FOOD AND DRUG ADMINISTRATION
CENTER FOR BIOLOGICS AND RESEARCH
Vaccines and Related Biological Products
Advisory Committee Meeting
February 19, 2009
Hilton Washington DC/Silver Spring
8727 Colesville Road
Silver Spring, MD
This transcript has not been edited or corrected, but appears as received from the commercial transcription service. Accordingly, the Food and Drug Administration makes no representation as to its accuracy.
CASET Associates, Ltd.
Fairfax, Virginia 22030
Table of Contents
Opening Remarks - John Modlin, Christine Walsh 1
Topic 3: Clinical Studies with Pandemic Influenza Candidate Vaccines in the Pediatric Population in the Absence of an Influenza Pandemic
Introduction - Douglas Pratt 7
EMEA Perspective - Bettie Voordouw 23
21 CFT Part 50 Subpart D - Additional Safeguards for
Children in Clinical Investigations - Robert Nelson 36
GlaxoSmithKline Presentation - David Vaughn 59
Novartis Presentation - Theodore Tsai 87
NIH Presentation - Richard Gorman 113
Open Public Hearing 125
Committee Discussion and Recommendations 130
P R O C E E D I N G S
Agenda Item: Opening Remarks
DR. MODLIN: My name is John Modlin, from Dartmouth Medical School. Iím the chair of VRBPAC. I would like to call this morningís meeting to order and welcome all our members, voting members and non-voting members, who are joining us and, of course, all of our guests as well.
Iím going to turn things over to Christine Walsh, who has a few administrative matters to update us on.
MS. WALSH: Thank you, Dr. Modlin, and good morning, everyone.
Iím Christine Walsh, the designated federal officer for todayís meeting of the Vaccines and Related Biological Products Advisory Committee. I would like to welcome all of you to this meeting.
Todayís session will consist of presentations that are open to the public. I would like to request that everyone to check their cell phones and pagers and make sure they are off or in the silent mode. I would also like to just let the committee know for those of you who were here yesterday, we had the microphones switched to push-to-talk so it will not automatically, so you do need to push to talk. Also if the committee could please not leave your cell phones or your Blackberrys next to the microphones because I believe that was causing some static also. Thank you.
I would now like to read into the public record the conflict-of-interest statement for todayís meeting. This brief announcement is in addition to the conflict-of-interest statement read at the beginning of the meeting on February 18 and will be part of the public record for the Vaccines and Related Biological Products Advisory Committee meeting on February 19, 2009. This announcement addresses conflicts of interest for Topic 3, the discussion of the conducting of clinical studies of pandemic influenza vaccine in the pediatric population in the absence of an influenza pandemic. This is a particular matter involving specific parties.
Based on the agenda and all financial interests reported by members and consultants related to Topic 3, no conflict-of-interest waivers were issued under 18 USC 208(b)(3) and 712 of the Food, Drug, and Cosmetic Act. Dr. Seth Hetherington is serving as the industry representative, acting on behalf of all related industry, and is employed by Icagen, Incorporated. In addition, Dr. Hetheringtonís spouse is employed by GlaxoSmithKline. Industry representatives are not special government employees and do not vote.
With regard to FDAís guest speaker for Topic 3, the agency has determined that the information provided is essential. The following information is being made public to allow the audience to objectively evaluate any presentation and/or comments.
Dr. Bettie Voordouw is head clinical assessor for antiinfectives, Pharmacotherapeutic Group 1, Medicines Evaluation Board, The Hague, the Netherlands. Her spouse is employed by Pfizer in the Netherlands, but is not involved in the human vaccine shield.
In addition, there may be regulated industry and other outside organization speakers making presentations. These speakers may have financial interests associated with their employer and with regulated firms. The FDA asks that, in the interests of fairness, they address any current or previous financial involvement with any other firm whose product they may wish to comment upon. These individuals were not screened by the FDA for conflict of interest.
The conflict-of-interest statement will be available for review at the registration table.
We would like to remind members and participants that if the discussions involve any of the products or firms not already on the agenda for which an FDA participant has a personal or imputed financial interest, the participants need to exclude themselves from such involvement, and their exclusion will be noted for the record. FDA encourages all other participants to advise the committee of any financial relationships that you may have with any firm, its products, and, if known, its direct competitors.
DR. MODLIN: Thanks, Christine.
I would like to ask each of those who are seated at the table to briefly introduce themselves and their institutional affiliations. I think we will start on this side, beginning with Dr. Nelson.
DR. NELSON: Robert Nelson. Iím the pediatric ethicist with the Office of Pediatric Therapeutics at FDA.
DR. PRATT: Douglas Pratt. Iím chief of the Clinical Trials Branch in the Division of Vaccine Applications, Office of Vaccines, FDA.
DR. BAYLOR: Norman Baylor, director of the Office of Vaccines Research and Review at CBER, FDA.
DR. WHARTON: Melinda Wharton, acting director, Immunization Safety Office at the Centers for Disease Control.
DR. DESTEFANO: Frank DeStefano, senior research epidemiologist, RTI International.
DR. STAPLETON: Jack Stapleton, director of the Division of Infectious Diseases, University of Iowa.
DR. LEVANDOWSKI: Roland Levandowski. Iím an infectious-diseases physician. I have no institutional associations.
DR. GELLIN: Bruce Gellin, director of the National Vaccine Program Office at HHS.
DR. JOFFE: Steve Joffe, a pediatric oncologist at the Dana-Farber Cancer Institute and Childrenís Hospital in Boston. Iím the Dana-Farber hospital ethicist.
DR. FOST: Norm Fost, professor of pediatrics and director of the Bioethics Program at the University of Wisconsin.
DR. HETHERINGTON: Seth Hetherington, senior vice president for clinical and regulatory affairs, Icagen, Incorporated, in Research Triangle Park, North Carolina.
DR. ROMERO: Josť Romero. Iím the section chief of pediatric infectious diseases and professor of pediatrics at the University of Arkansas for Medical Sciences, as well as the director of their clinical research program.
DR. MCINNES: Pamela McInnes, National Institutes of Health.
DR. GILBERT: Peter Gilbert, Vaccine and Infectious Disease Institute, Fred Hutchinson Cancer Research Center in Seattle.
DR. SANCHEZ: Pablo Sanchez, a neonatologist and pediatric ID at UT Southwestern, Dallas.
DR. EICKHOFF: Ted Eickhoff, University of Colorado, Denver, infectious disease.
DR. KLIMOV: Alexander Sasha Klimov, Influenza Division, Centers for Disease Control and Prevention, Atlanta.
DR. JACKSON: Lisa Jackson. Iím a vaccine researcher at the Group Health Center for Health Studies in Seattle.
DR. DAUM: Iím Robert Daum. Iím a professor of pediatrics at the University of Chicago, infectious diseases, a past member of the Pediatric Advisory Committee of the FDA, and a past chair of this committee.
DR. MODLIN: Thanks, Bob. I understand that Dr. Debold is stuck somewhere underground on the metro -- can of like Charlie on the MTA -- and will be joining us as soon as she can.
This morning the Vaccines and Related Biological Products Advisory Committee is meeting. We are joined by a few members of the Pediatric Advisory Committee, I understand. As Christine said, the focus of our discussion throughout the entire day will be the conduct of clinical studies involving children with regard to pandemic influenza candidate vaccines.
Dr. Pratt, I understand that you are going to lead with some introductory comments.
Agenda Item: Topic 3: Clinical Studies with Pandemic Influenza Candidate Vaccines in the Pediatric Population in the Absence of an Influenza Pandemic
DR. PRATT: Good morning.
As mentioned, the topic today is clinical evaluation of pandemic influenza vaccines in children.
This is an agenda for the dayís activities. I will be introducing the topic and provide some background for the discussion. Bettie Voordouw, who is a member of the Vaccines Working Party at EMEA, will follow with the European perspective on these issues. Skip Nelson will walk us through the Subpart D regulations, the additional safeguards for children in clinical studies.
There will be three sponsor presentations that will follow: from GlaxoSmithKline, by David Vaughn; from Novartis, by Theodore Tsai; and from NIAID, by Richard Gorman.
Then Iíll return later in the day to present some discussion points for the committee to consider.
My presentation this morning:
ē Will summarize the relevant epidemiology for H5N1 in relation to children. Iíll go over the definitions of pandemic vaccine and pre-pandemic vaccines, as FDA has defined them.
ē Iíll go over the Pediatric Research Equity Act.
ē Iíll introduce the Subpart D regulations, though Skip Nelson will go into those in far more detail.
ē Iíll go over age considerations for pandemic vaccines, touching on the allocation priority from the departmental guidance, as well as the seasonal influenza recommendations as they relate to age and children.
ē Iíll summarize the ongoing and completed H5N1 vaccine studies in children to date.
ē Iíll touch on some safety considerations.
ē Go over existing FDA guidance.
ē If time permits, Iíll preview the discussion points for the committee.
The WHO has been tracking cases of H5N1 since 2003. There have been 404 human cases of H5N1. Thatís updated as of February 2 of this year. Sixty-three percent of these cases have been fatal.
Most cases have been among people in contact with domestic poultry. All human cases have been in either Asia or Africa. No avian or human cases of highly pathogenic H5N1 have yet been reported in the Western Hemisphere.
These data are somewhat dated. They are from June of 2006. These were the last published data from WHO, breaking down the cases by age. One can see that most of the cases of H5N1 have been in children and in young adults, and relatively fewer cases among the older age groups.
With respect to severity of disease, the highest mortality is actually in the adolescent age group, 10- to 19-year-olds. You can focus here on the first column and the last column. The last column has the cumulative data. The 10- to 19-year-olds have a 73 percent case fatality rate for H5N1; also still high in the young adults; somewhat lower in the younger children and older adults, but still quite high.
At this point I would like to try to distinguish the indications for a pandemic vaccine versus a pre-pandemic vaccine. A pandemic vaccine, as FDA is using the term, is for active immunization of persons who are at increased risk of exposure to an influenza virus that has potential to cause an influenza pandemic, and then during a pandemic caused by the influenza virus subtype contained in the vaccine. We envision that these vaccines would be used under an emergency declared by the secretary.
By contrast, a pre-pandemic vaccine indication would be for active immunization of persons against an influenza virus subtype that has the potential to cause a pandemic, as a strategy for pandemic preparedness. The distinction here would be that this would be for use during the interpandemic period, before a pandemic is declared, for example, for population priming and boosting.
This distinction is important when considering vaccine development and the kinds of data one might expect for either indication. In a true pandemic, in an emergency setting, itís likely that any safety risks from the vaccine would be far outweighed by potential anticipated benefits from the vaccine. In a pre-pandemic setting, when one cannot know for certain the subtype or the strain of the future pandemic, the anticipated benefits are less certain, and one might expect more safety data to support the routine use in that situation.
Specific laws and regulations apply to clinical research conducted in children. The Pediatric Research Equity Act -- we refer to it as PREA -- was enacted in 2003. It was renewed in 2007 with the FDA Amendments Act. The intent of this legislation was to assure that children have access to safe and effective drugs, through proper testing for pediatric use. The regulation calls for a pediatric assessment for all relevant pediatric populations ‑‑ this is required -- and all applications or supplements for new active ingredients, new indications, dosage forms, dosing regimens, or routes of administration.
The pediatric assessment should contain data from pediatric studies for each age group for which the assessment is required. These data should be adequate to assess the safety and effectiveness for claimed indications in all relevant pediatric subpopulations and should be adequate to assess the dosing and administration.
FDA reviews these assessments. In fact, we have an internal FDA-wide advisory committee that participates in the review process. This is the PER (phonetic) committee. After review, based on provisions in the law, FDA may conclude that the assessment is adequate to support the safety and effectiveness in all age groups or -- item 2 here -- the findings in the assessment can be extrapolated from adults to children or between age groups, if the course of the disease and the effects of the vaccine are sufficiently similar.
Also FDA could, alternatively, defer studies, in which case they would be provided at a later agreed-upon date.
Finally, itís also possible that studies would be waived and never requested. There are specific conditions in which this would occur. For instance, if the necessary studies were impossible or impracticable to conduct or if there is evidence that the product would be unsafe or ineffective in children.
What have we done so far with respect to PREA? We have licensed one pandemic vaccine. This is the Sanofi Pasteur influenza virus vaccine H5N1, for the National Stockpile. The indication for that vaccine reads, ďFor active immunization of persons 18 through 64 years of age at increased risk of exposure to the H5N1 influenza virus subtype contained in the vaccine.Ē Of note here is that the indication is for adults; it is not for children.
Also of note here is the dosage and administration. The antigen is at a higher concentration than is in seasonal vaccines, and two doses are required, even for adults.
With respect to PREA, the pediatric studies required under PREA for this vaccine were deferred. The deferred status is to be reassessed upon review of data from a completed pediatric study. At the time that we met for the advisory committee to discuss this vaccine, this trial was mentioned briefly. The manufacturer has committed to submit these data to assess the safety and reactogenicity of the vaccine. The protocol is entitled ďA Randomized, Double-Blind Phase I/II Study of the Safety, Reactogenicity, and Immunogenicity of H5N1 Vaccine in Healthy Children Aged 2 Years through 9 Years of Age.Ē I ask you to make note of the age there because we will come to age considerations.
At this point, I would like to introduce the Subpart D regulations. Of course, children cannot provide informed legal consent, and a set of federal regulations were adopted to assure that research conducted in children preserves their rights and safety. These additional safeguards in Subpart D have applied to federally funded research since 1983, under the 45 CFR regulations. In 2001, the FDA regulations under 21 CFR were updated to incorporate these safeguards. These now apply to all FDA-regulated research, including research sponsored by manufacturers.
Again, Iíll just introduce these briefly. Skip Nelson will go over them in some detail.
Under Subpart D, research in children can be approved by an IRB, under four sets of conditions:
ē 50.51 states that if the research does not involve more than minimal risk, the studies can go forward.
ē 50.52 states that if the research presents the prospect of direct benefit for the individual subject, the research can be approved.
ē Under 50.53, if the risk represents a minor increase over minimal risk and the research is likely to yield generalizable knowledge about the subjectís disorder or condition, the research could be approved.
Iíll stop here and say that in the FDA briefing document, a somewhat restricted interpretation of ďdisorder or conditionĒ was taken, in that a healthy child would not have a disorder or condition and this might not apply. I think Dr. Skip Nelson will go over this in more detail, and perhaps a broader interpretation may apply here.
If clinical investigations involving children are not approvable under the three regulations that I have just cited, the research can still go forward if it presents an opportunity to understand, prevent, or alleviate a serious problem affecting the health and welfare of children. However, an IRB cannot approve that research. Under this provision, the trial must be referred to the FDA Commissioner for a panel review. The Pediatric Advisory Committee serves as the reviewing panel for FDA if these studies are referred under 50.54.
Once previously, a vaccine trial was referred under 50.54. I know a couple of members of the advisory committee have familiarity with this example. In 2002, a clinical trial of a live vaccinia smallpox vaccine in children 2 to 5 years of age was referred by an IRB. This trial was sponsored by NIAID. In fact, two IRBs approved the trial. One IRB referred the trial for panel review under Subpart D. They cited a lack of the prospect of direct benefit and also that determining risk in this situation was very difficult.
A panel of 10 experts was organized by the Office of Human Research Protections, and they provided written opinions regarding the approvability of the study. This was not a meeting in one place at one time. But these opinions are published and they are available publicly.
The panel did find that the research was approvable, and most agreed that it was approvable under the 50.54 regulation -- that is, that it provided the opportunity to understand, prevent, or alleviate a serious problem affecting the health and welfare of children.
Ultimately, this study was not conducted, because another means of addressing the availability of smallpox vaccine was pursued by the Department.
At this point, I would like to move on to age considerations. This figure is from the guidance from the Department on allocation of pandemic influenza vaccines in the event of a pandemic. It may be a little difficult to read from the back of the room.
In tier 1, which is the highest priority in a severe pandemic, that group includes pregnant women, infants, and toddlers; in tier 2, high-risk children -- that is, children with medical conditions that would place them at high risk -- as well as infant contacts. Then healthy children aged 3 to 18 are actually in tier 3 in this plan.
This slide summarizes the data from the earlier figure. Again, in tier 1 are pregnant women, infants, and toddlers, 6 months through 35 months of age. The plan determined that children under 6 months of age would not be a priority because there are no seasonal vaccines that are licensed in that age group, and there was thinking that there vaccines would not work in that age group. The way of addressing that was that household contacts of infants less than 6 months of age were considered high-priority in tier 2. Again, tier 3 is children 3 to 18 years of age who are without high-risk medical conditions.
So thatís one way of thinking about age of children in the event of a pandemic.
Also relevant to age of children as we discuss this issue are the seasonal recommendations for influenza vaccines. These are current from 2009, from the ACIP. The seasonal influenza recommendations say that the minimum age is at 6 months of age for the trivalent inactivated influenza vaccine and 2 years of age for the live, attenuated vaccine. All children should be vaccinated yearly from age 6 months through 18 years, and children receiving TIV should get .25 mL if they are aged 6 to 35 months and then .5 mL if they are aged 3 years or older. So there is an age break here. Children less than 9 years of age should receive two doses if they are receiving influenza vaccine for the first time or if they were vaccinated for the first time in the previous influenza season, but only received one dose. So again, we have two age breaks here at 3 years and 9 years of age.
Some clinical trials have already been initiated and some completed in H5 influenza vaccines in children. These trials were identified from a search of the Clinicaltrials.gov database. They are not all necessarily being conducted under a U.S. IND. Some of the information is supplemented from other public sources.
What I ask you to note here are the different ages that the different manufacturers are studying, again the different locations, and the sizes of these trials.
With the first trial noted here, the Sanofi trial in 2- to 9-year-olds, that is the trial that was mentioned earlier and was mentioned at the previous VRBPAC.
All of these trials so far are using the A/Vietnam/2004 subtype vaccine. Also of note is that some of these contain different adjuvants, either aluminum phosphate or aluminum hydroxide adjuvants, or in the case of GSK, their ASO3 adjuvant, and in the case of Novartis, their MF59 adjuvant.
Important to the discussion of risk in the context of the Subpart D regulations is the safety of the vaccines. I have listed here some safety considerations. These actually apply for adults, as well as children, but in the context of pediatric trials, we do note that a larger antigen dose may be required than is used for the seasonal vaccine. Certainly that was the case for the Sanofi vaccine that is approved.
They also may contain a novel adjuvant -- for example, the MF59 or the ASO3 that was just mentioned. Iíll note here that these so-called novel adjuvants are not components of any U.S.-licensed vaccine yet, though they may be in some vaccines approved in Europe. In the U.S., certainly we have no long-term extensive record of safe use that we would get in a large pre-licensure study or from a postmarketing experience.
There are theoretical concerns with respect to these adjuvants of possible nonspecific self-directed immune responses that might be stimulated. Acknowledging that theoretical concern, no serious safety issue has been identified, I think we could say, based on the published data. Of course, we have not reviewed these in the context of an application.
With respect to the live influenza vaccines, the live pandemic vaccines would pose risks associated with replication, shedding, transmission, and possibly reassortants.
In 2007, FDA did put out a guidance for industry on the clinical data needed to support the licensure of pandemic influenza vaccines. With respect to the pediatric indications, the discussion is quite brief, and Iíll read it here in the next couple of slides:
ďIt is anticipated that data will be collected in adults and in the pediatric population in a stepwise fashion. We assume that approval for use in the adult population, including the geriatric population, will be sought with the initial application. The amount of data needed for a particular sponsorís pandemic influenza vaccine to support approval for use in the pediatric population will depend on the available clinical data for that sponsorís U.S.-licensed seasonal influenza vaccine. The timing of the clinical development in the pediatric population warrants discussion with CBER.Ē
So thatís the extent of the guidance on pediatric trials that we published in 2007. Iíll note, coming back to this slide, that clinical data from the sponsorís U.S.-licensed seasonal vaccine -- that is somewhat problematic when the vaccines contain these novel adjuvants.
At this point, Iíll summarize the discussion:
ē Children at all ages are at risk for pandemic influenza, and we note that avian influenza has infected and killed children in Asia and Africa.
ē The timing, subtype, and clade/strain of the next influenza pandemic are not known. In fact, it may not be an H5N1 subtype.
ē PREA, the Pediatric Research Equity Act, promotes the proper testing to assure that products are safe and effective for pediatric use. However, extrapolation of data from adults and across age groups can address PREA requirements if the disease and effects of the vaccine are similar.
ē Subpart D provides an ethical framework for conducting the clinical investigations in children.
ē With respect to age, we have the seasonal influenza vaccine recommended dose and schedule, and we note that they do depend on the age of the child.
ē We have the departmental guidance on allocation of pandemic vaccine during a severe pandemic and note that infants and toddlers are in tier 1 and children without other medical conditions are listed in tier 3.
ē We also note that multiple manufacturers are investigating H5N1 vaccines in children, and the age groups being evaluated in the trials to date do vary among the manufacturers and the locations where these studies are conducted.
ē Finally, novel adjuvants used in some pandemic vaccine candidates lack extensive safety experience in children.
ē FDA guidance regarding pandemic influenza vaccine development for children does not address vaccines without a corresponding licensed seasonal vaccine or the specific age groups to be studied. They also do not address live, attenuated vaccines and the issues that might be relevant there.
So Iíll stop at this point. If time permits, I could preview the questions. Up to the Chair.
DR. MODLIN: Actually, I think that would be a very good idea. Why donít we go ahead and take a look at the questions the committee will need to discuss?
DR. PRATT: There are two discussion items, with some sub-discussion points.
Discussion point 1: Please discuss whether clinical studies in one or more pediatric age groups should be conducted using inactivated pandemic influenza vaccine candidates as part of pandemic preparedness.
If your recommendation is that no clinical studies in any pediatric age group should be conducted prior to use of an inactivated pandemic influenza vaccine in children, please discuss whether other data, if any, would be needed to support immunizing children in the event of an influenza pandemic.
Item 1, continuing: If your recommendation is that clinical studies should be conducted in one or more pediatric age groups prior to use of an inactivated pandemic influenza vaccine in children, please discuss your recommendations regarding which pediatric subpopulations should be studied and also the adult safety and immunogenicity that would be needed to support proceeding to pediatric studies with inactivated pandemic influenza vaccine candidates. In your deliberations, please consider the use of novel adjuvants and additional viral subtypes, other than H5N1.
The last discussion point: Please discuss what pediatric safety and immunogenicity data you would consider adequate to support licensure of an inactivated pandemic influenza vaccine candidate for use in one or more pediatric populations. In your deliberations, please consider use of novel adjuvants and other viral subtypes, in addition to H5N1.
DR. MODLIN: Thank you, Dr. Pratt, for very clearly laying out the issues that we will be discussing today.
In the interest of time, I think weíll move on to the next item, which will be the European regulatory perspective by Dr. Bettie Voordouw.
Agenda Item: EMEA Perspective
DR. VOORDOUW: Thank you very much for giving me the opportunity to say something about the European perspective here in this audience.
This is very new for me and, I think, for many Europeans, because we donít have a setting like this. So itís very interesting to be here.
First of all, I would like to say that also in Europe the discussion with regard to pre-pandemic influenza vaccines in pediatric populations is not at all a closed discussion. We are still in the middle of it. What Iíll be presenting now are the results of ongoing discussions over the last six to seven years.
What I would like to talk about is, first of all, very shortly, to say something about the European regulatory environment and then say something about seasonal influenza vaccines and how we apply those requirements for pre-pandemic influenza vaccines. I would like to stress in both cases what that that might mean for pediatric populations and what it doesnít mean for pediatric populations. Finally, I would like to say what we really donít know.
The pictures on the right have nothing to do with the presentation, but give you some flavor of the Netherlands, where I come from.
First of all, the European regulatory environment: The EMEA is the European counterpart of the FDA. I think all of you know the EMEA. The main difference between EMEA and FDA is that itís an administrative body that is responsible for all EU-wide regulatory activities. But in addition to that, as you can see there, we have also 27 national authorities, national competent authorities, because we are 27 EU member states. So next to European legislation and regulations, we also have national regulations. It is sometimes very difficult to combine the two.
Moreover, the EMEA uses the expertise from the individual member states. For example, for the committees that you see there below, like the Committee for Human Medicinal Products, CHMP, the body that is responsible for licensing human medicinal products, that expertise comes from all the different member states. But, for example, for the working parties, like the Vaccine Working Party, which I am a member of, that uses only specific expertise from some member states. In the Vaccine Working Party, we are from, I think, 10 or 12 different countries, and not 27.
So this is how the EMEA is constructed.
As I said, discussions concerning pediatric data in pandemic or pre-pandemic influenza vaccines are ongoing since we started talking about these vaccines. That was, I think, in 2002-2003. It was not until the EU Pediatric Regulation became active in 2007 that we got the tools in hand to bring the discussion much more actively forward.
The EU Pediatric Regulation led to the installation of the Pediatric Committee at the EMEA in 2007. The objectives of this committee are to:
ē Improve the health of children through the increase of high-quality, ethical research into medicines for children.
ē Increase availability of authorized medicinal products for children.
ē Increase information on medicines.
To achieve the above, it has to be done without unnecessary studies in children and without unnecessarily delaying authorization for adults.
It has a strong similarity with the Equity Act in the U.S. I think that much of it has been -- we looked very carefully to the U.S. situation.
This was the setting, seasonal influenza vaccines. In the EU, as in other countries, of course, we have an annual relicensure procedure for seasonal influenza vaccines. But in the EU, contrary to other regions of the world, we request a seasonal immunogenicity trial for every vaccine thatís licensed. That means that every manufacturer has to do a clinical study, an immunogenicity study, in 50 healthy adults and 50 healthy adult elderly, to support the new formulation. These are assessed based on the so-called CHMP criteria. They are defined according to the GMT increase, the seroconversion or a significant increase proportions, and the seroprotection rate.
Itís important to note that there are no validated criteria for children. These studies are also done only in healthy individuals over 18. We have no validated criteria for children at all.
The question, of course, is, what CHMP criteria are relevant for pandemic or pre-pandemic vaccines. Of course, when you look at seasonal vaccines and pre-pandemic vaccines, there is a difference in pre-vaccination antibody titers. We assume that most of the population is immunologically naÔve to a pandemic strain. There is a difference in vaccination or infection history. There is a difference in the use of adjuvants. Almost all pandemic and pre-pandemic vaccines are using adjuvants, and it will influence the primary response. Very important, there is no license for adjuvants in influenza vaccines in children.
I must say here that in Europe all the seasonal influenza vaccines have licenses for use in children. So itís different from here. But none of them has a license for adjuvanted vaccines, and none of them has any experience in children.
The choice of dose for pandemic influenza vaccine is targeted at availability instead of on the efficacy. There is a difference, of course, in the schedule with regard to priming versus revaccination and, again, experience in children.
Is it acceptable to extrapolate data to children? Itís evident from everything we have heard before that data in children are needed. Yesterday it became apparent; it was very clear from the discussions there that data in children are needed and appropriate influenza vaccines for children are needed. But itís very difficult to make the assumption that you can extrapolate seasonal data or pre-pandemic data to the use in children, because children are not young adults. They are immunologically more naÔve, but maybe not for the pandemic strains. In theory, from our perspective, they might be the best model for serological assessment of pandemic strains, if you would reconsider your thoughts.
Furthermore, we know the seroconversion rates in children increase with age, with less than 50 percent seroconversion in children below 6 months of age, and more than 80 percent will seroconvert after influenza vaccination over the age of 10 years. Thatís probably also the influence of natural priming.
It led for us to the idea at least that data from younger adults were probably most representative only for those likely to be naturally primed. In essence, that means that you might accept an extrapolation to older pediatric populations, maybe children over 9 years, but not in children younger than 9 years or younger than 6 years.
As I said, itís inevitable that data on children are needed:
ē As we have heard several times now, there is a difference in disease burden.
ē There are differences in clinical characteristics of the disease.
ē There is a difference in predictive benefit or assumed benefit.
ē Itís likely that immunological response will be different in children.
ē We do not have any dose-response data for influenza vaccines in children, also with the seasonal vaccines.
ē We have no long-term safety data in children, also for the seasonal vaccines.
ē We also donít have them on the adjuvanted influenza vaccines.
ē The question is, of course, are the seasonal criteria relevant for use in children? As I said before, they are not specific for children and they are not validated for use in children.
First of all, with all these limitations for use in children, how did we proceed for the pre-pandemic and pandemic influenza vaccines?
In Europe, different from here, we do have at some point to decide on the acceptability of a pre-pandemic or a pandemic vaccine and say whether it can be licensed, yes or no. When the manufacturer comes to us with a dossier, we have at some point to say whether it can be licensed, yes or no. We canít keep it in an IND.
Because of that, we installed two procedures, one for the pandemic vaccines and one for the pre-pandemic vaccines. For the pandemic, thatís called a mock-up dossier. Itís a dormant dossier and it means that the vaccine cannot be marketed in the pre-pandemic period. The indication will be based on official guidelines and the product can only be marketed after inclusion of the final pandemic strain.
For pre-pandemic vaccines, the situation is different. That is a full application. It means that the company is allowed to market the vaccine, but that will be up to the individual member states. One of my colleagues had a comment here and said that, depending on the decision of the member states -- for example, France has decided not to market a pre-pandemic vaccine, whereas Finland has decided to prime the whole population. So there are regional differences in Europe regarding that.
The other difference is that, because itís marketed, itís a full application, so the dossier and the requirements are different. The indication will be restricted to the age groups that are studied. So if there are no pediatric data, there will be no pediatric indication. Thatís different from a pandemic strain, because the national governments may decide to vaccinate the whole population.
The interesting thing is that in practice these vaccines are the same; only the procedures are different. So we have identical vaccines licensed with a pandemic and a pre-pandemic license.
Of importance also for the pre-pandemic vaccine is that they have to maintain their dossier with drifted strains. We get regular updates if things change.
To make our lives easier, we have developed a lot of guidelines. I wonít go into detail there. If anyone is interested, please go to the EMEA Web site and you can find them all, or on PDFs, if itís easier for you.
What do those guidelines, the two most relevant ‑‑ the pre-pandemic guideline and the pandemic guideline, the mockup guideline -- say with regard to getting data in children?
The pandemic vaccine guideline says that in a pandemic children may be vulnerable to infection, and so constitute a special target group for vaccination. Once data from adults are obtained, itís recommended that at least limited data on safety are obtained from healthy children. In a pandemic, priority should be given to assessment of immunogenicity of the pandemic vaccine in children.
The pre-pandemic guideline says that studies in children and adolescents are needed to evaluate the immunogenicity and safety after acceptable data are obtained from healthy adults. In principle, it says the data are needed prior to licensure. Studies in infants and toddlers should be only initiated when data from all the children and adolescents have been found acceptable.
Both guidelines say the data in different age strata are needed, and they recommend a stepwise approach.
What is the EU experience with regard to pandemic and pre-pandemic vaccines? As I said, we do have to license or we do have to say no. We have three licensed pandemic vaccines at this moment. We have M59 and ASO3, and we have a whole virion cell culture-based -- whole virion vaccine -- licensed in the EU as pandemic vaccines, and we have a few in procedure at this moment.
Furthermore, we have one licensed pre-pandemic vaccine, and again we have a few in procedure. We have several at the stage of scientific advisers pre-submission.
Itís very important that none of those licensed vaccines are licensed with pediatric data. There are no data in children in any of those dossiers. Itís interesting, because the guidelines for the pre-pandemic dossier said that those data had to be there. But they are not. That also gives the difficulty.
Most important, of course, is to get the level of data that you would need. Thatís mainly driven by the level of the safety database that we request. Thatís completely different, as I always understand, from what the U.S. expectation is.
We have very limited expectations for these vaccines. This is for pre-pandemic vaccines. We ask for adults from 18 to 60 years old. They have to exclude rare events. They are defined by an incidence of less than 1 in 1,000 vaccinated. That comes up with a sample size of about 3,000 vaccinated subjects with the pre-pandemic formulation. For specific age groups or specific risk groups, the uncommon drug reactions have to be excluded, which means a sample size of about 300 individuals. As I already said, we still donít have the data in children.
So far, for the use of children to define pediatric requirements, we define children between the ages of zero to 18 years. We know that there are differences in disease characteristics, immune responsiveness, and safety considerations. There are no data yet from pandemic or pre-pandemic licensed vaccines, but we also know that pre-pandemic trials have ethical limitations.
As you are here, we have also struggled with how to go forward. Maybe some of you know this picture, but maybe you donít. At this moment, many of us -- and I know the picture, so I know what to look for -- see a lot of white and black, but I hope that at a certain point we will see that there is a Dalmatian dog walking towards a tree in the snow.
Thatís actually where we should end: That we should somewhere see the picture on how to continue with pediatric data for pre-pandemic vaccines.
So now Iíll get to very quickly what our ideas are on how to go forward. I must say, this is still an ongoing discussion. We have been discussing it. Itís not finalized yet, so it may change. It may change because of discussions we have here. Maybe our views are helpful to your discussions here, but your views may also be helpful to our discussions.
So we have CHMP criteria for immunoprotection in adults and elderly for seasonal vaccines, but we use them also for the pre-pandemic and the core dossiers, the mockup vaccines, because we have nothing else at hand, and so we accepted the use of the CHMP criteria for adults and elderly.
For seasonal vaccines, as I said, we have no appropriate lab determinations. What we need is a pediatric plan for seasonal vaccines to establish the right dose and the number of injections. For that right dose and for those numbers of injections, we have to establish the efficacy and hopefully the correlates of protection, because thatís what itís all about, and to use that information to get criteria for immunoprotection in children for seasonal vaccines, and not for pandemic or pre-pandemic vaccines.
If we establish that, we hope that we establish the criteria for immunoprotection as we use them now for adults and elderly, but for the pediatric population. That would mean that we have the same situation as I had four slides before, but for children, and a situation where it is ethical to do these clinical trials.
Further on, if we do the trials for pre-pandemic vaccines, we use those same pediatric criteria, considering the fact that those criteria are established based mainly on a non-adjuvanted influenza vaccine. The same holds for adults and elderly. But at least we have something in hand, and now we have nothing in hand.
So we use the pediatric criteria of immunoprotection for the pediatric plan, and the pediatric studies should establish the immunoprotection data, and especially the safety data.
For the pandemic vaccines, then we say that we donít need any further data, that we accept the experience from the pre-pandemic dossiers.
In conclusion, since 2002, when we started off the discussions on how to proceed with pre-pandemic and pandemic vaccines, we all knew that there was really a need for common research protocols. As you see now, you see some different studies. Everybody does something, but it doesnít bring you further. So there is a need for common research protocols. They will speed up the timely development of vaccines and also will help the common schedules and recommendations.
There is a clear need for pediatric data and criteria for pre-pandemic vaccines, but also for seasonal vaccines. At least for the European situation, installation of the Pediatric Committee has freed up this discussion and give us a powerful tool to go forward.
I think we all would benefit from harmonized regulatory requirements for pandemic or pre-pandemic vaccines for EMEA, FDA, WHO, and other regulatory bodies as necessary.
But for the EU perspective, at this moment, we are focusing on pre-pandemic and pandemic vaccines, but relying on the seasonal vaccines to pave the way. It may be too late, but we have had seven years of opportunity that we didnít take. We hope itís never too late to start at some point. So we decided to go forward at this point.
Now I want to add special thanks to Daniel Brasseur, who is the Chair of the Pediatric Committee, who provided me those beautiful last slides.
DR. MODLIN: Thanks, Dr. Voordouw.
We do have some time for questions. It was a very clear and informative presentation. Letís see if there are questions from the committee or anyone else, for that matter.
I guess your presentation was so clear, there are none. So thank you very, very much. We certainly appreciate it.
The next presentation will be from Dr. Nelson, who, as advertised, will go through the Subpart D regulations in more detail for us.
Agenda Item: 21 CFR Part 50 Subpart D -- Additional Safeguards for Children in Clinical Investigations
DR. NELSON: Good morning.
Iím basically going to walk through what has been referred to as the Subpart D regulations and provide an overview, as you start discussing today the scientific and ethical issues involved in the development of vaccines to prevent pandemic influenza in children. This presentation will introduce you to the pediatric research regulations and the associated ethical principles which guide the design of pediatric clinical investigations.
Itís important to note that this presentation is not intended to turn you into an IRB -- hopefully, that would not happen -- nor is it to have you function as a federal panel, which Iíll get into at the end and point out the particular regulatory parameters around that panel, but simply to understand the ethical and regulatory parameters governing pediatric research to inform your discussion.
Iím going to cover the basic ethical framework initially and then walk through minimal risks and then what I call the ethical limits on pediatric risk, focusing on these two categories: minor increase over minimal risk, and then greater than minimal risk. I have given you the regulatory citations from 21 CFR 50. Iíll end up with a brief discussion of the referrals under 50.54.
The basic ethical framework of pediatric research is that research involving children either must be restricted to minimal or low-risk absent a potential for direct benefit to the child. I have given you some citations to the CIOMS Guideline 9, the ICH good clinical practice guidelines, which use the word ďlowĒ and was actually the format for the European regulations that were put in place in 2001, and then given you the FDA citations.
So absent direct benefit, minimal or low risk. Then, if there is the possibility of anticipated direct benefit, the risks must be justified that direct benefit. That risk/benefit balance must be at least as favorable as any available alternatives.
There is broad international consensus on this basic framework, although differences as to whether this is in a regulatory format or in a guidance format. The EMEA, for example, has a guidance that is similar to this, with some differences, given their regulatory context.
Letís start with minimal-risk research. I have put direct benefit, no direct benefit, and then the risk level, to give you a sense of the relationship. Thereís the citation. This is the definition: Minimal risk means that the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.
In the ethics literature, there is a large discussion of how you interpret this particular category. I should point out that the National Commission, which was the commission in the late 1970s that established this framework, defined minimal risks as those risks ďnormally encountered by healthy children.Ē That phrase was not carried forward in the regulations. Although itís not included in the current definition, most ethicists and additional federal panels -- the Secretaryís Advisory Committee for Human Research Protection, the Institute of Medicine -- agree with this interpretation.
ďRoutine immunizationĒ was used as an example by the National Commission to illustrate this concept, but I would suggest that the administration of an experimental vaccine is neither routine nor minimal-risk. So Iím going to go on past that category.
The other two categories are minor increase over minimal risk and greater than minimal risk. Iím going to start with the minor increase over minimal risk. Basically, these really are the two options, as I mentioned before. If, in fact, your investigational product presents greater than minimal risk, either the risk must be no more than a minor increase, absent the direct benefit, or it must be justified by the potential for the product to directly benefit the child and for that balance of risk and benefit to be comparable to alternatives. Again, those are the two categories available to us at this point.
Letís talk initially about this ďgreater than minimal risk, no direct benefitĒ category, known as a minor increased risk. These are, in fact, the three criteria that you will find in that.
The first is that the risk is only a minor or slight increase over minimal risk. One could argue that this implies that you have some idea of what that risk might be, in order to place it into that category.
The second is that the experiences are reasonably commensurate with actual or expected situations. This was meant to inform parental permission and child assent, although that as well is a point of discussion about how to interpret that category.
The final one is the yielding of generalizable knowledge of vital importance or understanding or amelioration of disorder or condition.
Iím going to talk a little bit in the next two slides about this ďminor or slight increase over minimal riskĒ and then this notion of ďdisorder or condition.Ē
A ďminor increaseĒ refers to a risk which, while it goes beyond the narrow boundaries of minimal risk, poses no significant threat to the childís health or well-being. I might point out, in the regulations themselves, there is no interpretation of ďminor increase over minimal risk.Ē What Iím giving you are quotations from the National Commissionís discussion behind that category.
The second is, given this conservative limit, the promise of substantial future benefits to children other than the subject does justify research that goes beyond, but only slightly beyond, minimal risk. That was the intent of the Commission in offering this particular category.
What about ďdisorder or conditionĒ? They didnít define it. They didnít define ďconditionĒ implicitly. You could argue that the notion of ďdisorderĒ was meant to be a disease. Again, the federal research regulations offer no definition either of ďdisorder or condition.Ē The only proposed definition that is on the table at this point was initially offered by the Institute of Medicine in 2004, modified slightly by the Secretaryís advisory committee. I give you the IOM definition, which basically says it is a specific or set of specific characteristics that an established body of scientific evidence or clinical knowledge has shown to negatively affect childrenís health and well-being ó meaning now ó or to increase their risk of developing a health problem in the future. This notion of ďat riskĒ Iíll call your attention to, because I will expand on that a little further as I walk through the regulations.
Actually, Iíll expand on it now.
Healthy children: In the vaccine arena, there is a lot of use of the term ďhealthy children.Ē The word ďhealthyĒ is not used anywhere in 21 CFR 50 and 56. I suspect, although I didnít do the research, that itís not found anywhere in FDA regulations at all -- but Iím not going to back that up with data at this point -- and can be misleading. A child can be healthy and at risk, which is a notion of having a condition. A child with a condition may not have the condition related to the research, and thus could be healthy relative to that research.
So my recommendation is that a more accurate designation, rather than using the word ďhealthy,Ē would be children with the disorder or the disease, or who are at risk for the condition which is the object of the research, or children without the disorder or disease or without being at risk, without having the ďat riskĒ condition which is the object of the research.
The word ďhealthy,Ē I think, is problematic, because often itís not clear which of those two definitions is being used at that time.
Letís talk now about this ďgreater than minimal risk, prospect of direct benefitĒ category, which is 21 CFR 50.52. Here are the criteria for approval, which you have heard before:
ē The risk is justified by anticipated direct benefit to subjects within each arm of the study.
ē The relationship of anticipated direct benefit to risk is at least as favorable as available alternative approaches.
Letís talk a little bit about ďprospect of direct benefit.Ē Very simply, the notion about a benefit being direct could be seen as itís my benefit, not your benefit. Itís a direct benefit to me, and not you, when you enroll me in the research. It results from the research intervention being studied, and not necessarily from incorporating other interventions into that protocol. Otherwise, we could fill up research protocols with a lot of health care and use that to justify the research risk of the intervention that is being studied. Thatís referred to as ďthe fallacy of the package deal.Ē
So the word ďbenefitĒ is often modified by ďclinicalĒ to indicate that direct benefit relates to the health of the enrolled subject.
I might also point out that the notion of ďprospect of direct benefitĒ is based on the structure of the intervention -- in other words, whatís the evidence in support of that? -- rather than simply the intent of the investigator to say, ďI think this is going to provide some benefit.Ē
I might point out, though, that the evidence in supporting the prospect of direct benefit would be weaker than evidence supporting efficacy. Otherwise, we end up in a circular situation: We canít initiate a pediatric trial until we have efficacy data, which is a nonsensical position. So whatever evidence one brings to bear on this notion of ďprospect of direct benefitĒ would necessarily be weaker than evidence that you would expect for efficacy.
You may well base this on a surrogate endpoint, which is a large part of the discussion in this arena -- immune response, which was illustrated nicely, I think, in the previous presentation. But again, what evidence do you have linking this chosen surrogate to clinical efficacy, which is, at the end of the day, what you are fundamentally interested in?
Hereís where the link to this notion of ďcondition,Ē under 50.53, is important. Whether a child may directly benefit from an intervention that prevents disease depends on whether that particular child is at risk for developing the disease in the future. Otherwise, there is no benefit of prevention, if you are not at risk for that disease. Thus, for a preventive intervention, both 21 CFR 50.52 -- in other words, the category that is directed at those interventions that offer the prospect of direct benefit -- and 50.53, which has the explicit language of ďdisorder or conditionĒ -- both require the child enrolled in the research to have the condition of being at risk for future disease.
In addition, 21 CFR 50.52 has this other language about the comparability of the risk and benefit. But both of them have this implicit notion of being at risk in the context of a preventive intervention.
It raises the issue about who is at risk for H5N1 influenza. This is simply a picture of the data that Douglas showed you earlier, to just illustrate this question. Itís not intended to answer the question. You are the ones that are going to have to answer this question.
I might point out, though, in applying the research regulations, the relevant population that is thought to be at risk, those who may be at risk for the disease, are the children that are enrolled in that particular protocol. So when we use the ďat riskĒ language, itís the children enrolled in that protocol that are the at-risk population that would be the focus of discussion for the purpose of the research regulations and guidelines.
Let me finish up with this notion of IRB referrals under 50.54. You saw the previous illustration that Douglas gave you for the smallpox vaccine. I might point out that that referral came in before the process Iím going to show you was, in fact, set up, since the Pediatric Advisory Committee did not exist at the time of that smallpox example.
An IRB may refer research involving children which does not meet 50.51-53 for federal review. The IRB is supposed to feel that the protocol is worth doing and would be a reasonable opportunity. But as mentioned, this research may proceed only if the FDA Commissioner, after consulting a panel of experts, and following public review and comment, determines either that one of the other categories is met and that the IRB referral actually was, to some extent, unnecessary -- that one of the other three categories applies -- or that indeed it presents a reasonable opportunity to further the understanding, prevention, or alleviation of a serious problem affecting the health and welfare of children and will be conducted in accord with sound ethical principles. As mentioned, the Pediatric Advisory Committee, which does have a permanent Pediatric Ethics Subcommittee as part of its structure, is by its charter the one that makes these recommendations to the Commissioner.
But let me show you some language in the National Commissionís report about why they thought it was worth setting up this process in the first place, which is seen as exceptions to this general framework. They recognized that, ďIn exceptional situations, dangers to children or the community resulting from a failure to involve children in research might exceed whatever risk is presented by that research. For instance, the threat of an epidemic that could be offset by developing a safe and effective vaccine might justify research involving risk greater than otherwise acceptable to establish safety, efficacy, and dosage levels for children of different ages.Ē
Thatís a direct quote, 1977.
They went on to say, ďThe ethical principles at stake are the moral obligation to protect the community or to come to the aid of certain sufferers within it and the moral prohibition against using unconsenting persons, at considerable risk to their well-being, for the promotion of the common good. These principles are of such moment and their observance so basic to a just and humane society that any debate about their application should be held at the most public level of discourse.Ē
Here you are: Even if you are not a panel, you are at least participating in that public level of discourse.
Let me leave you with a summary of this basic ethical framework, again with that question raised by the National Commission about exceptions to that framework: Risk involving children either must be restricted to minimal or low risk absent the prospect of direct benefit to the child, or must present risks justified by anticipated direct benefits to the child, and which are as favorable as any available alternatives -- again, with the question as to whether or not the research that you are being asked to ponder today fits within this framework or is an exception to that which could be conducted nevertheless.
With that, I thank you.
DR. MODLIN: Thanks, Dr. Nelson.
Letís ask if there are questions for Dr. Nelson, or comments. If anyone would like to discuss any of these issues, this would certainly be the time to do it, before we get into the more technical discussions a little bit later on.
DR. JOFFE: Thanks, Skip, for a terrific presentation. I want to just get you to say a bit more about the notion of a condition, who has a condition, and whether individuals who are at risk, particularly when that risk is spread population-wide, as I think it is in this case, can be considered to have a condition.
Thinking back to the SACHRP recommendations, which I didnít review specifically in advance of this meeting, I seem to remember -- and you will remember better than I do -- that they left open the possibility that the entire population of children -- Iím thinking about the pediatric recommendations from SACHRP -- the entire population of children could be viewed as having a condition by virtue of being at risk for something that might strike anybody in the population, and vaccines would be a prime example.
Can you clarify a bit more? You hinted that there were some subtle differences between the IOM recommendations and the SACHRP recommendations. Your slide from IOM suggested that one would need to have defined high-risk groups, as opposed to a population-wide view of being at risk for a condition. Can you say a bit more about that?
DR. NELSON: My comment about the subtle differences was not directed at that point. I think, consistent with the IOM recommendation, one could decide that a population is at risk as opposed to individuals within that population. I think one could think of examples of diseases where, if we fail to immunize the entire population up to a certain level of herd immunity, in fact we all would be susceptible to that disease. Measles might be one. There was, in fact, a case of measles at the University of Pennsylvania just the other day, among graduate students. So everyone is at risk that was in that building. So I think that would be very different.
The only difference in the definition was, there was a concern about using socioeconomic and other variables to place populations at risk. So they added something about living in a healthy environment, to try to address some of those issues. But it leaves open the interpretation. ďHealthy environmentĒ wouldnít necessarily mean that you are not at risk for the diseases that might exist. Obviously, diseases vary, depending upon the part of the world that you live in.
So I think that is an open question that it would be interesting to hear everyoneís thoughts on in the course of the day.
DR. MODLIN: Seth?
DR. HETHERINGTON: I wonder if you would, on slide 14, expand a little bit on one of the statements there, something to do with risks justified by anticipated direct benefit to subjects. Then in parentheses it says ďwithin each arm of the study.Ē How do you interpret that? Can you expand on that in the context of a placebo-controlled trial, where one arm of your subjects are not going to get any potential direct benefit if they are assigned to a true placebo. How does this fit in?
DR. NELSON: Briefly -- that was a whole topic of discussion at an Ethics Subcommittee meeting last June -- the notion of the placebo-group risk is one approach to that. Again, Iím not saying the regulations drive this approach. But one interpretation that would be consistent with this is that the risk to the placebo group is driven more by what they are not getting than it is by the risk of the placebo, so that you would limit the risk to that group to a minor increase over minimal risk, which is then more related to the question of what they would not be getting if they are not getting their experimental product. One certainly wouldnít have a placebo control that would be considered risky.
Thatís very separate from adding other vaccines in as controls, which I think are based on more the virtues of providing some active vaccine to the control population, as is often performed, say, with either a rabies vaccine or MMR or something that would be appropriate for that population. Itís not really a placebo-controlled trial, strictly speaking.
DR. HETHERINGTON: What about the procedures of participation in a protocol -- for instance, phlebotomy done on a routine basis for some interval of time, which I think represents more than the risk that a child would encounter as daily living? How does that fit into this?
DR. NELSON: I think most IRBs -- there are some differences of opinion, I think, but most IRBs would consider a certain amount of blood testing to fit within the minimal-risk definition.
DR. MODLIN: Bob?
DR. DAUM: Skip, thank you for a nice, informative presentation. I want to ask a question, where I sort of already know the answer, but I would like to hear you reflect on it. The situation where the IRB has deemed that it cannot rule and needs to go through the federal process that you outlined involving the Pediatric Advisory Committee -- could you just tell us what your current view is on how long that process takes? You could do it from the time you submit to the IRB or from several months later when the IRB actually realizes that it canít decide and needs to punt to the federal process.
DR. NELSON: Let me just go through the steps, because for me to say what I hope would happen versus what actually happens, I think, would not be useful. The steps would be the IRB referral process and then the acceptance, which becomes a little more complicated, as it was in the smallpox context of HHS-funded and FDA-regulated. Then it ends up as a collaborative process with the Office for Human Research Protections at HHS. At that point, obviously, we need to pull the panel together, which is the Ethics Subcommittee. Subcommittees cannot advise the Commissioner, so the Pediatric Advisory Committee then has to meet as well. That committee meets routinely three to four times a year, and some of the timeframe is the timing between when the next meeting is and when the referral comes in and how quickly we can get that process going.
After that, depending on the panelís recommendations, we generate a letter of the FDA Commissionerís determination, which, if it was not HHS-funded, would be the final step; if it is HHS-funded, the next step would be to go to OHRP, who would then write a letter to the Assistant Secretary for Health.
So that process -- I guess we can all speculate about how long it takes, but it has generally been on the order of six to 12 months.
DR. DAUM: Wouldnít you agree that sometimes it has actually been quite a lot longer than that?
DR. NELSON: Well, one could say the smallpox determination is still not made. But that was withdrawn by NIH because they decided to go in a different direction. But the determination on that one, I donít think, was ever actually made, to my knowledge.
DR. MODLIN: Dr. Fost?
DR. FOST: Just one brief comment on that and then another comment.
By the chair, Iím Chair of the Pediatric Ethics Subcommittee.
In some of the cases that went very long, it was at the local institution. That is, it took them a year to get their act together and debate whether to send it in or not, getting their materials together, and so on.
DR. DAUM: I totally agree with that, but long is long.
DR. FOST: Yes. I just want to go back to the subjectís-condition issue. Iíll come back to this later. I donít think itís necessary to go to that part of the regulations to approve the trials that we are talking about. But suppose one were in 50.52 and one had to decide whether the subjectís condition applied. I think itís important to remember why that is in there -- that is, the ethical reason for requiring that the subject must have the condition that you are doing the research on. This is a category, remember, for research that has no prospect of direct benefit. I donít think thatís true with these vaccines, but letís suppose you are talking about such a study. The idea here is that since the child canít, himself or herself, consent, you are trying to make a guess about whether he or she would if they had a moment of lucidity and was fully informed.
This was Richard McCormickís argument: That the child, if it was about his condition, he probably would consent, because he would want knowledge to advance about his or her condition.
If thatís the reason for it, if thatís the moral basis for requiring that it be about the childís condition, then children who are at risk for this kind of highly fatal flu -- itís not hard to construct an argument that, of course, they would want reasonable research, well-designed research, to go on that would protect them from this.
DR. MODLIN: Dr. Debold?
DR. DEBOLD: Going back to slide 14 again, you mentioned risk to the placebo group if they were getting another vaccine that could potentially confer benefits. What about the situation where the placebo is actually the adjuvant, which may not, in fact, confer benefits at all, but increases their exposure to risk of adverse events?
DR. NELSON: I have never been asked that question before, and Iím a little loath to just give an off-the-cuff answer. But I think you would have to evaluate the risk/benefit to that group where you are giving them the adjuvant without the antigen. I think itís a reasonable question. Without getting into the details on the data, I usually donít even answer questions like that off the cuff internal to the FDA, and Iíll doubt I will do it external, without getting into the data as well.
But that would be a question that would have to be addressed.
DR. MODLIN: I could take a quick shot at that. It may very well depend on the value of the data. The overall understanding of the use and application of the vaccine to that population would be the question, I think, that would need to be asked.
Dr. Nelson, let me ask you whether or not you or the agency considers the decision that was made with respect to smallpox vaccine to have in any way set a precedent, not necessarily in the legal sense, but that it does provide at least a precedent that we can work from in making these decisions as we go forward, with not only this vaccine, but perhaps other vaccines that are being tested in children in similar situations?
DR. NELSON: No.
DR. MODLIN: Dr. Joffe?
DR. JOFFE: I just want to make a comment in response to a question that Dr. Hetherington raised a little while ago with respect to the blood draws and other procedures that might involved in a study like this and whether those -- setting aside the exposure to the vaccine ‑‑ might be considered minimal risk.
In the regulations from OHRP, there is a set of criteria for what can be reviewed in an expedited fashion by an IRB. One criterion is that the study has to involve minimal risk. Then there are, I think, eight specific categories of studies that can be reviewed in an expedited fashion and donít have to go to a full board. One of those categories is studies involving blood draws in children, presuming that they meet certain thresholds of blood volume. So within certain bounds, they can be considered minimal risk.
So thatís already in OHRPís guidelines and may be helpful in realizing that at least some studies that involve just the blood draws might fall into the minimal-risk criterion. That already seems to be accepted in guidance and regulation.
DR. MODLIN: Any other questions or comments?
DR. NELSON: Let me just give one clarifying comment to Steveís remarks. The expedited criteria are, in fact, in FDAís acceptance as well. What it says is that it needs to be both minimal risk and on that list. Itís possible that an IRB might decide that being on that list doesnít make it minimal risk, but I think itís fair to say that the overlap between minimal-risk determinations and being on that list is probably nearly 100 percent.
DR. MODLIN: Thank you. Bob?
DR. DAUM: I would like to ask Skip to reflect on one other issue that I have been thinking about. It goes to the map of the world that you showed, with the cases of H5N1 that are currently known in humans. It also seems to me that one of the things that may be on the table here, but not said, is, is there a risk to children of H5N1 virus spreading this week, next week, this year, next year? Of course, the magnitude of that fear, threat -- I donít know what the right word to use is -- is not known. But it seems to me that itís a number that is not zero, in our minds. Otherwise, we wouldnít be sitting here discussing this. In other words, the whole discussion of whether we want to test H5N1 vaccines for a total theoretical idea that no one has ever thought about, has never happened before, couldnít possibly happen here -- we wouldnít be discussing that.
So we all must believe, at some level, that there is some admittedly indefinable threat of this virus coming to children in our country and European countries in the world. Itís that risk of disease -- not risk of procedure, but risk of disease -- that, it seems to me, is driving this discussion to a fair extent. We donít say that enough, because it goes to how we categorize the research, believing that there would be benefit if there is a risk of the virus coming and there would be no benefit if there is no risk of the virus coming.
Do you see where Iím going? I would really like to hear your comments on that. Just showing us a map like that, implying that there arenít any cases here, and therefore there is no concern, which I know isnít what you mean -- how do you deal in your deliberations about this with the level of concern of the virus coming, and therefore the need to protect our children?
DR. NELSON: I think thatís the question this committee is being asked to discuss. That map was simply a reflection of existing, current cases. It has nothing to do with the determination of risk, which is future cases.
DR. MODLIN: Thatís why weíre here.
Any other comments?
If not, I would like to thank Dr. Nelson for a very clear and helpful framework. It does provide a nice framework as we continue on with our discussions.
We are about 30 minutes ahead. I would like to suggest that we go ahead and take our break now. Letís return at 10:30 sharp, if we can, with the manufacturersí presentations.
DR. MODLIN: We come to the next portion of the meeting, which will be a series of presentations from influenza vaccine manufacturers. The first presentation will be by Dr. David Vaughn, from GlaxoSmithKline.
Agenda Item: GlaxoSmithKline Presentation
DR. VAUGHN: Good morning. Thank you for this opportunity to present GSK data for children who have received adjuvanted H5N1 vaccines and to provide a rationale for continued testing in children.
I hope you will understand that itís a little bit awkward for manufacturers to be presenting data on clinical trials in children so that you can decide if itís appropriate to conduct clinical trials in children. But we will do the best we can.
My presentation will touch on the following topics: First, indeed, children do suffer from interpandemic and pandemic influenza, and vaccination, for the most part, protects them. We will want to vaccinate children during the next pandemic, as outlined in this guidance document from HHS. However, the next pandemic could occur very quickly, with little time to test vaccines in children. It is the position of GSK that limited testing of pandemic vaccines in children should occur during the pre-pandemic period. This data could be very beneficial to pandemic planners, and perhaps reassuring to public health decision makers, health-care providers, and parents, to be working with a licensed vaccine with an indication for use in children, rather than to be working under an emergency-use authorization.
GSK and other companies have already conducted some trials of adjuvanted H5N1 vaccines in adults and children, and we will present some of that data this morning.
GSK believes that an ethical framework does exist for continued testing in children.
Influenza, as has already been pointed out, is common in children, even in non-pandemic years. Each year, 15 to 42 percent of preschool and school-age children are infected with virus. Of course, not all of them become ill, but many do. Hospitalization rates for children under 2 years of age are similar to those for adults 65 years of age and older -- as high as 450 per 100,000 in children in the first six months of life, 4 to 21 times higher for children with underlying chronic medical conditions. Of course, there are many more outpatient visits, 10 to 250 times more.
Fortunately, deaths from influenza in children are unusual, but they do occur, at estimated rates of around 2 per 100,000 for infants in the first year of life.
Chris Potter reports that there have been nine pandemics in modern times, meaning from 1700. There will be another influenza pandemic. We do not know when, we do not know the subtype, and we do know the severity. The 1918 pandemic was particularly rapid and lethal. Two percent of those who became ill died. There were an estimated 50 million people, ultimately, who died during this pandemic, two-thirds of whom died over approximately a six-month period.
If you look at this graphic from Dr. Potterís publication, you will notice that there is an axis break here at 3.5 million deaths. To capture this peak in 1918, this graphic would need to be 14 times taller than is shown here.
In terms of the impact of the 1918 pandemic on children, Alice Reid has reported recently that infant mortality increased by 50 percent in Derbyshire, England in 1918. Thatís compared to 1917 and 1919 to 1922. Denmark reported peak disease incidence in children 5 to 15 years of age, though mortality was highest in the children in the first year of life and in young adults, 20 to 34 years of age.
The pandemics in 1957 and 1968 were much milder by comparison. But even here, children were three to five times more likely to get sick than their parents or adults the age of their parents.
While we do not know which subtype of influenza A virus will cause the next pandemic, the current concern is H5N1, obviously, the reason for this meeting today. This graphic has already been shown. We know that approximately half of cases have occurred in children and that the mortality has been extremely high, in excess of 70 percent for children 10 to 19 years of age. We donít know why there are more children affected. It could be that they are more susceptible to infection, it could be increased risk of exposure to H5N1 virus, or perhaps there is some relative resistance to infection among older adults. Of course, it could be some combination of these.
A public health officialís worst nightmare has to be an influenza pandemic caused by a virus with the transmissibility of the 1918 H1N1 virus and the lethality of the current H5N1 virus.
Regardless of the disease burden, we need to have some assurances that vaccination will provide some protection. I wonít go through the details here. Whereas individual clinical trials to look at efficacy in children vary in terms of the results, there have been several meta-analyses that consistently show the benefit to vaccination and serve as the basis for current ACIP and AAP recommendations for yearly vaccination of children.
Vaccines not only protect children in the seasonal setting, but since children play a central role in the spread of influenza, also protect the greater community. Children are more susceptible to infection. They shed virus at high titer and for longer periods, up 10 to 14 days. Once they start walking, they are socially mobile and highly interactive.
Models my wife ran all suggest that vaccinating 20 percent of children in a community would reduce the amount of disease across all age groups by 46 percent; vaccinating 80 percent of children would reduce total cases by 91 percent.
Iím sure VRBPAC members are all very familiar with the movie that this graphic represents. It dramatically makes the point that there may not be adequate time for pediatric trials once a pandemic begins.
I wonít go into the specifics of this particular scenario, but just point out that the models suggest that virus may circulate in the United States unrecognized for 24 days and then increase dramatically in terms of frequency, up to 4.5 million new cases in a single day, just 61 days later.
Neil Ferguson suggests that a pandemic that starts anywhere in the world could be in the United States within one month. So we may be talking only about a four-month period before much of this is over with. If it takes four to six months to produce the first doses of strain-specific vaccine, you can appreciate the problem.
The good news is in the lower graphic, where the rapid deployment of a stockpiled vaccine, even if of modest efficacy, could dramatically change the outcome, particularly if children are targeted for vaccination.
As a part of GSK contributions to pandemic preparedness, GSK is utilizing alpha-tocopherol-based adjuvant system number 3, or ASO3, in combination with hemagglutinin antigen, using conventional production processes. We make the antigen in two sites. The D-Pan antigen is manufactured in Dresden, Germany, ďDĒ for ďDresden,Ē using the Fluarix process, which is licensed in the U.S. To date, we have safety data in more than 5,600 adults and 300 children aged 3 to 9 years. We will turn to that data momentarily.
This vaccine has been licensed in Europe, registered as Pandemrix, as a mockup vaccine for use during a pandemic, and Prepandrix for pre-pandemic use.
Q-Pan antigen is produced in Quebec Province of Canada using the FluLaval process. We have safety data in more than 3,600 adults. This vaccine has been shown to be immunogenically equivalent to D-Pan.
The potency of this vaccine comes from the adjuvant ASO3, which results in an antigen-sparing vaccine, at least 24-fold. This means that rather than using 90 Ķg of antigen only twice, we can use 3.8 Ķg of antigen. In theory, this could take the current 12.2 million courses of vaccine reported by Secretary Leavitt to be present in the U.S. stockpile as of September 2008 and extend that to 292 million courses of vaccine, approaching the population of the United States. As I will mention in a moment, we anticipate that the dose for use in children will be 1.9 Ķg of antigen, which would extend the stockpile even further.
With the adjuvant, the vaccine is highly immunogenic in adults 75 years of age and old. It prolongs the duration of measurable antibody. A single dose primes for a later robust anamnestic response many months later. In humans, we see cross-reacted antibody generated, both in terms of hemagglutination inhibition antibody and neutralizing antibody.
While we wonít have protective efficacy data in people until a pandemic occurs, there is an animal model, ferrets. We are able to demonstrate cross-clade protection in ferrets. That is, if you vaccinate them with a clade 1 antigen vaccine and then later challenge them with a clade 2 wild-type virus, they are protected.
GSK anticipates submitting a BLA for Q-Pan this year. The indication that we will be seeking will be for use in adults at increased risk of exposure to H5N1 virus. The submission will be based on the following:
ē First, dose-selection studies using D-Pan and Q-Pan.
ē A demonstration that Q-Pan is highly immunogenic, both in adults 18 to 64 years of age and in adults 65 years of age and older.
ē There will be Q-Pan-specific safety data for more than 3,600 adults and more than 1,000 adults 65 years of age and older.
ē An integrated summary of safety that will look at pooled data from D-Pan studies and Q-Pan studies.
While there is a large safety experience with inactivated split influenza antigens in children 6 months of age and older, there is currently very limited pediatric safety experience using the ASO3 adjuvant. European regulatory authorities, as we heard this morning, have determined that these vaccines should be evaluated in children, including infants, and should include an evaluation of the effectiveness of boosting. The viewpoint in the United States has been quite similar, at least until this morning. We have contracted by HHS to conduct pediatric trials with Q-Pan.
Letís turn to the data now. Following the completion of several D-Pan vaccine assessments in adults, GSK conducted a preliminary study, under U.S. IND, in 405 children aged 3 to 9 years. This was conducted in Spain in 2007 and 2008. The virus antigen used was A/Vietnam, a clade 1 virus. Each formulation -- and there were three -- was first tested in children 6 to 9 years of age and then, in consultation with an independent data-monitoring committee, or IDMC, also tested in children 3 to 5 years of age. By the end of the trial, 300 children had received H5N1 vaccine, D-Pan, and 105 children had received two doses of Fluarix as an active control.
The following slides will have data only from Phase A, which has previously been publicly disclosed. But there were three phases, Phases A, B, and C.
In Phase A, children received half the standard amount of hemagglutinin antigen and half the standard content of the adjuvant ASO3, which we refer to here as ASO3B. In Phase B, children received the standard amount of antigen and ASO3B. In Phase C, children received the standard amounts of antigen and adjuvant.
These green arrows indicate where the study advanced in consultation with the IDMC.
Based on the hemagglutination inhibition assay, the D-Pan vaccine was highly immunogenic. The data here is the proportion of subjects who had HI titers of 1 to 40 or greater. This is an HI level thought to be associated with benefit and in this particular study is synonymous with seroconversion rate, as all the children were seronegative at the onset of the study.
On the left half we have data using an HI assay with homologous HI antigen, A/Vietnam, clade 1, and on the right half, using heterologous antigen, A/Indonesia, clade 2.1. We have data for children 3 to 5 years of age and 6 to 9 years of age.
For the children 6 to 9 years of age who received Fluarix, there was no immune response to the H5N1 HI assay. For children who received the D-Pan product, 30 percent seroconverted by day 21 and 100 percent seroconverted by day 42, 21 days after the second dose of vaccine, which was given on day 21.
For children 3 to 5 years of age, 12 percent seroconverted by day 21 and 96 percent by day 42.
In an assay using heterologous antigen, the seroconversion rates were 71 percent and 74 percent.
Immune responses across clades support the preparedness strategy of stockpiling.
In terms of geometric mean titers for the children 3 to 5 years of age, GMTs were around 1 to 400, and for the slightly older children, in excess of 1 to 500. When tested using heterologous antigen, the GMTs dropped by about a log to about 1 to 60, which is still about 10 times higher than baseline.
Solicited local symptoms were similar between D-Pan recipients and Fluarix recipients, with the exception of injection-site pain, in the center of this fairly complicated graphic. This has also been seen in our adult studies. Subjects are much more likely to complain of injection-site pain with this adjuvanted vaccine.
For children 6 to 9 years of age who received D-Pan, 86 percent complained of injection-site pain, compared to 67 percent of subjects who received Fluarix. In the 3- to 5-year-old children, 61 percent complained of injection-site pain compared to 39 percent of those who received Fluarix in that age group.
There were some grade 3 complaints of pain in the D-Pan recipients and none in the Fluarix recipients. This is similar to what we see in the adult studies, with the possible exception that in adult studies where we give antigen only or even saline placebo, we usually see grade 3 complaints in 2 or 3 percent of adult subjects.
The duration of this injection-site pain -- the mean was two days in both the D-Pan group and the Fluarix group, meaning they complained of pain on the day of inoculation and then the next day, on average.
Complaints of general symptoms were fairly homogeneous between groups for the 6- to 9-year-olds. In the 3- to 5-year-olds, while overall the numbers of complaints were low, there were more in the D-Pan group than in the Fluarix group.
The IDMC did not raise any safety concerns throughout this trial. There were no SAEs in Phase A. Until recently, there were no SAEs in any phase of this trial. There was one SAE reported in a child who had elevated liver enzymes on day 0, the day of vaccination, prior to vaccination. Eleven months later, this child underwent liver biopsy for continued elevations, and a diagnosis of autoimmune hepatitis was made. This child has remained asymptomatic throughout this time, but still, as of February, this month, he does have elevated liver enzymes.
Thatís D-Pan data.
GSK suggests that a Q-Pan pediatric protocol can be appropriately reviewed by IRBs in alignment with Title XXI, CFR Part 50, as was discussed earlier this morning. Itís possible that an IRB could review such a protocol under Subpart 50.52, meaning they think the risk of an H5N1 pandemic is real and that HI antibody suggests the possibility of protection, or at least solid priming for protection with a subsequent dose.
However, we do not have definitive efficacy data for this vaccine, and we wonít until there is a pandemic. Currently, highly pathogenic H5N1 virus is not circulating in North America. So other IRBs may want to review this protocol under Subpart 50.53 or 50.54.
What I have on this slide is an outline of thoughts under 50.53. As we heard, there are four regulatory requirements:
First, the risk to study subjects must represent only a minor increase over minimal risk. This assessment could be made by the IRB, based on more than 9,000 adults who have been exposed to this type of vaccine. What is expected is self-limited, short-term reactogenicity.
Second, study subjectsí experiences in the trial must be reasonably commensurate with those inherent in their actual medical situation. Again, this the ďhealthy childrenĒ discussion, where they routinely receive childhood immunizations, some of which are fairly reactogenic, and venipuncture is also a fairly common childhood experience.
Third, the study must be likely to yield generalizable knowledge that is vital to the amelioration of the condition. Here we have the discussion of what is the condition. Some would say that children in North America are not at risk for H5N1 infection. The counterargument would be that the condition is susceptibility to an influenza pandemic, a condition that we all share.
There is the opportunity, if the studies are conducted in advance of a pandemic, to demonstrate that the vaccine is immunogenic in children, and therefore potentially protective. It gives us an opportunity to identify a proper dose. Also it reduces the risks associated with exposure to an untried pandemic vaccine. If we are talking about 61 days to the peak of the pandemic, we will want to vaccinate very quickly once the pandemic is declared, perhaps unless less than ideal circumstances.
Finally, adequate provisions must be made to solicit the assent of the children and the permission or consent of their parents or guardians.
To conclude, I would like to emphasize a few points:
ē First, children in the United States will be among the first to receive pandemic vaccines, in agreement with the HHS pandemic plan.
ē Second, a conventional inactivated antigen approach may not meet the public health needs.
ē There are potential benefits to the use of adjuvanted vaccines to address a pandemic. These include enhanced cross-protection, enhanced priming of immunity, and, because of antigen-sparing, the ability to rapidly deliver to all U.S. citizens, including children, vaccine.
ē GSK believes that the initiation of carefully planned and well-controlled clinical trials of novel pandemic vaccines is ethically appropriate and in the U.S. publicís best interest.
DR. MODLIN: Thank you, Dr. Vaughn.
We have time for one or two quick questions. Dr. Fost?
DR. FOST: Dr. Vaughn, with regard to the prospect of benefit, that would be highest in the areas where the risk of getting this kind of flu is highest. That would argue for doing the trials in the highest-risk countries, in Asia, et cetera. Do you agree with that, that there is an argument for doing the trials in the highest-risk areas?
DR. VAUGHN: There have been approximately 200 children who have suffered, many who have died, in at-risk regions. Iím not sure how high that risk is, even there. We are focused on seeking a licensure in the United States, so the FDA may want to see data in children from this region.
But, yes, a study could be conducted in other areas.
DR. MODLIN: Ted?
DR. EICKHOFF: Dr. Vaughn, could you reiterate what you said about duration of local and systemic symptoms, particularly the injection-site pain issue?
DR. VAUGHN: What I said was that for this pediatric study, the mean duration of injection-site pain was 2.1 days. This was both for D-Pan recipients and for Fluarix recipients. Two days would mean the day of inoculation and the next day. That would be the average. The other half of the children would complain of injection-site pain on the third day, maybe 48 hours after injection.
DR. EICKHOFF: Were parents instructed to treat injection-site pain in any fashion?
DR. VAUGHN: I donít know the specifics of that. I imagine that they were not given specific instructions to treat, but they certainly would be free to treat with paracetamol or acetaminophen.
DR. MODLIN: These immunization protocols often include specific recommendations for management of injection-site reactions.
DR. VAUGHN: There was no prophylactic administration. In fact, itís an exclusion criteria if children receive prophylactic medication on the day of vaccination.
DR. MODLIN: Josť?
DR. ROMERO: A question and a comment. Could you tell us about the fever? Was the magnitude of fever in the group that received the H5N1 vaccine higher than that in the group that received Fluarix?
The comment is, soliciting headache in this age group is a difficult symptom to elicit. There was a large meningitis trial study that failed to reach significance, in part because that symptom is very hard to elicit in that age group. Thatís the comment.
DR. VAUGHN: In terms of fever, there was actually more fever among Fluarix recipients -- slightly more -- than among D-Pan recipients in the 6- to 9-year-olds. For 3- to 5-year-olds, there was no fever in the Fluarix group, but about the same frequency of fever in the D-Pan group as in the 6- to 9-year-olds.
In terms of headache, yes, it can be difficult to elicit that. We only ask that question to 6- to 9-year-olds. That question was not asked to 3- to 5-year-olds, though it could be an unsolicited complaint.
DR. ROMERO: Let me just follow up. Perhaps I didnít explain the question well enough on fever. Was the magnitude of fever -- that is, how many kids had fever greater than 102 or 103, significant fevers -- in the two groups, were they comparable or were they not comparable?
DR. VAUGHN: There was grade 3 fever in a Fluarix recipient or two. I believe that is 39 degrees or higher. No fever over 39 degrees among D-Pan recipients. It looks like there was one in the 3- to 5-year-old age group.
DR. MODLIN: Dr. Debold?
DR. DEBOLD: The adjuvant, ASO3, would be new to the United States. Itís something that we havenít used in vaccines here. I understand that it contains squalene.
DR. VAUGHN: Yes.
DR. DEBOLD: And I understand that one of the concerns with those products is the potential for autoimmune illnesses, responses. This was, as I understand it from reading the documents that were provided to us, a relatively small study in children. To have one child out of about 100 with autoimmune hepatitis is something that I find concerning.
Can you talk a little bit about preclinical-trial information about this adjuvant, the extent to which you actually have experience with it in children?
DR. VAUGHN: I might just comment on the autoimmune hepatitis case. That was a preexisting condition, in that the child had elevated liver enzymes prior to the studyís start.
In terms of the squalene issue, I think that has been maybe of less importance since the WHO comment on it in 2006. They felt that concerns about squalene were unfounded. That was their final assessment, at the end of that publication. That was largely based on Novartis studies with MF59, which also contains squalene -- to not see an increase in anti-squalene antibodies in those subjects.
The last question was about our safety experience with ASO3 in children. You have seen it -- 300 children.
DR. MODLIN: Lisa, do you have a question?
DR. JACKSON: Just a point of clarification. The D-Pan pediatric trial was 300 total or Phase A was 300? The data we are looking at are from how many --
DR. VAUGHN: The total trial was 300. Phase A was 101 children.
DR. MODLIN: Roland?
DR. LEVANDOWSKI: Just one quick technical question on that study design. There is both ASO3B and ASO3A. I didnít quite catch what the difference between those was, why the study population -- you needed to do those two cells for those two different types of ASO3.
DR. VAUGHN: The ASO3 in this study is a simple one-to-one dilution of the standard concentration of ASO3. Itís half the content of the active components, the alpha-tocopherol and the squalene.
DR. LEVANDOWSKI: And thatís the difference between the B and the A?
DR. VAUGHN: Yes.
DR. LEVANDOWSKI: I see.
DR. MODLIN: Pam?
DR. MCINNES: Iím interested in your study age groupings -- just on this quick look, really not seeing differences in either safety or in immunogenicity that look meaningful between the 3- to 5-year group and the 6- to 9-year group. I think itís a very conservative approach that you took in terms of step-down. Do you think, if you were to do it again, you would feel you could now put 3- to 9-year-olds together, 2- to 9-year-olds together? What is the thinking about it?
DR. VAUGHN: The division in age here is largely due to the maturity of the child. We have different adverse events that we solicit in the younger age groups. Thatís why the data is presented separately. Now that we do have data in 300 children -- it would depend on the IRB and the regulatory authorities that we consult with, but, yes, I would think we could do that entire age range in a single study.
DR. MCINNES: So 3 was the youngest?
DR. VAUGHN: Yes.
DR. MCINNES: We normally think about 6 months to 2 years old. What was the thinking about 3?
DR. VAUGHN: This follows roughly the thinking for seasonal vaccine, where you use a smaller dose in children under 3. From children 3 to 8, you use two doses, and so on. So it roughly follows the age breakdowns in the recommendations for seasonal flu.
DR. MODLIN: Dr. Gilbert?
DR. GILBERT: Given the possibility -- and maybe likelihood, from what Iím hearing -- that a pandemic strain might be antigenically mismatched to a pre-pandemic vaccine, Iím interested in the issue of developing a standardized panel of avian flu isolates to evaluate. I suppose, ideally, that panel would be representative of the spectrum of strains that could potentially become pandemic viruses. By getting a fairly comprehensive assessment of the breadth of neutralization or HI titers in the study population, one could gain a better prediction of what the efficacy would likely be for a real outbreak.
I noticed in your approach, from what you have shown today, you have one strain that was a heterologous target. But Iím wondering if you could share any other thoughts you would have on the approach to getting a more robust assessment of the diversity of responses or the breadth of responses.
DR. VAUGHN: There are a number of H5N1 viruses that can be used in laboratories that are not high-containment, as they have been reassorted. At GSK, we are looking at a number of different H5N1 virus antigens for testing, both by HI and by neutralizing antibody. The example I gave is fairly extreme, in the sense that itís cross-clade, from clade 1 to clade 2. In experiments where we may look at clade 2.1 versus 2.2, there may be even more cross-reactivity in terms of antibody responses.
DR. MODLIN: Bob?
DR. DAUM: Just a quick clarification. In your slide entitled ďFramework for Ethical Review,Ē you mention that venipuncture is a common childhood experience. I would just like to get some information from you as to how common you think that is. In my experience, itís not very common, but I donít take care of healthy children in the community or well children in the community.
DR. VAUGHN: I donít have specific data for that. Using needles with children is fairly common, from the newborn nursery, with a heel stick or a finger stick or actual venipuncture, as we are doing here. Iím not sure the pain is that much different between those different approaches. Healthy children may have infrequent venipuncture. I donít have specific incidence or numbers for children in the U.S.
DR. DAUM: Might I suggest that itís a pretty rare childhood experience, and you might want to improve that language a little bit.
DR. VAUGHN: Well, my perspective is from a pediatrician, so is fairly common where I work. But I understand your point.
DR. MODLIN: Bruce?
DR. GELLIN: We spent most of yesterday talking about how to make better vaccines to provide better cross-protection for seasonal flu. Given what you have here with your adjuvant, can you comment about the companyís plans to use an adjuvant for seasonal flu? I guess you have already told us that this is the body of information in children, but you might want to tell us some more about these plans if they are to introduce that product and what the development plans would be, particularly to bring a product like that down to younger children.
DR. VAUGHN: Yes, the company does have plans to utilize the ASO3 adjuvant with seasonal antigens. There is a trial under way now in 43,000 elderly to look at that age group, and also plans to look at children, where the adjuvant may provide benefit or children donít respond so well to antigen alone.
DR. MODLIN: Melinda?
DR. WHARTON: Given that there is a more robust body of experience with the ASO3 adjuvanted vaccines in adult, could you share with us a little more information about the safety experience, particularly addressing the issue of some autoimmune conditions, which I think is the thing that raises at least a theoretical concern?
DR. VAUGHN: GSK is sensitive to the issue, the concern about autoimmune with new adjuvant systems, such as this one. We have come to agreement with CBER on a list of adverse events of special interest, many of which are presumed to be immune-mediated. Some are not, but many are. We are actively looking for those types of cases, not only in the Q-Pan program, but the D-Pan program and other programs that are using ASO3, such as the efficacy study in the elderly thatís under that I just mentioned.
Not surprisingly, we are finding some cases on this list. To date, they are not above expected background rates. We continue to monitor that.
I think on an earlier slide I mentioned that as part of our submission packet, we will be putting together an integrated summary of safety that will look across D-Pan and Q-Pan programs to assess that risk.
DR. MODLIN: Seth?
DR. HETHERINGTON: Just a little more information on the child who had the elevation in liver enzymes. It sounds it was a preexisting condition, but if the condition worsened post-vaccination, it would be considered a treatment-emergent adverse event. Can you comment on whether the liver function tests in that child increased subsequent to the immunization or were they stable?
DR. VAUGHN: I have limited information about this child. My understanding is that liver enzymes were, ALT around 200 or so on day 0. Since that time, they have gone as high as 300 to 500. They have gone lower. They have gone up and down over the last year. So it would be a judgment call as to whether this is an exacerbation of a preexisting condition.
DR. HETHERINGTON: The fluctuations, then, were probably within some range that would keep it within the same grade, I would guess. It doesnít sound like it increased in grade. There are specific criteria for various grades of liver enzyme elevation.
DR. VAUGHN: Correct. And the child has remained asymptomatic.
DR. MODLIN: Dr. Debold?
DR. DEBOLD: Iím concerned about how we make decisions about vaccinating people who may be potentially at risk, particularly children. Iím assuming that if you had it to do over again, you probably wouldnít have enrolled this child in your trial. How should we go about deciding who should and shouldnít receive a vaccine with ASO3 in it?
DR. VAUGHN: If you are talking about the declaration of a pandemic, the decision of who should receive vaccines, and which vaccines, will be left to government authorities. If we are talking about 3 million to 7 million dying in the next pandemic, thatís one decision; if we are talking about 350 million people dying in the next pandemic, thatís another situation. There is a risk/benefit ratio to be considered there.
Were you going to ask about clinical trials prior to --
DR. DEBOLD: I think you said something about putting this in the seasonal flu vaccine as well. Iím assuming that this is going to eventually unroll here.
DR. VAUGHN: The question is how to decide the use of an adjuvant like ASO3 for seasonal vaccines. I think thatís probably a different question than what we are discussing today. But you look at data in adults first, as is being done, and start in a small number of children, if you move to children, and gradually increase that until there is satisfaction of regulatory authorities and advisory committees that the risk is acceptable from the vaccine to prevent a known quantified risk of seasonal influenza.
DR. MODLIN: Norm?
DR. BAYLOR: A comment was made, and I just wanted to make sure we address it. There was a comment made about any preclinical data that you had. I think Dr. Debold mentioned that. Do you want to comment on any of that?
DR. VAUGHN: Preclinical data in terms of efficacy or safety?
DR. BAYLOR: Primarily safety.
DR. VAUGHN: I wonít be able to cite it completely for you. Small animal studies have been done and have been included in the IND submissions and will be updated for the BLA submission.
I think I may have missed the point of the question.
DR. BAYLOR: I just wanted to make sure -- Dr. Debold mentioned that. I interpreted her question as whether we could get some background on what the company has done preclinically, before going into these studies that you have gone into in humans.
DR. VAUGHN: There has been work done in small animals prior to going to clinical trials.
DR. MODLIN: Does someone want to respond from GSK?
PARTICIPANT: Yes. In the context of adjuvant systems, like we have done for the ASO for adjuvant, we are looking at the mode of action in the sense of the adjuvant does in the immune system. In light of rare events -- because thatís what we are talking about here -- there is no recognized animal model that can help you with that. So basically the best way to progress on that is from these two studies and establish as best as is possible the mechanism of action of the adjuvant on the innate immune response, as well as the adaptive.
DR. MODLIN: Thank you. I basically heard you say that there is no recognized animal models to study human autoimmunity.
DR. MODLIN: We do need to move on. Dr. Vaughn, thank you very much. We have kept you up here longer than we intended, but the information that you have presented, I think, has been extraordinarily valuable.
We will go on with the next manufacturerís presentation, which will be by Dr. Theodore Tsai from Novartis.
Agenda Item: Novartis Presentation
DR. TSAI: Iím Ted Tsai, representing Novartis Vaccines. Thank you for the opportunity to comment on the important public health and ethical question of pandemic influenza vaccine development for children.
We support such development under Subpart D 50.54. Although the avian H5N1 influenza virus currently is the focus of attention, because other avian and non-avian influenza viral subtypes also pose a pandemic threat, we believe the clinical research on vaccines against those subtypes also can be justified under a similar rationale.
Novartis is one of several companies that has conducted H5N1 vaccine trials both in adults and in children. To help inform the committeeís discussion, I will review that clinical experience. But because the safety of adjuvants is part of the larger concern, I will first review our experience with our adjuvant, MF59, and our adjuvanted seasonal influenza vaccine, Fluad.
We suggest that clinical trials of pandemic influenza vaccines in children can be supported under Subpart D 50.54, for a number of reasons. A pandemic, by definition, will affect the United States and large segments of its population, including children. By common consent, everyone should have access to a pandemic vaccine. In times when vaccine distribution must be prioritized, as mentioned previously, the Department of Health and Human Services and the Department of Homeland Security have placed infants and toddlers at the top tier of recipients.
We believe that the justice principle applies equally to pre-pandemic vaccination. Indeed, a strategy of gradually priming the population over time could facilitate outreach to people who might have reduced access during an emergency.
As mentioned previously, children have borne a disproportionate share of the H5N1 cases since 1997, and along with adults under age 40, are also susceptible to H2N2 virus, which further strengthens the rationale for pediatric vaccine development.
Lastly, as mentioned previously, the approval of a smallpox pediatric vaccine trial under Subpart D 50.54 suggests that similar development of a pandemic vaccine for children also could be justified, since arguably the threat of a pandemic in the United States is no less than that for a domestic smallpox outbreak.
You have seen a reference to this document previously. This is the Department of Health and Human Services and Department of Homeland Security guidance on allocating pandemic vaccine. You can see in red that health-care workers and other essential workers are placed in the top priority to receive pandemic vaccine under every pandemic scenario -- severe, moderate, and less severe. But within the general population, only pregnant women and infants and toddlers have received that prioritization.
I would like to describe our MF59 adjuvanted H5N1 vaccine, Aflunov. Aflunov is made by the same manufacturing process as our seasonal vaccine, Fluad, that also is adjuvanted with MF59, which has been licensed in Europe since 1997, with more than 40 million doses distributed to people over 65 years of age. As a monovalent vaccine, Aflunov contains .75 Ķg of the H5N1 hemagglutinin, along with MF59. Formulations with the clade 1 Vietnam and clade 2.2 Turkey/Turkey strain have been produced.
As mentioned earlier, Europe allows for approval of a mockup pandemic vaccine file, and such a vaccine, using the same manufacturing process and formulation as Aflunov, has been approved in Europe under the name of Focetria.
MF59 is an oil-in-water emulsion comprised of squalene, which is a naturally occurring oil and a precursor of human cholesterol synthesis. It also contains two surfactants in an aqueous buffer. The particles in this emulsion are stable for years.
There is an extensive clinical database and clinical experience for MF59 from the distribution of 40 million doses of Fluad since 1997 and actively collected clinical trial data from more than 25,000 recipients of MF59 adjuvanted vaccines. No safety signals have been observed from that experience.
The data from these clinical trials has been compiled, analyzed, and submitted to the FDA as a biologics master file. That master file comprises 94 clinical trials, many of them conducted under U.S. IND, and will compare the safety experience of recipients of Fluad and their non-adjuvanted vaccine counterparts, focusing on local and systemic reactogenicity and adverse events, including autoimmune disease, the new onset of chronic disease, cardiovascular events, hospitalizations for any cause, and deaths.
The more than 25,000 subjects in this database comprise principally adults, but it also includes 748 children, and the preponderance of all the subjects was enrolled at influenza vaccine trials.
Other sources of safety data in adults will be forthcoming from a large-scale observational study being conducted in Lombardy, Italy. This study links the vaccine and medical database and, n recipients of Fluad and non-adjuvanted seasonal vaccines, will compare influenza effectiveness outcomes, as well as vaccine-related, medically attended adverse events. We are seeking to enroll 150,000 subjects. One hundred thousand have been enrolled. The final analysis is expected in 2010.
Lastly, we have a published analysis of the pharmacovigilance database for Fluad over an interval during which 27 million doses were distributed. In that analysis, no safety signals were detected for selected adverse events.
I would like to turn now to describe clinical data on Fluad, both in the indicated population of adults over 65 and experimental data in children.
Fluid is more locally reactogenic than unadjuvanted seasonal vaccines, but most of the adverse reactions are mild or moderate in severity and are transient. In a meta-analysis of 22 studies that traced Fluad and non-adjuvanted vaccine recipients over three consecutive seasons, there was on increase in reactogenicity with repeated vaccination.
On the other hand, MF59 also amplifies the immune response. In the same meta-analysis, Fluad and non-adjuvanted HI antibody responses for the there subtypes were compared and are shown here as a geometric mean ratio, with ratios above 1 indicating a higher Fluad response. I think you can see, for the three subtypes across all three years, the Fluad antibody responses were significantly higher.
MF59 not only raises the antibody response, but also broadens it to heterovariant or drifted strains, influenza viral strains. I have shown here just the results for HI antibody responses to the H3N2 component of a vaccine in which the H3N2 component was a Wisconsin strain. The data are mapped against the CHMP criteria, the EU criteria described earlier for seasonal vaccine, with mean fold increase, seroconversion rate, and seroprotection, which is defined as proportion of subjects achieving HI titers of 40 or higher.
You can see that the Fluad responses, in red, met those criteria, not only for the Wyoming strain contained in the vaccine, but also for three other H3N2 strains, including a California strain that circulated the following year and a Wisconsin strain that circulated two years later. Responses to the unadjuvanted vaccine were lower, shown in green.
So even though this vaccine was formulated to meet the criteria for the Wyoming strain, it would have met the CHMP criteria for H3N2 strains that did not circulate until one or even two years later.
Because the efficacy of TIV unadjuvanted seasonal vaccine in children is lower than that in healthy adults, we undertook a proof-of-concept study of Fluad immunogenicity and safety in 6- to 35-month-old children in Finland. The children in this trial were randomized to receive half of the adult dose of Fluad or a licensed split vaccine comparator. After the first 101 subjects were enrolled, an interim safety analysis was conducted. No safety concerns emerged, and the trial resumed without interruption.
The subjects were invited to return after the initial season for revaccination the next year.
Fluad tended to be more reactogenic than the unadjuvanted split vaccine, but, as shown here for the first dose, only swelling at the injection site was significantly more common Fluad recipients compared to split-vaccine recipients, occurring in 12 percent of the Fluad recipients and 5 percent of the split-vaccine recipients. The pattern of reactogenicity after the second dose, after the revaccination dose, was similar.
The immune response data are shown here as seroprotection against the three subtypes. After the first dose, the antibody responses to Fluad were significantly higher for the H1N1 and H3N2 strains, and after the second dose, to the H1N1, with an even more pronounced difference for subtype B. Nearly 100 percent of the Fluad recipients achieved putatively protective HI antibody titers to the B antigen, compared to 30 percent of the split-vaccine recipients.
Of note, after just one dose, 90 percent of the Fluad recipients achieved putatively seroprotective antibody titers to the H3N2 subtype, meeting the CHMP criteria for young adults.
I show the same immune response data now as GMTs, geometric mean titers. In addition to the post-dose 1 and post-dose 2 titers are shown the results of a follow-up bleeding six months later, ahead of the next season. The Fluad responses were significantly higher at every time point for all three subtypes. The sustained elevated antibody titers, especially for influenza B, may be of significance, because, as we discussed yesterday, influenza B affects children disproportionately and also often occurs in springtime outbreaks.
Similar results were seen in another study that enrolled children up to 5 years of age. In this study Fluad was compared to a U.S.-licensed comparator. The homologous antibody responses to the H3N2 and B antigens contained in the vaccine were significantly higher in the Fluad recipients. The sera also were tested against strains that circulated the following year and that were mismatched. The Fluad responses against all three of these heterologous antigens were significantly higher.
I would like to leave the data on seasonal vaccine and turn now to the MF59 adjuvanted H5N1 vaccine, Aflunov. I will describe the data first in adults. There is a growing database on Aflunov, approaching 10,000 subjects, including 334 children.
Here is the tolerability profile of Aflunov, shown in yellow, for young adults, on the left, and older adults, on the right. The Fluad reaction rates are shown in brown for comparison, because Fluad was used as a comparator vaccine in the trials.
Aflunov tended to be less reactogenic than Fluad, and most of the adverse reactions were mild or moderate in severity, with very few that were severe. You can perhaps see on the bottom of the bars an orange bar, which indicates the severe reactions.
The antibody response data are shown here -- again for young adults on the left and older adults on the right -- to the homologous clade 1 Vietnam antigen contained in the vaccine, in yellow. After two doses, 85 percent of the young adults and 79 percent of the older adults achieved putatively protective neutralizing antibody titers to the homologous clade 1 antigen. Six months later, their antibody titers had declined, but with a third dose, also containing the clade 1 antigen, antibody titers returned to protective levels in 94 to 97 percent of the young and older adults, respectively.
The sera also were tested against a heterologous clade 2.2 antigen, the Turkey/Turkey antigen. After the booster dose, 70 and 60 percent of the young adults and older adults, respectively, achieved putatively protective neutralizing antibody titers to the heterologous response. Please note that these subjects did not receive the Turkey antigen and were vaccinated only with the Vietnam antigen.
In another study, a smaller group of individualsí antibody responses were tested to a broader array of heterologous antigens, including a clade 2.1 Indonesia strain, a different clade 2.2 strain, and a clade 2.3 strain. These were pseudotyped neutralization assays that are done because of safety concerns. It shows that the Vietnam-immunized vaccinees developed heterologous antibody responses to all there heterologous clades.
Expanding on this observation, and also looking at the persistence of immune memory, Novartis is fortunate to have a cohort of subjects who were vaccinated eight years ago with a clade 0 H5N3 adjuvanted vaccine. In 1997, after the Hong Kong outbreak, Chiron made an experimental H5N3 vaccine, because at that time there was no recombinant dilution H5N1C virus available. The H5N3 virus is considered to be antigenically related to an H5N1 clade 0 strain.
The primary neutralizing antibody responses to the adjuvanted H5N3 vaccine were significantly higher than to the unadjuvanted vaccine. Six to eight years later, the subjects were reconvened and immunized with Aflunov, the H5N1 vaccine. Within seven days of receipt of that H5N1 dose, all subjects made high levels of neutralizing antibody, not only to the H5N1 clade 1 strain, but also to viruses in clade 2.1, 2.2, and 2.3 that all had been associated with human H5N1 infections. Control subjects who had not been primed required two doses of vaccine to reach protective antibody levels, as you have seen previously.
This unique set of data suggests that in primed individuals, one booster dose of an adjuvanted vaccine can rapidly induce neutralizing antibodies to putatively protective levels to a broad array of related antigens after an interval of up to eight years after priming.
Proactive pre-pandemic vaccination could be facilitated if that priming could be achieved with just a single dose. In this study, subjects received one dose of the adjuvanted clade 1 vaccine, shown in yellow, and then a year later, a second dose containing a clade 2.2 antigen. Within a week after receipt of that second dose, both groups reached putatively protective neutralizing antibody levels to the respective antigens. Approximately 90 percent of subjects in both groups achieved those protective levels.
The second dose of this vaccine was given concomitantly with an unadjuvanted seasonal TIV without interference to the immune response to those seasonal antigens.
So these observations suggest that a single adjuvanted prime and boost dose with different but related antigens spaced a year apart could provide high antibody responses to both strains.
Finally, I would like to turn to available clinical data on children.
The pediatric H5N1 vaccine trial in children that Iíll describe was approved by the national regulatory authority, as well as the local IRB, without significant delay, in part because the clinical trial site previously had conducted the Fluad pediatric trial, but also because the threat of H5N1 disease in Europe is tangible. H5N1 virus and other highly pathogenic avian influenza viruses have been isolated in wild and domestic birds in Europe, and human H5N1 cases have occurred in nearby Turkey. So participation in pandemic vaccine trials may have a benefit that is considered to be less than hypothetical.
The trial in Finland enrolled children 6 months to 17 years of age simultaneously, and all children who received Aflunov received the full 7.5-Ķg dose. The children were in three age cohorts -- 6 to 35 months old, 3 to 8 years old, and 9 to 17 years old -- and were randomized three-to-one within these groups to receive Aflunov or Fluad as the comparator, and Fluad was given in the age-appropriate dose.
An independent data safety-monitoring board monitored the safety responses after the first 30 subjects were enrolled, after each of the doses. There were no safety concerns, and the trial was conducted without interruption.
I have only the data for the primary responses. The booster response data are pending.
This figure displays the local and systemic reactogenicity after the first dose, but the pattern is similar after the second dose. In general, rates of adverse reactions after Fluad and Aflunov were similar. About 15 to 25 percent of infants and children had adverse events consisting of rhinitis, otitis, and cough. These were the most commonly reported adverse events. In adolescents, headache, pharyngitis, and dysmenorrhea occurred in about 5 percent of the subjects. There was no difference between the Fluad group and the Aflunov group in the incidence of these reactions.
The immunologic data are shown here, against the CHMP criteria, seroprotection rate, seroconversion rate, and geometric mean ratio. For each of the three age groups, you can see 97 percent, 97 percent, and 89 percent reached the CHMP criteria for seroconversion and seroprotection after two doses. For the third CHMP criterion, geometric mean ratio, that criterion was reached after the first dose, with a substantial increase of antibody rise after the second dose, to 129-fold, 117-fold, and 67-fold, respectively, reaching GMTs of 688, 585, and 344.
I have shown a large body of data that I will try to summarize, reflecting both the seasonal vaccine and the H5N1 vaccine experience, both in adults and in children. We have a substantial database of 25,000 subjects, with no safety signals detected thus far. The safety database for children is more limited, with 750 subjects exposed.
In seasonal vaccines, MF59 provides higher and broader responses, and in children, a single dose has