you were looking around.

I just, you know, I think it is very important for us to take seriously patterns. And that's what Ms. Marra was trying to communicate. We see this often.

Certainly at the National Vaccine Information Center, we see many patterns with the reactions that are being reported. I certainly hope that we will follow up on those patterns.

But I would like to ask the FDA: now that LYMErix has been voluntarily withdrawn from the market by the manufacturer, what is the process for LYMErix to be reintroduced in the U.S.?

Specifically, can this vaccine be reintroduced and used in children, without supporting data first being presented to this committee for a vote on safety and efficacy?

CHAIRMAN DAUM: Okay. Thank you.

Let's ask someone from the agency to respond.

Thank you, Dr. Midthun.

DR. MIDTHUN: This vaccine has never had an indication for a pediatric use, and so if that scenario were to ensue, clearly that indication would have to be sought.

As you know, we routinely bring these types of things to the Advisory Committee for your input.

CHAIRMAN DAUM: Thank you, Dr. Midthun.

And again thanks to the people who took the time, and effort, and energy to come to address up today. We assure you we will take your comments seriously and under advisement.

With that, I declare the meeting adjourned.

(Whereupon, at 3:35 p.m., the meeting was concluded.)