FOOD
AND DRUG ADMINISTRATION
CENTER
FOR BIOLOGICS
EVALUATION
AND RESEARCH
VACCINES
AND RELATED
BIOLOGICAL
PRODUCTS ADVISORY COMMITTEE
September
25, 2008
Crowne
Plaza Hotel
8777
Georgia Avenue
Silver
Spring, Maryland
“This
transcript has not been edited or corrected, but appears as received from the
commercial transcribing service. Accordingly the Food and Drug Administration
makes no representation as to its accuracy”
Proceedings
by:
CASET
Associates, Ltd.
Fairfax,
Virginia 22030
Table of Contents
Call
to Order and Opening Remarks 1
John Modlin, M.D., Acting Chair
Administrative
Matters 1
Christine Walsh, R.N., FDA
Topic 1: Update:
Office of Vaccines Research
and Review Site Visit Report
Response
to the OVRR Site Visit Report 7 Erik Henchal, Ph.D.,
FDA
Committee
Questions/Clarifications 26
Topic 2: Use of MDCK Cells for Manufacture of
Live Attentuated
Influenza Virus Vaccines
Introduction 39
CDR Timothy Nelle, Ph.D., FDA
Manufacturer’s
Presentation 53
George Kemble, Ph.D., MedImmune, Inc.
Tumorigenicity
and Oncogenicity Issues 103
Keith Peden, Ph.D., FDA
Open
Public Hearing 153
Committee
Discussion 154
Adjourn
Meeting
P R O C E E D I N G S (8:30 A.M.)
Agenda
Item: Call to Order and Opening Remarks
DR. MODLIN: Good morning
everyone. I am John Modlin from
Dartmouth Medical School, the Acting Chair for Vaccines and Related Biological
Products Advisory Committee. I would
like to call the meeting to order. We
will start with some announcements from Christine Walsh.
Agenda
Item: Administrative Matters
MS. WALSH: Good morning.
I am Christine Walsh the Designated Federal Officer for today’s meeting
of the Vaccines and Related Biological Products Advisory Committee. I would like to welcome everyone to this meeting. Today’s session will consist of presentations
that are open to the public, as described in the Federal Register Notice of
September 3, 2008. Of the Committee
Members, Dr. Jackson, Stapleton and Gilbert, were unable to attend today’s
meeting.
I request that everyone please
check your cell phones and pagers to make sure that they are all off or in the
silent mode.
I would now like to read into
public record the conflict of interest statement for today’s meeting.
The Food and Drug Administration,
FDA, is convening this September 25, 2008 meeting of the Vaccines and Related
Biological Products Advisory Committee under the authority of the Federal
Advisory Committee Act (FACA) of 1972.
With the exception of the industry representative, all participants of
the Committee are special government employees, SGE’s, or regular federal
employees from other agencies and are subject to the federal conflict of
interest laws and regulations.
The following information on the
status of this advisory committee’s compliance with federal ethics and conflict
of interest laws, including but not limited to 18USC 208 and 712 of the Federal
Food and Drug Cosmetic Act, are being provided to participants at this meeting
and to the public.
FDA has determined that all
members of this Advisory Committee are in compliance with federal ethics and
conflicts of interest laws. Under 18USC
208, Congress has authorized FDA to grant waivers to special government
employees and regular government employees who have financial conflicts when it
is determined that the agency’s need for a particular individual service
outweighs his or her potential financial conflict of interest.
Under 7.12 of the Food Drug and
Cosmetic Act, Congress has authorized FDA to grant waivers to special
government employees and regular government employees with potential financial
conflicts when necessary to afford the Committee their essential expertise.
Related to the discussion of this
meeting, members and consultants of this Committee have been screened for potential
financial conflict of interest of their own, as well as those imputed to them,
including those of their spouses or minor children and for the purpose of 18USC
208, their employers.
These interest may include
investments, consulting, expert witness testimony, contract and grants,
teaching, speaking, writing, patents and royalties and also primary employment.
For topic 1, the Committee will
be briefed on the Office of Vaccines Research and Review response to this
Committee’s Office Site Visit Review Report. There was no conflict of interest
involved with this briefing.
For topic 2, the Committee will
hear presentations and discuss the use of Madin-Darby Canine Kidney, MDCK Cells
for Manufacture of Live Attentuated Influenza Virus Vaccines. This is a particular matter involving
specific parties. Based on the agenda
and all financial interest reported by members and consultants, conflict of
interest waivers have been issued in accordance with 18USC 208 B.3 and 712 of
the Food, Drug and Cosmetic Act.
Related to Dr. John Modlin. Dr. Modlin’s waivers include a consulting
arrangement with two firms that could be affected by the Committee’s
discussions, topic 2. The waivers allow
Dr. Modlin to participate fully and vote on the Committee discussions.
Related to Dr. Pablo
Sanchez. Dr. Sanchez’s waivers include a
consulting arrangement with the firm that could be affected by the Committee’s
discussions, topic 2. The waivers allow Dr. Sanchez to participate fully and vote
on the Committee discussions.
FDA’s reason for issuing the
waiver’s are described in the waiver document’s, which are posted on FDA’s
website at www.fda.gov/OHRMS/dockets/default.htm. Copies of the written waivers may be obtained
by submitting a written request to the Agency’s Freedom of Information Office,
Room 630, of the Parklawn Building, Rockville, Maryland.
Dr. Seth Hetherington is serving
as the industry representative, acting on behalf of all related industry and is
employed by Icagen, Incorporated.
Industry representatives are not special government employees and do not
vote.
This conflict of interest
statement will be available for review at the registration table. We would like to remind members, consultants
and participants, that if the discussions involve any other products or firms
not already on the agenda for which an FDA participate has a personal or an
imputed financial interest, participants need to exclude themselves from such
involvement and their exclusions will be noted for the record.
FDA encourages all other
participants to advise the Committee of any financial relationships that you
may have with the sponsor, its product, and if known, its direct competitors.
That ends the reading of the
conflict of interest statement. Dr.
Modlin, I turn the meeting over to you.
DR. MODLIN: Thanks, Christine. Let us begin by asking the members of the
Committee to introduce themselves and where they are from. We will start with Dr. McInnes.
DR. MCINNES: Pamela McInnes. National Institutes of Health.
DR. KEITEL: Good morning.
Wendy Keitel, Baylor, College of Medicine.
DR.
DESTEFANO: Frank DeStefano, RTI
International.
DR. WHARTON: Melinda Wharton, Centers for Disease Control
and Prevention.
DR. ROMERO: Jose Romero,
University of Arkansas for Medical Sciences.
DR. GELLIN: Bruce Gellin, National Vaccine Program
Office.
DR. COOK: Jim Cook, University of Illinois.
DR. DEBOLD: Vicky DeBold, National Vaccine Information
Center.
DR. HETHERINGTON: Seth Hetherington, Icagen, North Carolina.
DR. HUGHES: Stephen Hughes,
National Cancer Institute.
DR. SANCHEZ: Pablo Sanchez, University of Texas
Southwestern Medical Center.
DR. SELF: Steve Self, Hutchinson Cancer Research
Center, University of Washington.
DR. BAYLOR: Norman, Food and Drug Administration.
DR. HENCHAL: Erik Henchal, Food and Drug Administration.
DR. WILSON: Carolyn Wilson, Food and Drug Administration.
Agenda
Item: Topic 1: Update:
Office of Vaccines Research and Review Site Visit Report
DR. MODLIN: Thank you everyone. The first item on the agenda will be an
update from OVRR on the site visit that actually took place by in March of
2006. Dr. Henchal will be beginning the
proceedings.
Agenda
Item: Response to the OVRR Site Visit
Report
DR. HENCHAL: Good morning. I am always pleased to be
the first speaker at sessions like this because we have got a long day ahead of
us today.
I am Dr. Henchal. I am the OVRR Associate Director for
Management and Scientific Affairs. The
purpose of this presentation is to provide you with our response to an office
site visit made actually in May of 2006.
This morning our review comments
and recommendations from the Office Site Visit review progress towards building
a new strategic planning and research management process, review our future
plans for implementing the office site visit team recommendations, and then I
will take your questions.
We are especially grateful for
the constructive and encouraging comments and recommendations provided by the
2006 office site visit team presented here. Team recommendations were carefully
considered and have been used to enhance the strategic planning and management
of our scientific programs.
On this slide you see the tasks
that were given to the office site visit.
These include commenting on the contributions of OVRR research to the
scientific mission of CBER and the FDA and to biological product regulation, in
general.
Secondly, to comment on ways to
improve visibility of the OVRR research programs and on its integration into
the biologics regulation process.
Thirdly, we asked the team to
recommend opportunities for research expansion, redirection and new
collaborations/leveraging of other programs.
Fourthly, we asked the office
site visit to identify research management strategies for anticipating future
biological products and related research programs.
The office site visit team was
exceptionally generous in their overall assessment of the value of the research
program to the FDA regulatory mission.
However, the team strongly felt that OVRR could benefit from the
development of more comprehensive strategic plan to address its broad missions
more effectively. And we agreed. We found that the office lacked a coherent
research management system to provide a framework for future strategic
planning.
One of the first steps that we
took was to build a new research management system within the office. So, since 2006, OVRR, with CBER’s support and
encouragement, has developed a new strategic planning and research management
system. The planning process actually
begins with the Science Management Committee, which is chaired by the Associate
Director for Research, and includes the division directors and representatives
from both junior and senior scientific staff in the Office.
The Scientific Management
Committee is responsible for taking input of external and internal to the
organization and looking at the broad issues that could affect the research
program.
Through their analysis global
priorities are set for the program.
These recommendations are passed to the left hand side of the screen, to
the laboratories where the research efforts are executed.
Key to this process is an intense
dialogue that begins actually with the principal investigator and his
laboratory chief, to discuss how their programs can best sit within the
priorities of the organization. Also
with a key towards what progress is already been made in their research
projects and what future opportunities exist in the research community.
Laboratory chief then, discussing
their programs with the Division Chief, can focus the program on priorities and
make adjustments as necessary based on the resources available. The Division Chiefs are then responsible for
formulating an integrated program, which is then presented to the Science
Management Committee for their review.
The Science Management Committee then reviews all the programs in the
Office and develops recommendations for the review and approval of the Office
Director.
Contingent on his approval, the
Associate Director for Research then builds an annual research plan, and one of
my functions is to integrate that plan into an overall operational plan for the
Office.
The annual research plan contains
the long term objectives for the program and the short term research aims.
As part of this process, as part
of our strategic process, we do include a look at the future regulatory
workload and evaluate how issues emerging in the reviews of products coming to
us from different sponsors and manufacturers, effects the focus and intensity
of work that must be accomplished in the scientific program. This process is repeated annually, updated
annually, so that we can continually make small adjustments to the specific
aims of the projects.
This list above is not in priority
order, but qualitatively identifies possible programmatic areas for our
efforts. You might note that influenza
vaccine and other emerging infectious diseases are also included on the list.
Using a combination of both
internal and external input into our efforts the Scientific Management
Committee has identified global priorities of the program through 2009. These
include; improving and developing new methods that enhance the safety of
vaccines and related products. Improve
and develop new methods that enhance the effectiveness of vaccines and related
products. Thirdly, facilitate the
development of new biological products for high-priority public health threats,
and emerging infectious diseases and biological threats. Also develop and
evaluate novel scientific technologies and standards to improve biological
product regulatory pathways, availability and quality of vaccines and related
products.
CBER divisions and laboratories
have used these to shape research activities in specific areas, such as the development
of new methods for the detection of adventitious agents to improve vaccine
safety.
In other ways the research
program has used these priorities to development new potency assays for
polysaccharide based vaccines for meningitis and pneumonia. In other ways we
have concentrated our expertise to animal models that will assist the
evaluation of biological threats, such as anthrax, plague, and orthopox
viruses.
In addition, we are assessing new
technologies to replace the older, single immuno diffusion assay that used to
establish the potency of influenza vaccines.
Similarly, we are preparing for
the next influenza pandemic by assisting the development of new methods from
manufacturing delivery of influenza vaccines both seasonal and pandemic, in the
future.
Lastly, we are exploring new
rapid microbial methods using new gene amplification and probe detection
methods to enhance the timeliness and safety of biological product testing lab
release. Including our efforts where unique applications of analytical methods
based upon an MMR and mass spectroscopy to characterize complex vaccines
components. The Science Management
intents to validate these broad priorities annually.
OVRR priorities are aligned with
the FDA and CBER goals. In this way,
OVRR scientific activities are linked to not only Center and Agency priorities,
but national priorities as well, to overall improve public health.
During this process and with a
look towards the future, we have identified several emerging strategic areas
that need our attention. These include;
studies to facilitate the development of safe and effective vaccines for
malaria and Staphylococcus aureus.
Second, to investigate the
mechanism of immunopotentiation by novel and adjuvants being used in new
vaccine formulations.
Thirdly, we recognized that we
need to extend our research focus to human studies including the development of
diagnostic reagents and immunoassays and tools for use in those studies, to
assess the performance of vaccines.
Fourthly, we need to evaluate
many of the new technologies that are now used in vaccine delivery with the
potential for vaccine interference, as well as characterize new attenuated
vaccines, toxoided vaccines, new vaccine vectors, and all other variety of new
vaccine technologies that are regulated by the Office.
These are all complicated issues
and it is critical that the Office be able to make contributions and have the
visibility of these contributions to the overall community. So the Office Site Visit Team recommended that
we increase the visibility of our research efforts through representation on
scientific advisory committees for major collaborative groups and centers for
relevant research.
I am proud to say that our
scientists are actively sought as scientific experts and advisors on numerous
advisory and study groups. You may not
be able to see all of the details on this slide but I hope you will appreciate
the breadth of these activities in which our scientists participate. It is important to note that many of these
activities are undertaken by our scientists in addition to an extraordinary
laboratory workload, as well as carrying on their own scientific projects.
It is through these activities
that OVRR staff, working with the scientific community, are helping shape
national/international efforts to respond to infectious diseases. Working with these advisory committees and
study groups is not the only network available for OVRR scientists. As noted here by the Committee, by the office
site visit team, our relations with academic scientists and clinical
investigators are critical to performing our scientific mission.
This slide represents some of the
many academic collaborations, particularly national and international centers
of research excellence that have been established by scientific staff. These relationships strengthen our program
and provide new opportunities for scientific programs.
In 2008, OVRR scientists
participated in over 61 academic collaborative efforts with other
investigators. At the same time our
program has been able to compete competitively for external funding with these
partners. Since 2007, we have enjoyed a
key relationship with the National Institutes of Allergy and Infectious
Diseases, at the NIH. Specifically, with
the NIAID Biodefense program.
Our scientists have been
enormously successful in leveraging a variety of federal programs and other
funding sources to enhance the intramural research program. With so many internal/external opportunities,
we recognize that along with the office site visit team, that we needed a new
approach to managing the program. We
agreed with the office site visit team when they recommended that we develop a
research plan that can integrate the separate laboratory programs that exist in
our Office.
However, this is not done easily
and it did require a cultural shift from the autonomous investigative driven
programs of the past to more integrated efforts of the future. Beginning in 2007, OVRR began a process of
linking individual investigator driven programs and projects into integrated
programs with common goals. As one of
the first steps was to unify our projects in the influenza area, OVRR influenza
program that has united scientific efforts across the Office to provide a
coherent and productive program to enhance our response to seasonal and
pandemic influenza.
Similarly, our biodefense
research program has been united under an umbrella that has been predominantly
funded through the NIH. This important
program is making contributions to the National Biodefense Strategy to develop
medical countermeasures against a broad range of biological threats. These programs are creating a new paradigm
for our scientists and new ways for them to collaborate. We anticipate that we will continue to grow
additional integrated programs in the coming years that will focus our efforts
on vaccine safety, the study of adjuvants and new vaccine delivery
methods.
However, exploitation of these
new research opportunities and in order to maintain our agility in the research
program, we need to attract and retain outstanding scientists. We agreed with the office site visit team
when they reported that the career pathway in OVRR should be structured to make
it more appealing to new members. Our
response was to participate with CBER to better define the scientific career
paths, as depicted on this slide.
This system has been the product
of many years with input extensively from the scientific community within
CBER. This slide actually shows a system
characterized by flexibility and opportunity.
As you can see, there are many entry ramps to this pathway. It can begin with pre-doctoral and
post-doctoral Fellows first coming to the program through the Contract
Fellowship Program called ORISE. But new
scientists can also enter the program through a competitive Staff Fellow
Programs, and progress in three or four years, and with evaluation by office
site visits and by evaluation of the individual performance in the Promotion,
Careers, Evaluation Committee, they can become staff scientists in the program.
There are also opportunities for
the most outstanding of this group to compete as Senior Staff Fellows or in
this case, Visiting Scientists, these are Visiting Associates, these are
scientists coming to us who may not be U.S. citizens, these Staff Fellows gain
additional time in order to demonstrate their prowess and their expertise and
their contributions to public health – which can be evaluated also through site
visits and the Promotion, Careers, Evaluation Committee, to finally become principal
investigators and leaders in the programs in the Office.
Recently, these opportunities in
this program for scientists to advance, has been enhanced by new programs
sponsored by the Commissioner to increase training of regulatory
scientists. Now we have a program that
can allow scientists to gain their initial experience, continuing to make
contributions as staff scientists, continuing to develop as leaders in their
programs – to lead laboratories and divisions and also become the expert
reviewers that we need in the FDA.
So you might get a sense of the
size of our staff and the future pipeline for research reviewers, please
consider the data on this slide in which we have 30 principal investigators, 10
staff scientists, 5 senior staff fellows on a path to become principal
investigators, 39 staff fellows on a path to become staff scientists, and 82
ORISE or IRTA – these are also contract fellows or pre-doctoral fellows that
can join our programs.
Opportunities for career FDA
scientists are expected to grow with the addition of the FDA Commissioner’s
Fellowship Program, that creates this additional source of future regulatory
scientists in the program.
However the office site visit
team was concerned with how we shared and acquired new scientific information. And we agreed when they said a seminar
program would be useful to maintain awareness of activities not covered by
internal programs, and that we should seek out speakers from company labs, as
well as from academic and government labs.
We have responded, the OVRR currently has an extensive set of seminar
programs that includes monthly invited speakers, weekly work-in-progress
seminars by product area, influenza-specific work-in-progress seminars
now. And we have the NIH jointly
participated, NIH-FDA Biodefense Program Seminars. Throughout the year we have special topic
seminars and training sessions that enhance the exchange of knowledge
concerning the products that we regulate.
But I should mention that and it
is really a test to the high regard for the research achievements of our staff,
that OVRR scientists participate in over 150 invited seminars, symposia, and
conferences annually.
This slide just depicts just a
brief sampling of the external speakers that we have had in our program and
includes everything from new vaccine production methods to new ways to evaluate
the immunity after immunization.
As mentioned previously,
retention of outstanding scientists in our program is essential to maintain our
regulatory mission. The Office Site
Visit Team recommended that we offer both fiscal and non-financial awards and
look for ways to create new support systems for junior and emerging scientific
leaders in our program.
While we have finite resources at
the FDA, we can offer scientists with substantial support that will assist
their studies in the laboratories. I’d like to mention that since 2006, CBER
has matured a critical set of core capabilities that support our projects. These include; oligonucleotide and DNA
sequencing, peptide synthesis, a variety of advanced analytical methods, and
new proteonomic tools.
At the same time, OVRR has been
developing partnerships for critical analytical capabilities and flow
cytometry, mass spectrometry, and have enhanced the sharing of equipment to
contracts.
CBER also provide comprehensive
and more care support on the NIH campus.
There are other ways that CBER and the FDA provide training and support
to our scientists. This includes;
training course of communication skills, scientific and technical writing. How
do you write grants? How do you manage
your time and your priorities and negotiation skills within communities?
FDA provides general assistance
or holistic assistance through its FDA employee assistance program. We can support transportation, healthcare
issues, and the Center and the Agency also sponsor formal mentoring programs
that links junior scientists with more experienced staff members in the agency.
Because our laboratories are
predominantly located on the Bethesda NIH campus, they also have access to a
variety of NIH based support similar to the scientists that work for our sister
agency. This includes the NIH training
courses. Our scientists can also
participate in FAES graduate school in a variety of subject areas. Our scientists have representation on the NIH
campus fellows committee. That also
enhances their training and preparation in a manner similar to the NIH
postdoctoral fellows.
FelCom-sponsored workshops, and
also an annual job fair for these scientists.
In summary, I hope that you have
an appreciation that we value the research regulatory staff that we have in the
office. And these support the science
base review that we conduct and the regulation of vaccines and related
products. I hope you have gained an appreciation for our unique research
program that serves to recruit and train and retain highly qualified
scientists.
And that we now have a system to
not only manage our programs but set priorities that will be focused on the
safety, purity, potency, and effectiveness of vaccines and related products.
I appreciated your attention this
morning and I am willing to take your questions.
DR. MODLIN: Thank you, Dr. Henchal. Speaking for everyone
on the Committee, we very much appreciate the update even though I think very
few of us sitting around the table today, were actually on the Committee in
2006. I noticed that Pam of course, was
a member of the site visit team. What we
do I think, very much understand the importance of these site visits. Many of us have participated in similar ones
and very much understand the importance of the interchange that goes on during
them.
Let me open this up to questions
and comments. Why don’t we start with
you, Pam.
DR. MCINNES: Erik, thank you. That was really very helpful to get the feedback
that was a synthesis of comments given and how it fits and how it could be
helpful. I have some interest in the FDA
Commission’s Fellowship Program that you have described. I am interested in number of slots, duration
of support, how much is the support? Are
there any dedicated to CBER or is it going to be open competition across all the centers?
DR. HENCHAL: Those are excellent questions. Actually this is a brand new program and I
think I would like to call on Carolyn Wilson, who can speak from the Center’s
point of view.
DR. WILSON: I can answer those questions. The Update Fellowship Program this year is
starting with a total of 40 slots that are distributed across the FDA. Of those, the Center is receiving eight
Fellows. It is a two-year program the
way it is currently structured. The
first year the Fellows that are recruited will go through an intensive series
of courses that will be focused on areas of science and regulation that are
important to the FDA. Things like
epidemiology, bio-statistics, - a lot of regulatory training courses, and so
on.
During that time they will also
start their Fellowship. The Fellowships
vary from things that may be full time regulatory positions – so they are in an
actual full time review position where they would be involved in doing
regulatory work. To those that are
laboratory based so that they may spend time also doing research that is
related to regulatory issues, as Erik as pointed out.
In addition to also having the
exposure to hands-on regulatory work as well.
The second year will be more focused on the actual work, depending on
their placement within the Agency of their regulatory/research duties. In addition to just a little bit more
training in the second year, but primarily a lot of training courses in the
first year.
Does that answer all your
questions?
DR. MCINNES: And the support level is a flat amount or is
it the number of years out of post-doc?
DR. WILSON: These are Staff Fellow positions that are
FTEs. So they will be based on a typical
sort of GS structure based on number of years experience and that kind of
thing.
DR. MODLIN: Other questions or comments? Yes, Bruce.
DR. GELLIN: Thanks for that. I would imagine that your fellow regulators
in other countries may have similar programs.
I wonder what conversations you have had or even more than that, you
might have with others to see how their approaching their various regulatory
science issues?
DR. HENCHAL: That is an excellent point. I don’t think we have explored that aspect yet. I know that we interact with those
scientists, but I have not evaluated whether there are any elements in their
programs that we could incorporate into ours.
I appreciate that comment.
DR. GELLIN: Another one related to partnerships and maybe
it is just the way the slides came out on the page, but you have the influenza
program and the counterterrorism program that come next to each other. In one the little box has CBER, CDC, WHO
industry partnerships and the other CBER-NIAID partnerships. I am interested because this is about
research and obviously, NIH is a big research machine.
How do you integrate your
work? I mean, there are these
partnerships, but is there a series strategic plan? Are your priorities theirs or vice a versa?
DR. HENCHAL: You are speaking specifically about the
biodefense program?
DR. GELLIN: I am actually speaking about CBER-NIH and
pick your program.
DR. HENCHAL: Okay.
It is a little easier to talk about the CBER NIH relationship in the
biodefense program because they had established goals that we just focused our
program on.
I think in that program we are
more closely aligned probably, with the NIH goals. In the case of the influenza program, that
has been structured a little differently.
To put emphasis on the needs for our regulatory reviewers to review
influenza and related products. In that
case I think those two programs are a little different. The influenza program is really geared towards
the new standards, new assays, needed for evaluating influenza vaccines, where
the biodefense program really, I think, is probably more closely aligned with
the NIH strategic goals and plan, as you are aware.
DR. BAYLOR: If I could just follow up – pretty much that
same structure exists also for the counterterrorism program. What we do to further those collaborations,
they are unique, even though we are sister agencies and we have some overlap,
there are things that are unique to the FDA that are really supporting our
regulatory. That is what we do. There are research projects that the NIH or
DoD may not focus on, specifically related to that regulatory. We can do that. We have the expertise to do that. We have the background, we know what the
regulatory issues are in support of that.
So we provide sort of a unique role in those partnerships.
Agenda
Item: Committee Questions/Clarifications
DR. MODLIN: Questions?
Yes.
DR. DEBOLD: Hi.
That was really interesting. I
was interested in the strategic planning slide as it relates to the goals that
are set up and that they are aligned with FDA and CBER goals. In some of the materials that were provided
to us prior to the meetings, some of those goals were broken out into more
detail.
Under goal number 2, improve
patient and consumer safety. Objectives
2.2 and 2.3 are fairly explicit in terms of providing information and
transparent communication with patients and consumers. I was just wondering, to what extent
consumers are involved in helping you to more clearly articulate some of these
goals and mechanisms to perhaps better achieve these outcomes?
DR. HENCHAL: We can receive comments from consumers and
the public through this Advisory Committee, obviously. In this case, communications with the
consumer are actually channeled through the Office of Communication and Manufacturer’s
Assistance in the Center. But you are
right, we haven’t asked the consumer and the public to directly review our
research program yet.
DR. HUGHES: So, obviously, the desireability of this is a
place to work depends on the success of the junior people who would come
here. Obviously you have more junior
people than you are –
DR. HENCHAL: As any program would.
DR. HUGHES: -- senior slots. I think I would ask three things. First, particularly for those who are
interested in a career that is in some sense at least, in part research. What sort of research support is available to
post-doc or its equivalent here?
Second, if someone comes here and
most of them should, moves on, where do they go? Where do the people go that don’t stay? Do they end up in academia? Do they end up in industry? What is, in a sense, the record of success of
people who train here?
Finally, connected to that, you
talked about the fact that you have opportunities but they did not really in a
sense, sound like special mentoring opportunities where the young people would
get trained for whatever career they are going to be in. I am sure that they exist, but I would be
grateful if you would elaborate a little bit on how this Institution or this
unit mentors its young people, particularly, the ones that won’t stay.
DR. HENCHAL: Those are all excellent questions. With regard to support, of course the support
for a post-doctoral fellow is actually determined by the PI that they are
working for. And then they are supported
by the Laboratory Chief.
About 60 percent or 70 percent it
would be fair to say, of funding for laboratories comes from external
forces. About 20 to 30 percent,
depending upon the laboratory, comes from the intramural program. We have been, within the last year, we have
seen enormous growth in the research program.
Of course the post-doctoral fellows get to share the equipment and the
resources of the laboratory.
I don’t get a feeling that they
are not receiving the support they need to do the basic studies that they have
to do in their capacity. At least not in
the last several years.
You asked about what happens to
some of these folks. My impression is
actually that many of them, if they don’t stay in our program, do go off to
other training opportunities. Some go to
Academic Centers. Actually, FDA
experience is particularly valued in the pharmaceutical industry, as you would
well expect.
With regard to mentoring, the
mentoring of the junior post-doctorals – fellows, they are being mentored by
their principal investigators. They are
mentored by their laboratory chiefs, but they can also receive mentoring
through the review committees. I think
that is an important concept in that our reviews are done on a team basis. These teams all contain different specialties. We have on our review teams with
statisticians, clinicians, with scientists of different life science
specialties.
Actually, junior scientists who
can participate on review teams actually can be mentored by the more senior
staff on those review committees. The real issue is a rich community available
at FDA to train these junior scientists.
DR. COOK: So a follow up question on that. Thank you for the presentation. The question about your ability to retain and
support your scientists – just thinking about it it is hard to get a sense of
the percentage of time they are able to spend doing research versus
administration. What kind of support
there is per laboratory? If 70 percent
of the support comes from outside sources – and that is the $5.8 million you
site, how well are the laboratories funded? How well are the laboratories
funded? How much time do they have to
spend doing research if they are going to be mentoring junior investigators
that are going to go on to academic careers?
How do you see that?
DR. HENCHAL: This is always a challenge. What we like to say is we like to make sure
that our junior scientist establish their scientific credentials first. So, it really depends upon the
laboratory. It depends upon the
individual, but the job of the post-doctoral scientists when they first come to
our laboratories, is to contribute to the scientific mission of the Center.
We do find that as these
scientists mature. They have got to take
on more and more regulatory workload. It
is our primary mission. It is the first
thing that we do at FDA.
So we generally say that
scientists on the average, spend about 50 percent of their time doing
regulatory work and 50 percent of their time in research. I don’t think this contrasts negatively with
the academic scientists who may spend just as much time doing teaching as they
do research, for example. Time
management always becomes critical.
The difference at FDA is that
when a submission comes to us we have got to focus most of our attention on the
review of that submission and use all of our expertise to complete our review
goals.
DR. MODLIN: Dr. Cook, that was an excellent
question. I tell you that having
participated on several of these site visits that that issue that you raised is
always right up there at the top in terms of the site visit teams. I think we have a pretty good appreciation
for just what Dr. Henchal mentioned, which is attention between the primary
mission and of course, the desire to be able to spend enough time in the laboratory
in order to be successful as a scientist.
I think, for the most part, the site visit teams have been as supportive
as they possibly can in encouraging the FDA to create the resources and the
time available to the extent that they possibly can.
Other questions or comments? Yes.
DR. KEITEL: What is the plan for monitoring the success
of the enhancements to the program and what are your endpoints or outcomes?
DR. HENCHAL: You’ve got me on the spot here because we
haven’t decided upon those metrics yet.
What I would look towards is two things.
First is, what is our achievements with regard to publications, patents,
other contributions to the program, manuscripts, reports? But also I would look
towards the impact is going to be a little harder to measure. Which is what is the impact on the regulatory
process? I think we still have to have a
lot of dialogue on how to measure what that impact might be.
I am going to throw that back to
the Science Management Committee and see if they can come up with a better
answer for what might be a measurable amount for our successes. Thank you for that question.
DR. BAYLOR: Just a couple of comments on that. When you think about endpoints in a program
such ours, the endpoints are really different because again, as Dr. Henchal
described, our mission is regulatory. We
really are fortunate to have research program and what I consider to be a
successful research program, but in the end we are preparing scientists to be
able to review submissions. Have the
expertise to review submissions. The
science is evolving all the time and we have to be in a position to be able to
evaluate new science coming down the road.
That is really what – that is one
of our critical endpoints, is to be able to have a staff of scientists who can
review state-of-the-art technology to move products, facilitate new products,
into the market. I think – we did not
show the slides – but if you look at our success rate as far as the number of
new vaccines licensed over the last five years, it is quite impressive.
DR. WILSON: I’ll add one more point to this which is also
at the Center level, we have an annual process called Research Program
Reporting. Which involves an online
database which each PI provides an annual update of their research where it is
very specific. They have specific
aims. They need to report on their
progress during the past fiscal year.
What they anticipate their work in the next fiscal year and outcomes
will be.
In addition to this, as Dr.
Henchal mentioned, each office takes that information from the PIs and
integrates this into a research program plan, which should very clearly link
their research priorities to the scientific accomplishments of the
scientists. Those are accomplishments,
as Dr. Henchal said, can be in the form of scientific publications, but also
they should also and likely are in many cases, contributions to policy guidance
documents, and the like.
I think we have some concrete
tangibles that we can measure as well to look at the success of the program in
terms of how the research is actually also contributing to the facilitated
development of safe and effective vaccines, as well.
DR. MCINNES: I have two comments. I think people may not – Erik you raised it,
but I think people may not fully appreciate that you are really at the mercy of
applications coming in the door. You
don’t control the number. You don’t
control the frequency and the clock starts ticking. So you have to perhaps drop everything and
focus on that. You really don’t have
negotiation control about timing in those cases.
The other is the ongoing issue of
keeping track of people whose lives have become intertwined with the Agency for
a period of time and they move on. I
think this is a difficult thing to do and we are sometimes constrained by
evaluation issues. But it is something
that I think you should give a lot of thought to because to try and go back and
do it retroactively is incredibly difficult. Google helps a lot, but it doesn’t
find everybody. I think you clearly will
have a role in training many, many people who go on into other aspects and that
training you have provided is a critical thing down the road. I think if some – I don’t know if you can do
it yourselves – you need a Friends of the FDA organization to do it for
you. But I think this is something to
try to do to stay up with people whose lives have been part of FDA and they
move on.
DR. MODLIN: Good points.
Other comments? I guess I could
have the last and that is even though the site visit committees get to focus in
and see a bit more of the details of the science that is being conducted. This committee doesn’t really have that
opportunity. Maybe at some point in
time, if we had a half a day on the agenda, where we could spare it, it would
be enjoyable to just have a sampling of some of the current research that is
now going on because it is excellent research and I think this Committee would
appreciate the opportunity to see it.
DR. HENCHAL: Thank you very much.
DR. MODLIN: We are a few minutes ahead on the
agenda. We are due to take a break but
it is awfully early. I might suggest
that we go ahead and get started with the MedImmune presentation, if that is
possible. That was a question. Is the MedImmune team here yet? No.
DR. COOK: I had a follow up question. It seems to me that there is a – it is not
really a conflict maybe, but maybe it is attention between the ability to do
all of the developmental research oriented things that are listed in the plans
for the future and the time that is required to do other regulatory work. I was wondering if there had ever been
thought given to some kind of partnering with academic institutions or with
laboratories at the NIH to bring people in to increase the numbers of people
you have doing science, maybe to enrich the environment a little bit so that
there is more opportunity for the laboratories to get involved with some of
these developmental things with a limited budget you have to maybe leverage
that with – I mean there are many programs that have graduate students that are
going to have to have jobs in the future.
Maybe you could interest them in some kind of careers like this if there
were real research opportunities and maybe that would expand your abilities to
do science.
DR. HENCHAL: I think actually we do do that. And actually for example, this past summer,
we bring in students all the time. We
probably had 20 students from local universities – actually from across the
country, who work with us. You are
right. There are issues of confidentiality
that have to be dealt with but there are opportunities for scientists, as well
as for students, to come and do training in our laboratories and in our
programs.
DR. WILSON: I was just going to add one more thing. We are also eligible to participate in the
NIH’s Graduate Partnership Program, which partners with a lot of the local
universities to get graduate students to come into the labs. We are just starting to engage but we have
had a few graduate students come into the labs through that program as
well.
DR. COOK: So the NIH will pay those people to come work
in your laboratories with your investigators?
DR. WILSON: We pay a portion, yes. But they provide the administrative framework
for the program.
DR. MODLIN: Let me welcome the MedImmune team that is
here. Please go ahead and take your
seats and we will get started. We are
actually running a little bit early here so I appreciate your willingness to
help us out.
Agenda
Item: Topic 2: Use of MDCK Cells for Manufacture of Live
Attentuated Influenza Virus Vaccines
DR. MODLIN: We are going to move on to Topic 2: Use of
MDCK Cells for Manufacture of Live Attentuated Influenza Virus Vaccines. The sponsor is MedImmune. I think it would probably be a good idea to
ask the members of the Committee to reintroduce themselves. Again, we will start with Dr. McInnes.
DR. MCINNES: Pamela McInnes, National Institutes of
Health.
DR. KEITEL: Wendy Keitel, Baylor College of Medicine.
DR.
DESTEFANO: Frank DeStefano, RGI
International.
DR. WHARTON: Melinda Wharton, Centers for Disease Control
and Prevention.
DR. ROMERO: Jose Romero, University of Arkansas for
Medical Sciences.
DR. GELLIN: Bruce Gellin, National Vaccine Program
Office.
DR. COOK: Jim Cook, University of Illinois.
DR. MODLIN: I am John Modlin from Dartmouth Medical
School and Acting Chair of the Committee.
MS. WALSH: Christine Walsh, Food and Drug
Administration.
DR. DEBOLD: Vicky DeBold, National Vaccine Information
Center.
DR. HETHERINGTON: Seth
Hetherington, industry representative from Icagen in Durham, North Carolina.
DR. HUGHES: Stephen Hughes, National Cancer Institute.
DR. SANCHEZ: Pablo Sanchez, University of Texas,
Southwestern Medical Center.
DR. SELF: Steve Self, Hutchison Cancer Research Center,
University of Washington.
DR. BAYLOR: Norman Baylor, Food and Drug Administration.
DR. WEIR: Jerry Weir, Director of the Division of Viral
Products.
DR. PEDEN: Keith Peden, Division of Viral Products.
DR. MODLIN: Thank you.
We will lead off with Commander Nelle’s presentation.
Agenda
Item: Use of Madin-Darby Canine Kidney
(MDCK) Cells for Manufacture of Live-Attentuated Influenza Virus Vaccine
DR. NELLE: Good morning.
My name is Tim Nelle. I am a
primary reviewer in the Division of Vaccines and Related Product Applications
at CBER. I will be providing a brief
introduction to today’s topic, which is the use of Madin-Darby Canine Kidney
Cells, or MDCK cells, for the manufacture of live attenuated influenza virus
vaccine or LAIV.
The specific vaccine to be
discussed today is analogous to the license flu mist vaccine, which is
currently produced in eggs. Today’s
discussion will focus on moving the production of the similar vaccine from eggs
into cell culture, specifically, MDCK cells.
Why is there such interest at
this point in time, from moving vaccines such as this, influenza vaccines in
general, moving the production from eggs into cell culture?
The impetus for moving production
into cell culture arises from the nation’s recent desire to enhance pandemic
influenza preparedness. In May 2006, the
National Strategy for Pandemic Influenza Implementation Plan was released. It translated the 2005 National Strategy for
Pandemic Influenza into many actions, timelines, and metrics, for the federal
departments and agencies.
One of these priority actions
mentioned, is to advance technology and production capacity for influenza virus
vaccine, with the ultimate goal of having enough surge capacity to provide six
million doses of pandemic vaccine within a six month time period.
While the goal of this is
primarily directed towards pandemic influenza, it is difficult to develop such
systems and infrastructure for the sole purpose of pandemic vaccine
production. Therefore, manufacturers are
developing cell culture based systems or seasonal vaccine production as a first
step.
Why is there such interest in
using cell-culture technology instead of eggs – which have served us well for
nearly 50 years. Here are some of the
advantages that cell-culture provides.
It is important to understand that
the current license influenza vaccines which are produced in eggs, occurs
through a process that takes nearly nine months. Scientists much first select the virus
strains that they anticipate will be predominant in the coming flu season. These strains are then adapted to growing
eggs, and manufacturers subsequently inject each adapted virus strain into
millions of fertilized eggs and allow the virus to grow. These viruses are then
harvested and purified to make our annual influenza vaccine.
Very large batches of eggs are
required by this process. Such large
quantities are not available on-demand – which is a major stumbling block in
terms of being prepared for a pandemic.
In many cases, it takes several months to us to a year in advance, to
order these eggs.
In contrast, cell-culture uses
cells which we freeze down and are available upon demand. So they are always available. This provides more flexibility and allows the
scale of manufacturing to be determined at the start of production so there is
very little advance planning needed.
In a few days to weeks, the
amount of cells required for production can be grown from very small quantities
of frozen cells or cell banks, for this purpose. This offers a distinct
advantage particularly in terms of in a pandemic event, where a large quantity
of vaccine will be needed in a short period of time.
In addition, the cells used for
production in the cell-culture system are extensively characterized during each
phase of the utilization. This allows a
very consistent proliferation system for virus production. This is in contrast to the use of eggs since
the eggs vary within natural limits.
This also leads to some variability in virus growth, which creates
challenges in terms of meeting their goals and making the amount of vaccines
that they would desire.
Extensive characterization of
these cells, as I previously mentioned, also offers the advantage of insuring
that the cell substrate that will be used for virus growth, is free from
adventitious agents. It also allows
closed manufacturing systems such as analogous to fermentation systems to
provide additional level of safety from adventitious agents.
This is in contact to the current
egg-based manufacturing which occurs primarily in an open manufacturing type
system.
Of course since the cell-based
systems are free from the use of eggs, such vaccines could be recommended for
people with egg allergies.
The advantages that I have just
covered are independent of the type cell line to be used. Where there are several cell lines that could
be used for a cell substrate influenza vaccine production, such as Vero and
Percy Six, today’s focus will be on the suitability of using MDCK cells.
One might ask, why choose MDCK
cells over others? When choosing a cell
line it is desirable to have one that supports the most efficient replication
of the virus and also allows large quantities to be produced in the shortest
amount of time. Of course, you would
also want the cell line to be capable of supporting the replication of a wide
variety of influenza strains.
It is also desirable to have a
cell line that can be grown in a chemically defined media as such to avoid
animal derived components. MDCK cells
meet the requirements and also offer the additional benefit of being widely used
in influenza research and influenza surveillance programs.
This sounds pretty rosy – pretty
good for MDCK cells – but what are some of the regulatory challenges that are
associated with the use of MDCK cells?
First there is limited
information about their derivation of this cell line, which originally was
described in 1958 by Drs. Madin and Darby, in this case they derived the cell
line from the kidney of a healthy Cocker Spaniel. More details about this will
be presented in a later talk by Dr. Keith Peden.
However, it has been important to
note that this cell line is a neoplastic cell line or a continuous cell line
and we do not currently understand how these cells became neoplastic.
Another challenge is, unlike Vero
cells, MDCK cells have not been extensively used for vaccine production for
vaccines that are for human use. Since
MDCK cells are neoplastic, they also have the potential to become
tumorigentic. Furthermore, as reported
in the 2005 VRBPAC meeting, some MDCK cell lines display an increasing tumor
forming capacity with increased passage.
The next two slides will focus on
the sources of concern when using neoplastic cell substrates and our regulatory
approach for addressing each concern.
First, there is a concern that
intact cells may find there way into the final product and lead to tumor
formation. While this possibility has been discussed in previous VRBPACs, the
consensus was that this possibility was deemed unlikely due to zenograph
rejection. However, we still take this possibility seriously and ask that
manufacturing systems assure removal of intact cells using validated methods.
The presence of residual cellular
DNA is also a concern due to the potential presence of oncogenes, which could
confer the transformed phenotype or the possibility of infectious genomes also
being present. Our regulatory approach
for addressing such concerns has been to limit the amount of DNA in the final
product and to reduce its biological activity via digestion or oculation to
reduce the size the DNA fragments.
We also recommend in-vivo
oncogenicity testing of the cell substrate DNA for an added level of
assurance.
The next source of concern is
adventitious agents. Here the concern is
– which is unique to neoplastic cells or transformed cells, that there is a
concern that such cells may contain agents that led to their original
transformation to neoplastic cell line.
The approach that we are taking is to minimize the concerns about this
concern is to use expanded adventitious agent testing to detect latent viruses
and to also use in-vivo testing of cell lysates for the general detection of
oncogenic agents.
The regulatory guidance that I
have just mentioned in the last two slides, have been previously discussed with
VRBPAC back in 2005. During this meeting,
MDCK cells were considered for the production of inactivated sub unit influenza
vaccines. This is opposed to today’s
discussion, which is for the use of MDCK cells for the production of live
attenuated vaccine.
Before introducing today’s topic,
I would like to briefly recap the discussion that occurred back in November of
2005 for the use of MDCK cells for inactivated vaccine production.
During this meeting, two sponsors
presented their production systems. They
were Solvay and Kyron, which is now Novartis.
They presented data on their MDCK cell banks and their proposed
manufacturing processes. In both cases the MDCK cell lines that were to be used
were found to be tumorigenic and their tumors varied by sponsor. While the details regarding the tumorigenic
results will be presented by Keith Peden later today, it is suffice to say,
that these cell lines displayed a moderate to high level ability to form tumors
when injected into immunocompromised mice.
Both sponsors both noted that
their cell lines exhibited increased capacity formed tumors with increasing
numbers in passage in cell culture.
While this raised concerns about the possible presence of unrecognized
occult agents, the conclusion at the end of the day was that the Committee was
comfortable with the use of MDCK cell lines for the production of inactivated
subunit influenza vaccines due to the inactivation and purification steps that
were employed during manufacture.
However, they also noted that
additional discussion would be needed in other situations, which is why we are
here today.
The purpose of today’s meeting is
to discuss the sponsor’s MDCK cell line for the production of a intra nasal
live attenuated influenza vaccine. You
will next hear a presentation from MedImmune, followed by a discussion of
tumorigenicity and oncogenicity issues by Dr. Keith Peden. This will be followed by an open public
hearing and committee discussion.
It is important to note that
today’s presentations will be primarily focused on tumorigenicity and
oncogenicity testing data. While the
sponsor has performed extensive adventitious testing, they will not be
discussed in great detail today. These
results can be found in the appendix of the sponsor’s briefing package.
We have elected not to go into
great depth in terms of adventitious agent testing for several reasons. One, the adventitious agent testing
recommended, and this instance is the same as previously recommended for the
cell lines that were discussed back in 2005, which showed a higher level of
tumorigenicity. These assays were discussed in detail back 2005 and were found
to be comprehensive in nature. So far,
the data that has been received and reviewed, we have not noted any unusual
findings so far.
Before yielding the floor, I
would like to also preview the questions that will be posed to the Committee.
The first is; Are the preclinical
data discussed today adequate to support the initiation of phase 1 clinical
studies for live attenuated influenza vaccines manufactured in MDCK cells?
The second question is; What
additional data would you recommend prior to moving into larger clinical
studies with this product?
Thank you.
DR. MODLIN: Thank you, Dr. Nelle. We will have plenty of opportunity to discuss
the issues that Dr. Nelle raised in great detail and you will have much more
detailed presentations coming from both the sponsor and from the FDA. But let me just ask if there are questions
specifically, from Dr. Nelle’s presentation?
DR. COOK: Just a general question to understand the
position. Is the question as posed
restricted to intranasal vaccines?
Because it says, phase I clinical studies of live attenuated vaccines.
Does that mean that if this is considered to be appropriate it can be used for
any kind of vaccine injected or otherwise, or just intranasal?
DR. NELLE: I believe we are going to restrict it to
intranasal vaccines.
DR. MODLIN: I am pretty certain that it is for the
intended use for this product would be for intranasal use only would be my
guess.
Let me just ask a very brief
question. I understand that this is a
matter of semantics. Why are we
characterizing MDCK as a neoplastic cell line?
They presumably came from normal tissue.
I can easily see why it would be characterized as an immortal cell line
but why do we characterize it as a neoplastic cell line?
DR. NELLE: I think I will let Dr. Peden answer that
question.
DR. PEDEN: It is a matter of semantics. What we have taken to call all cells that are
transformed in any way as neoplastic.
You are correct, clinically it did not come from a neoplasm – it came
from a normal kidney. But it is just
over the years we have made the instigation – probably Andrew Lewis, who is a
clinician and studied these issues for many years, we have decided for
consistency to call them all neoplasticly transformed, starting with
immortalization through to tumorigenetic cells.
They are all part of that neoplastic process.
DR MODLIN: Seth.
DR. HETHERINGTON: I just wanted to go back to the question
about whether this is restricted in its application to the LAIV. I would assume since we are talking about a
specific cell line, that these kinds of discussions would be applicable to any
other vaccines in the future that may be using MDCK cells.
Also I would assume that based
what we are going to hear in the future from Dr. Peden about this standards
that are set for safety for a cell line no matter what its origin is, that this
discussed will have implications for future vaccines developed from other cell
lines. That we are not going to be
recreating new standards every time we come up with a new cell line unless
there are specific issues related to that cell line.
I guess the question I am asking
is isn’t there some generalizability from the discussions today to other cell
lines or other vaccines?
DR. MODLIN: Norman, I think you are the right one to
answer that.
DR. BAYLOR: That may be true but we would like you to
focus on today is the particular product – the intranasal LAIV from
MedImmune. We may be able to, depending
on the outcome of this discussion, use this discussion as we move forward with
other similar products. But I would like
to hold off on that because we really don’t know the nuances that will come
with future products. I would like to
focus on the MedImmune product today, the LAIV intranasal.
DR. MODLIN: Dr. Coffin has just joined the
Committee. Do you want to briefly
introduce yourself, Dr. Coffin.
DR. COFFIN: I’m John Coffin, Biology and Microbiology,
Tufts University.
DR. MODLIN: Welcome.
If there are not any other questions –
DR. GELLIN: If he would repeat again the adventitious
agent part of this discussion because what is set up here is sort of a go/no-go
decision about moving forward on a phase I trials. Just keep us informed about other
considerations besides the specific cell lines.
DR. PELLE: So, in terms of the adventitious agent
testing, the data we received so far we have not found any concerns about
that. There is still some outstanding
data but the main points are that the testing that we have recommended for this
cell line, which appears to be non-tumorigenic, is the same testing that we
requested for MDCK cell lines, which showed very high capacity of formed
tumors.
DR. GELLIN: Just to paraphrase, then the only remaining
issue on the table to go to phase I trial –
DR. PELLE: For today’s decision, we really want you to
focus in on the issues surrounding oncogenicity and tumorigenicity in making
your recommendation.
DR. MODLIN: Thank you, Dr. Nelle. Having asked the sponsor to come in and join
us, I also noticed that you have an hours presentation. I think probably the thing to do is would be
to ask the Committee to take a break now.
If you will indulge us, we will come back at 10:10 a.m. on the dot.
(Break)
DR.
MODLIN: I guess we will go on with the
manufacturer's presentation and Dr. George Kemble from MedImmune will be
leading off. Dr. Kemble welcome.
Agenda Item: Manufacturer’s Presentation
DR.
KEMBLE: Thank you and I want to thank
CBER and the Committee as well, for giving us this opportunity to talk about
our program today which we think is a very exciting program to go forward
with.
So
what I want to tell you about today is the live attenuated influenza virus
vaccine manufactured in MDCK cells and the aim of this program overall is to
develop a safe and reliable vaccine technology both to enhance the nation's
supply of annual influenza vaccine as well as be a significant contribution to
increase pandemic preparedness.
We
have outlined three goals today. One is
I want to be able to describe the benefits of switching from cell cultured to
egg based production of the live attenuated vaccine. I want to spend a significant amount of time
describing the safety of production of LAIV in MDCK. I will talk about the characterization of the
cell line, describe the manufacturing technologies that we use that help us
ensure that we have a high quality, safe product and then finally spend a few
moments on defined risk assessments that help us model the risk associated with
some of the residuals in the NV vaccines.
Then finally to enable VRBPAC to recommend to move forward with the
clinical development of a cell cultured produced live attenuated influenza
vaccine.
So
really there are three major components to what I am going to describe here
today. Shown on the left side is our
cell substrate and I will be describing the extensive characterization
infesting to that substrate. Then also
make you aware that we do have bank cells and they are ready for use and have
been used in fact for our first clinical trial material production.
We
combine that with our robust manufacturing process which removes cells and
reduces other sub cellular components in the vaccine. It is a Aseptic manufacturing process. We combine that with a comprehensive testing
program for each bulk lot of vaccine. We
bring those elements together the cell substrate, the manufacturing process,
the vaccine testing, that really helps us ensure that we have got a safe and
high quality MDCK produced LAIV.
I
am going to spend a few minutes on the background of who MedImmune is, a little
reminder of FluMist, and then get into some of the differences between the egg
and cell based production technologies.
MedImmune is the worldwide biological unit of AstraZeneca. It is headquartered here in Gaithersburg,
Maryland. And since 2003, has been
marketing Flumist, which is the live attenuated influenza vaccine produced in
eggs and between licensure and the end of last season, over 11 million doses
have been distributed commercially. Again
that experience combined with the clinical development of that program has
demonstrated the safety as well as the effectiveness and the efficacy of that
product.
LAIV
is an important component of influenza prevention on an annual basis as well as
pandemic preparedness. These are some of
the attributes that go along with this program.
There is established efficacy against seasonal influenza including cross
protection against mismatched strains.
There is a strong immune response seen after even a single dose in
immune-naive population, which is children for the seasonal vaccine and that is
a very important aspect of a pandemic vaccine.
Something
that may not be fully appreciated is there is considerable manufacturing
efficiency with the LAIV product. So if
you think about the production of the influenza vaccine and the dose that we
use 10 to the 7th FFA, or infectious units, is actually a very small amount of
the HA protein. So as an example one egg
typically is sufficient to make a dose of inactivated vaccine. That same egg or in this case millet tissue
cell culture fluid, has anywhere from a 100 to thousands of doses of LAIV. There is a great economy of scale and efficiency
when trying to produce massive quantities of vaccines in response to a
pandemic. Of course it comes along with
an innovative intranasal delivery route.
There
have been over 57 completed clinical studies some placebo controlled, some
active controlled with the inactivated vaccine.
These studies in their context demonstrate that FluMist was efficacious
in both adults and children. It was
efficacious across multiple influenza seasons and in trials that were conducted
throughout the globe.
Turning
now to cell cultured-based production, a little comparison and contrast to the
two techniques which many of you are aware of.
Probably the most important part of thinking about egg-based production
is the vulnerability of that production during a time of crisis. We use what are called for our vaccine, SPF
or specific pathogen-free eggs, yet the flocks while guarded are still out
exposed to the environment. So at the
time that you are trying to produce a pandemic vaccine there will certainly be
avian influenza viruses as there are at any time circulating and those flocks
are vulnerable. So at the time you are
trying to produce massive amounts of pandemic vaccine your production
substrate, the egg is vulnerable because the chicken flocks are
vulnerable. No chickens, no eggs, no
vaccine. Obviously the cell culture
production does not have that same risk associated with it.
Preproduction
characterization is another difference.
So taking an egg and trying to characterize it for use in production is
really not feasible. Taking the cell
banks and characterizing extensively as I will describe today is, and that is
an important part of assuring the quality of your production substrate.
The
other part which most of you are aware of is egg have an inherent microbial
contaminations. Eggs are not
sterile. We have to put procedures in
place during manufacturing to deal with that.
That exposure has led to a number of issues over the last several years,
in fact in the 2004‑2005 season was partly responsible for a shortage of
influenza vaccine in the United States.
The cell banks used for cell culture production are tested to be sterile
and of course handled in an aseptic environment. And finally as I pointed out the egg-based
vaccines come along with a contraindication to individuals who have an egg
allergy. With the cell culture vaccine
that is not an issue.
Now
what about the scale on the production of cell culture-based influenza
vaccine? So certainly you can make
larger quantities of both doses more rapidly than you can in eggs. As an example if you want to produce 150
million bulk doses of the vaccine it will take about three months doing that in
eggs in our case running at a fairly high rate, higher than we could normally
run. In contrast we can make that same
volume of bulk dose in about four weeks or a third of the time using two 2500
liter reactors which is our current expectation for our manufacturing
scale.
Increasing
the scale is also faster and again it is a fairly common sense approach. If you want more eggs you need more chickens
and it takes about a year to build up the flock numbers to get more eggs out of
them. Scale for cell culture production
is really limited on the availability on cell culture reactors that are
appropriately configured to run your process.
As we speak we actually in addition to those two 2500 liter reactors
have two slightly smaller reactors available to us and that would not quite
double output in the same time frame. So
it is an efficiency that you gain from that type of production system.
Now
I want to turn more specifically to FluMist or the live attenuated influenza
vaccine. LAIVs are 6:2 genetic
reassortments. What that means is they
contain a flu as an eight segmented negative strand RNA virus. It contains the wild‑type,
hemaglutinin, and neuraminidase gene segments.
The other six gene segments in our vaccine strains are derived from a
master donor. They are the same six gene
segments in all of our vaccine strains and it is what confers the attenuation
and the characteristic in vitro and in vivo phenotypes that we have on our
vaccines.
Something
important to consider as we move through today is those gene segments for the
egg-based and the gene segments for the cell cultured-based vaccines are
identical. We have not modified the gene
segments in order to produce more vaccine and MDCK cells.
How
we actually make these reassortments these days is shown on the slide in the
middle. So on the left side you have a
wild‑type virus shown in purple, we isolate using recombinant DNA methods
the HA and the NA gene segments as cDNA plasmids. It is that process that then safeguards us
from any potential contaminants that might be in that wild‑type human
isolate by going through a recombinant DNA process isolating the cDNA in a
plasmid and sequencing that plasmid. We are
quite clear that we have only that egg chain from that wild‑type
isolate. We combine that with six other
plasmids which represent those internal genes of the master donor virus, we
transfix cells and what comes out is a defined 6:2 genetic reassortment shown
there on the right which has the HA and the NA of the wild‑type, the six
internal gene segments of the master donor virus. It expresses the classical in vitro and in
vivo phenotypes associated with our vaccines and confers the attenuation
properties.
So
that you are aware this process is what we currently use to make our commercial
vaccine. This is currently used. It is what has been used for this season to
make all three strains in FluMist and it is part of the egg-based process. Again just shown on the bottom is an
important reminder that it is this process that we feel gives us a great
confidence in starting the qualification of our production by isolating our
vaccine from ever having exposure to a potential agent that might have been in that
wild‑type isolate.
We
also did a number of preclinical studies and we wanted to take the same
starting material, grow it in eggs, grow it in cell culture and then compare it
in preclinical models. So one of the
things that we did is we completely sequenced the genomes of these viruses and
compared them and they were indeed identical whether you grew in it eggs or in
cells. We looked for the characteristic
in vitro phenotypes of cold‑adapted and temperature sensitivity and they
both expressed them to the same level.
We looked at a variety of other parameters the ability to infect
different host cells, virus protein expression, virus morphology. They were all comparable between these
two.
Then
importantly we did a series of animal studies where we looked at the
replication and attenuation in ferrets which is a classical animal for
influenza and we that we use a lot with our vaccine. We look for immunogenicity and protection
from wild‑type in these animals and again showed the egg‑based and
cell derived vaccines were identical. We
have also done GLP safety studies and animal models and again demonstrating
that the egg and cell‑based vaccines were comparable. Other preclinical analyses on these vaccines
again demonstrated essentially that the production substrate did not have an
impact on the performance of the active components of these vaccines.
Cell
culture production to sum this section up is a significant advance for public
health. It allows us to increase the
reliability and the supply of influenza vaccine. It accelerates the speed and quantity of
vaccine supply and it retains all of the advantages, all of the performance
advantages that we have associated with LAIV.
I
now want to turn to the heart of this presentation where the empirical data is
which is producing a safe reliable LAIV and cell culture. First I want to take some time to go through
our cell line selection provided with MDCKs and talk about how we derived our
bank. Talk about the cell line testing
results, the manufacturing technologies that we use, and finally end up with
product testing.
We
had a number of criteria established before we went to actually choose the MDCK
cell line and they were laid out as follows:
First we wanted to have a readily characterized cell line that could
help us assure product safety. We wanted
to ensure that the cell line that we chose did not have a history of known
oncogenic agents associated with it. It
had to support replication of different influenza stereotypes and strains. This is critical when you are manufacturing
seasonal or pandemic vaccines. You must
have your substrate ready to be able to respond to variants and subtypes that
may emerge.
We
needed a consistent cell growth and high virus productivity in order to manufacture
a vaccine at large scale and we wanted the cell line that can be grown in serum‑free
media so we could eliminate can exposure to serum and any potential
advantageous agents that might come along with the product.
When
we looked through those criteria we looked at a number of different cell lines,
thirteen different cell lines and really the MDCKs were the only ones that had
all of those requisite characteristics for the production of LAIV. We looked at a variety of other substrates
that are used in a vaccine manufacturing including MRC‑5 and WI‑38s,
we looked at FRhLs, a number of different avian cell lines, and again the
culmination of all of this research demonstrated that the MDCK was the only one
that really fulfilled all of the criteria that we had set out to produce our
vaccine.
To
reiterate on some history in 1958 the cell line was derived from a normal
healthy cocker spaniel. In 1964 it was
deposited with the ATCC and in 2001 we obtained a vial from the ATCC of a
preparation of our premaster and master cell banks.
Again
to remind the committee the MDCK cells contain different subpopulations and
there are a number of different publications and reports out there that help us
describe this. First you can look in the
literature and find different groups of isolated different subclones of that
parental ATCC MDCK cell that have different biochemical properties. So that was one line of evidence. The second one comes from the tumorigenicity
studies some which were described in the briefing documents and earlier
today. You can see a lot of the studies
done with the parental cell types are in the top three rows there. You see very limited tumorigenicity to no
tumorigenicity when you look in nude mice models.
Then
this committee is well aware that in 2005 three years ago two manufacturers
came and demonstrated that they had isolated banks of cells derived from MDCK
cells some which had a moderate level 10 to the 5th cells required to make
tumors in one case or in a much higher level tumorigenicity. In that case you will see those 10 cells
could produce tumors. The interesting
note on that last one is those were suspension cells and that is different than
what most of those other studies are showing different than what our cell bank that
I will describe that it is comprised of.
This
is just a diagram to remind everybody what we are doing. The ATCC parental cell population we believe
is a heterogeneous population. The
literature and the tumorigenicity studies over the years have affirmed that
hypothesis. What we are going to
demonstrate now is we are going to subclone cells out of that parental
bank. We can get we think a variety of
different phenotypes as shown in the literature over the years. On the right you should see a suspension
clone which maybe has a moderate to high level of tumorigenicity moderate virus
yield. On the left is one that we are
more interested in an inherent cell with low tumorogenic potential and a high
virus yield.
Something
that I will remind you of and just for purposes of this cartoon and I will
describe in the manufacturing section will remind you that most modern
manufacturing processes remove intact cells.
So in many ways the tumorigenicity characteristic is an important
characteristic to the cell line in general but we do not put intact cells in
our vaccines.
We
wanted to isolate a MDCK cell population with low tumorigenicity so we focused
on three key areas. First we cloned the
isolated through limit dilution cloning our cell subclones. We required them to be contact
inhibited. We wanted inherent cells for
our system and we wanted them to be able to grow in a robust serum‑free
growth media. So as I said we obtained a
low passage vial from the ATCC. We
cloned them three times by limiting dilution and wanted to establish that
uniform population.
Secondly
while we were doing that we took the time to evaluate and choose a subclone
which supported high levels of vaccine strain replication and I will show you
that data in a moment.
Finally
it gave us a point in time where we could very definitively track every product
that those cells and that bank were exposed to.
We then transferred those cells to a serum‑free media which again
eliminated exposure to any potential adventitious agents associated from animal
derived products and finally produced the cell banks in compliance with current
good manufacturing practices.
We
chose adherence clones and what we did is we took subclones and we placed them
on multiple 96 fold dishes(?) and then we infected some of those dishes with
our LAIV strains. Then we ask the very
simple question how much virus is put out after a certain time of replication
and then we ask what is the titer?
Shown on the histogram you can see a lot of the cells on the left
produced 7.6 logs or less. What we were
interested in is those subclones that are on the right of this diagram. They were produced using more virus in those
wells of 96 fold dishes than the other subclones. We wanted to understand that that was
actually a stable characteristic of these subclones so we then passaged them
over 25 times and asked the question again, are you still producing higher
levels of virus than the parental cell and the answer is yes they are. We chose one of those, we put it in a serum‑free
media, and then we produced the cell bank under a CGMP in serum‑free
media and it is the testing of that cell bank that I will now describe.
As
you are all aware there are three main questions that come along with using a
continuous cell line to produce a vaccine in some cases in cell line for
production of a vaccine. The first is
adventitious agents, the second is obviously is tumorigenicity. Just to define our terms today tumorigenicity
is really the property of that intact cell when put into an animal model to
form a progressive tumor. The third
element is oncogenicity. We wanted to
evaluate whether there is any cellular components so not the intact cell but
cellular components when injected into an animal model that can elicit and
therefore have evidence of oncogenicity inherent to that sub cellular
component.
How
do we address these issues? Well the
adventitious agents who did a number of in vitro and in vivo testing for both
specific and general agents for tumorigenicity revaluated that in a nude mouse
model in fact we used two different studies that I will describe. For oncogenicity the sub cellular components
we focused on were the MDCK DNA as well as the MDCK cell lysate and we used
multiple rodent species for that analysis.
I
am going to spend just a few moments on the adventitious agent testing. Just to summarize it we did a number of
general tests which detect a broad range of different microorganisms. We did for sterility mycoplasma,
microbacteria, in vivo where we put the material into a number of different
animal species, in vitro safety where we put it on a number of different cell
lines and we asked the question is there any evidence of an adventitious
agent. The summary of these studies is
there was nothing detectable in our cell line.
We
then turned to a number of specific tests and these tests looked for agents
that we think might not be as readily detected in a general test and many cases
also provide some redundancy with those tests.
We did over 30 different PCR tests as well as other types of testing
focusing really on human, simian, canine, rodent, equine, and porcine agents
and again the answer was there was nothing detectable.
Then
we did a series of studies and I am going to spend a few more slides explaining
these which are called induction studies.
We take our MDCK case cells and we stress them by putting different
chemical agents on the cells. We then
take the results of that and we ask is there any evidence of a latent RNA or
DNA virus in our MDCK cells?
Here
is how we perform the studies. We took
our MDCK cells and in one case we induced them with a four mole ester and
sodium butyrate which are very potent chemical inducers of latent DNA viruses. On the right side we induced with AzaC in
iodo-uridine again trying to get any retro virus that might be potentially in
the cell line to be induced.
We
then took materials from those inductions and in the case of the DNA viruses we
used degenerate primers so that we could cover a wide variety of different
members of these different families including herpes viruses, polyomaviruses,
adenoviruses and papilloma viruses. The
answer was we were unable to detect by these PCR based assays or electron microscopy
studies any latent DNA virus.
On
the right side of the diagram again we looked at the materials by using
electron microscopy as well as an enhanced reversed transcript assay. We also took those materials and we sub
cultured them for 42 days on a number of different detect or cell lines. We took the results of those detect or cell
lines and we also looked at them by electron microscopy as well as these
enhanced reversed transcript assays. The
culmination of all of this testing demonstrated that we had no evidence of an
adventitious agent that was inducible with these very potent chemical agents in
our MDCK cells.
So
this sums up the AVA testing which is we had multiple testing strategies and we
had no evidence of an adventitious agent in our MDCK bank.
Now
to turn to the tumorigenicity and the oncogenicity so first of all the first
thing to note is we use what is called EOP or end of production cells. So we
take our cell banks and we passage then over and over again to a point past
where we would ever use them in the manufacturing setting. The idea behind this is we are going to try
to put those cells through a system of population doublings and growth to
really give them some additional experience past what we would ever use them
at.
We
then took those EOP cells and we injected them into animals or sub cellular
components. We observed those animals
for six months and we conducted all of those studies in compliance with good
laboratory practices.
For
the tumorigenic potential we did two different studies one in adult athymic
nude mice and one in newborn nude mice.
For the oncogenicity studies we did newborn rodents, nude mice,
hamsters, and rats and I will go through the data here next.
This
is an overview of our tumorigenicity study.
We took groups of 44 animals and again these are adult or newborn nude
mice. We inject them anywhere from one
to ten million MDCK per animal. We have
a group that gets TBS and a group that gets HeLa cells. We observed the animals for six months and at
the end of that study we examined them for the presence of progressive tumors
at the site of inoculation as well as systemically. Remember with all of these studies the
criteria for identifying a tumor is that it actually establishes and progress
and that is what the HeLa cells do in this type of model.
On
the next slide is the summary of the results.
Our top is the adult mouse study and on the bottom is often what is
considered a more sensitive study for tumorigenicity which is the newborn nude
mice. Let me talk you through the
different groups to start with. If you
look at the negative control groups who simply received TBS there were 76
animals in the two studies. There were
no tumors at the site of injection.
There were two tumors noted in the negative control group of the adult
mouse study. This is consistent with
doing six month long studies in a nude mouse.
You occasionally see spontaneous nearing(?) tumors. In this case one was a lymphoma one was a
bronchoalveolar adenoma.
When
you look at the positive control groups you can see in the adults 37 to 41
animals had progressive HeLa cell tumors at the site of injection in fact most
of these animals did not make it for six months because the tumors became so
large. Same with the newborn mice 44 of
44 had tumors at the site of injection again most of the animals did not make
the six month duration because the tumors had become so large.
For
the MDCK cells you can see there was a group of 176 animals that got anywhere
from ten to ten million cells at 44 animals per group. For the nude newborn mice 171 animals again
getting anywhere from ten to ten million cells per animal and there were no
tumors noted at the site of injection.
There are also no tumors in the newborn mice distal from the site of
injection. There was one tumor noted in
the adult mouse in the 10 to the 5th cell group distal from the site of
injection and this was a histiocytic sarcoma.
I
am going to show you the data in just a moment but that histiocytic sarcoma is
a nearing tumor. There were no MDCK
cells in that tumor and also no evidence of MDCK DNA in that tumor. Here are some of the data we used to identify
that tumor. So some of the top two
panels left and right is that histiocytic sarcoma from the adult group at 10 to
the 5th cells. When we stained it with
an antibody protein canine ezrin that we know is on MDCK cells and you can see
that in the bottom left that is the control of the in vitro grown cells they
stained quite well with this antibody; that tumor section did not stain at all.
When you then go in with the control
antibody which is a murine antibody you can now see on the top right hand side
of this slide that they stained very actively again demonstrating that again we
had no evidence of MDCK cells in this section.
We
then went on and wanted to confirm this even further using molecular methods so
we used a canine SINE which stands for short interspersed nuclear element PCR
based assay. This is a very important
assay as we go through our tumorigenicity and oncogenicity studies. What it does is it is a PCR probe to a
repetitive element and this element is highly distributed within the canine
gene of infected. It is present on every
5 to 8.3 kilobases which means that it has got linkage to many, many different
parts of the canine chromosome. Overall the whole coating capacity it makes
anywhere between 1.8 to 3 percent of the genome.
We
took that histiocytic sarcoma from that one animal, we applied the canine SINE
PCR and the answer is we did not detect any signal. We then had an inverse set of primers that
would detect rodent sign elements. In
that case this tissue amplified very readily.
So again backing our conclusion from the earlier slide this was a
spontaneous nearing tumor it was not derived from MDCK cells.
Now
I am going to turn to the oncogenicity studies.
Here is where we took some cellular components and in one case lysate
from MDCK cells and in another case DNA purified from MDCK cells and we gave
them either to newborn nude mice, newborn hamsters, or newborn rats. We watched the animals for six months and
again examined for the presence of a tumor both at the site of the injection as
well as distally. What is important to
note is that the 100 micrograms is intact so we have not gone to any lengths to
shear this DNA in an effort to detect any potential activated oncogene or other
element that might be present in that chromosomal DNA. That 100 micrograms also represents anywhere
from 100,000 to 1,000,000 fold excess of DNA from what is in one dose of what is
in our vaccine.
Shown
on this slide another revoke(?) study and let me take you through the various
parameters. Shown in the top two rows
are either 25 animals per group who were not injected or in that middle row 45
animals per group that were injected with TBS.
You can see that one mouse that was not injected had a tumor. This was another bronchiolo-alveolar adenoma
again that was shown to be a spontaneous nearing as expected. In the TBS group we had two rats with distal
site tumors and again shown to be rodent.
On
the bottom line then are the results from the test articles. We had one rat that had been given the
equivalent of ten million cells worth of lysate that did have a tumor which was
a hind leg carcinoma. We had one mouse
and one hamster that had distal site tumors that had been given 100 micrograms
of MDCK DNA. Again we went through the
very same process. We wanted to
investigate these tumors at a very basic level so we took out our SINE probes,
we applied the canine SINE DNA probes to these the PCR probe in each case there
was no detectable amplification. We used
the rodents SINE PCR probes in each case and they became quite positive so
again these are murine tumors and we believe to be of spontaneous origin. So all tumors were rodent origin no canine
DNA was detected. These are similar to
tumors that are absorbed in other studies of these animal systems that have
been observed before and finally there is a balance frequency and a low frequency
between both the negative control groups as well as the test article
groups. So we have no evidence of
oncogenicity in either of our cellular DNA or in our cellular lysates.
To
sum up this section no evidence of adventitious agents by comprehensive testing
regimen. No evidence of local or
systemic tumorigenicity when we used up to ten million cells per animal and no
evidence of local or systemic tumors caused by an oncogene in either our DNA or
our cellular lysate.
Now
we have addressed the three potential risks of using a continuous cell line
through testing. Now we want to turn to
addressing those risks through our manufacturing process. Again the same three risk adventitious
agents, tumorigenicity, and oncogenicity.
How are we going to address them as we manufacturer our vaccine for
adventitious agents? As I said we start
with a plasmid rescued seed which gives us a high degree of confidence of
partitioning off our vaccine from any potential ADA that might have been with
that wild‑type virus and we have closed aseptic manufacturing
processes.
For
tumorigenicity we remove all of the cells through multiple filtration steps and
I will be showing that data in a few moments.
For oncogenicity we reduced the quantity and size of residual MDCK DNA
in a product as well as the quantity of MDCK proteins in the product.
This
is an overview and it is not meant to go through step by step but let me just
outline some of the main features shown on the top left. We start with our cell bank we grow them on
microcarriers and reactors. We infect
them and it is that system using that high quality cell substrate, that high
quality seed, and enclosed bioreactor that helps partition away our vaccine
with adventitious agents. We then go
through a filtration called filtration step number one which is designed to
remove intact cells from the vaccine. We
then go through another series of purification steps where we reduce the DNA
quantity, we reduce the DNA size, and also reduce the protein load. Then we finalize one last filtration step
again another level of assurance that we are removing intact cells and finish
with a sterilizing step as well.
This
is just a cartoon of relative size. What
it says is filtration removes intact host cells. So shown in that top star is a 15 micron MDCK
cell. The bottom two rows are either
0.45 or 0.22 micron pore size in the filters we use and you can see that the
cell is significantly larger than that of the porous. We have experimentally shown that we can
remove intact cells at both of those filtration steps and in fact when we go
through the numbers we clear an order of 10 to the 21 cells or have the
capability of removing that many cells from our process. That is obviously a lot more cells that we
have in any one particular bioreactor.
In fact we would have to run our reactors one hundred billion times
before we would have that many cells we would have to clear. So I think our filtration capacity is
significantly high with its safety margin.
Just
for a moment I want to describe how we get to those numbers. Ten to the 21 seems like a very big number
when you look at filtration step one as an example we state that it can clear
about 10 to the 12th cells. How do we
get that number? We start with a
suspension of bacteria that we can enumerate.
We use bacteria because they are smaller than the MDCK cells providing a
more challenging filtration step. We can
also enumerate them much more sensitively.
You start 10 to the 12th cells and you filter it and you end up with
nothing on the other end so that gives us a clearance factor of 10 to the
12th. Could it be 10 to the 13th? Possibly but we did not do that
experimentally one because we could not put that many bacteria in and secondly
it is becomes more difficult to run the studies at that level. So that is how we come up with these
numbers.
For
the first filtration step 10 to the 12th for the second one around 10 to the
9th combining them for a 10 to the 21 power again it is those multiple
filtration steps combine in that same process that we think gives us that level
of safety.
Now
I just want to talk about using multiple steps to reduce the quantity of the
MDCK DNA and the protein. So our
purification scheme removes over 90 percent of the MDCK DNA from the starting
material. One dose of our vaccine
currently contains less than a nanogram of MDCK DNA. In fact our first clinical trial material
contains about a tenth of a nanogram of DNA.
WHO just to remind you has a guidance of ten nanograms as a limit for
injectable products from continuous cell lines.
Shown on the right is a gel of the residual DNA in our product
and what you can see are smears there.
These are three different monovalent bulks. The median size of that DNA is approximately
450 base pairs and we really cannot detect much above that but when we try and
quantitate these gels what we can claim is that 90 percent of the DNA is less
than 1000 base pairs. You can also see
markers on this gel up at 2 kb and higher and we do not have any residual DNA
of that size.
Just
showing for scale purposes again is the size of the average coating region of
the million oncogene about 1900 base pairs compared to the size of our DNA, the
median size of the DNA in our product.
Then again for the proteins our process removes over 90 percent of the
MDCK protein in that starting material.
Our current vaccine has less than a half of a microgram of protein in
the final dose.
One
final element we wanted to understand since we had sheared this DNA or degraded
it down to 450 base pairs if we administer it either into an animal model
either by an intranasal injection or an intramuscular injection how long does
that DNA stay around that animal in its tissues? So what we did is take groups of rats and we gave
them either 100 micrograms of sheared MDCK DNA intramuscularly or 100
micrograms of sheared DNA intranasally, took out our assigned TCR assay and
asked over time how much canine DNA do we detect in that animal.
So
in this graph are the results. What you
can see is in that dark blue pyramid there is the amount of DNA residual in
that animal following an intramuscular injection. You can see it degrades actually quite
rapidly in just a 24 hour period of time.
On the bottom left of this graph is actually a lighter blue bar showing
the intranasal amount. It is difficult
to see so what we have done is put it on a log scale so you can start to see
the signals. The first thing that you
can appreciate is six hours after injection or intranasal administration of
these animals we had about a five to six order of magnitude difference between
the intramuscular and the intranasally delivered DNA with the intramuscular
being much higher. It took about two
weeks for that intramuscular DNA to clear out of these animals. It took approximately one week for the
intranasal DNA to clear out of these animals and again the levels of DNA by
intranasal administration were actually quite low and right around the
sensitivity of this PCR assay. Again the
respiratory tract is doing what we would expect it is dealing with foreign
materials and clearing them out of the host.
How
do we test the bulk vaccines? Before we
filter it out, before we subject it to that first filtration step we actually
use that material to look for any adventitious agents. We look for mycoplasma by traditional methods
and we also take that vaccine material and we neutralize the influenza virus
and then we put that material back into in vitro systems or in vivo systems and
we ask is there any evidence of an adventitious agent. We also look for the potency of that virus as
well as the bio burden and these are studies that we do with our egg‑based
vaccine as well.
Once
it is filtered then we look to ensure that it is indeed sterile, we look for
the characteristic genotypes and phenotypes associated with our vaccine strain,
and of course we measure residual host cell DNA and protein in the vaccine as
well.
To
sum up this portion again with our manufacturing process with control all of
our materials at the upfront portion there is minimal exposure to animal
derived components by using serum‑free media. We use highly characterized MDCK cell banks
and we use highly characterized vaccine seeds.
The production systems and environment are isolated from the
environment, their aseptic manufacturing processes, and we use multiple
purification steps to help us ensure product safety. We remove all of the intact cells, we reduce
the quantity and the size of the DNA in the product, we reduce the most cell
protein in the product, and we finish up with a sterile filtration.
Now
I want to turn from the empirical to our modeling approach so we can try to
enumerate a calculation for risk factor assessments due to some of the
potential residual materials in the vaccine manufacturing process. This is a defined risk assessment. It was employed based on CBER guidelines as
well as in line with what other manufacturer's have done to address the
concerns associated with either intact cells, oncogenicity of the DNA, or
infectivity of the nucleic acids. This
really helps us to reinforce product safety by giving us a quantitative element
to consider going forward.
So
for tumorigenicity or the risk of an intact cell I will remind you the
observation is we can have the process that is capable of clearing up to 10 to
the 21 cells from the process. So how do
we calculate a safety margin? At first
they may have to say if flu did not like a cell and if our process did not
clear any cells how many cells would be in that vaccine? The answer would be around 10 to the
5.6. Again we take that, we look at our
clearance factor which is 10 to the 21 and we look for the difference and there
is 15 orders of magnitude difference between those two numbers. That is the margin of safety that we have to
work with. That translate is what is the
risk of one dose of vaccine having an intact cell it is around 10 to the minus
16th it is a very difficult number for us to think about so we invert it and
say how many doses would it take before you would find an intact cell? The answer is using this modeling assumption
it would be one out of 6.3 quadrillion doses or if we gave every individual on
the planet a vaccine every 50 minutes for the next 100 years then we would have
the risk of having one cell in that vaccine.
Now I will remind you that cell is in different strains of animal models
to not be tumorigenic when we administer 10 million of them.
Oncogenicity,
I am going to spend a few more minutes describing how we came up to these
numbers. I will remind you again our
MDCK DNA was not shown to have any active oncogenes in those animal experiments
where we delivered very, very high doses of DNA to the animals. We have to assume for modeling purposes that
it does because otherwise we cannot calculate anything. So we assume a genomic DNA with an active
oncogene in it. We then look back to the
laboratory studies that were done and what was put in the briefing documents
was it would take about a nanogram of oncogenic DNA, plasma DNA to illicit a
tumor in an animal model system. That is
what is we are going to use to frame this calculation.
The
first thing we have to ask is how much genomic DNA will it take to deliver the
same number of oncogenes as that nanogram of plasma. That is the first part of the calculation. Once we get that number we get how much MDCK
DNA that would take and then we say how much DNA is in our vaccine and you can
translate how many doses does that translate into. After we get that first cut of the number
which I am going to describe in a moment then I am then going to walk you
through what we believe is the safety factor the digestion of that DNA can
potentially add to the factor.
Again
using one nanogram as the exemplar that it takes one nanogram of the highly
active oncogene in an animal study to cause a tumor then it would take about a
million doses to deliver that same oncogenic content into the population
assuming we had an active oncogene in our DNA.
What
I have below is what we have in the briefing document which is when we looked
at the literature that was out there it took at that point 25 micrograms of two
separate plasmids. Using that scenario
now it would take over five billion doses to again deliver that same dosage of
oncogene into the population.
That
does not consider digestion of the DNA and is there a quantitative element that
we can add to this that helps us understand the safety factor and there is and
here is how we determine it. Again on
that top line I am going to show the hypothetical genome that has an intact
1900 base pair active oncogene in it.
For consideration if you were to take that two billion base pair genome
and cut it once the likelihood that you are going to cut inside your oncogene
is infinitesimally small. So a single
digestion in the context of an entire genome does not add any safety. Shown on the bottom is what we think we have
actually done which is we have digested it down to 450 and we think if there
were any hypothetical oncogenes they are degraded. But again that is not something that we can
calculate into the risk assessment.
We
had to look at what we would consider a worst case scenario. What if all of the DNA in our vaccine was the
same size of an oncogene around 1900 base pairs? That gives us a tool to try to start to model
and add an actual element back into the calculation. Again the model says we have got a nanogram
of DNA in our vaccine which I will run you is tenfold more than we actually
have and that it is digested to 1900 base pairs in length which again is much
larger than we currently have. Using
those numbers what you end up and you use calculation that people use who work
with cloning genes out of chromosomes they have models now to help us
understand how often will there be an intact gene. How often will there be a degraded gene? Using those criteria that my oncogene is 1925
base pairs and the DNA digested fragments of about that same size about one in
every 2000 of those oncogenes will be intact.
It gives us an additional factor of 2000 to add back to the safety
factor. Now the risk of an oncogene per
dose goes down to 10 to the minus 10th and the safety factor goes up to 2.4
times 10 to the 9th again that is modeled off of only one nanogram of plasmid
required to simulate a tumor in an animal model. Again it would take over two billion doses to
deliver that oncogene content into the population that is used in one animal to
deliver a nanogram of active plasmid which translates into about four hundred
thousand liters of vaccine.
Infectivity,
the exact same steps were taken. Some of
the assumptions were slightly different so again I will remind you we did not
have evidence of an active provirus to either our latent induction studies, our
DNA studies in animals but it does not allow us to calculate anything so we
have assumed a small canine retrovirus that is active in our DNA.
We
then looked back to the in vitro studies now and these studies demonstrated
that when you take 150 nanograms of an HIV‑1 plasmid you put those in the
cells you get an infectious HIV‑1 back out. When you degrade that DNA down to the size of
about 650 base pairs you will eliminate the infectivity of that DNA so that
gives us a way again to think about calculating the safety factor due to our
processing.
The
first thing we ask is how much genomic DNA will it take to give you the same
dosage of provirus that was in that 150 nanograms of plasmid. Then we look at the amount of MDCK DNA in our
vaccine which we are going to assume is a nanogram and then we will ask how
many doses of vaccine will it take to administer that dosage of provirus into
the population. We also again are going
to model in the impact of DNA digestion into this factor. So when you put all of that together what you
get is a risk of provirus in one dose in about 10 to the minus 12th or a safety
factor of about 7 times 10 to the 11th and I will remind you that includes at
this point some modeling where we have a nanogram of DNA in our vaccine which
was digested to only 7000 base pairs the size of the provirus we assumed would
be in there. Again about tenfold larger
than what is in our current vaccine product.
Using that scenario and the extrapolation for the in vitro data we would
say there would be no infectivity detected in only 700 billion doses of
vaccine.
This
is a summary of what we have just gone through.
We wanted to address the three risks that come along with using a
continuous cell line for vaccine production the risk of adventitious agents,
tumorigenicity, and oncogenicity. We
addressed them through testing where we showed by both general and specific
tests on the MDCK cells. We had no
agents. We test the vaccine bulks again
demonstrating we have no agents. We
manufactured using plasmid rescued seeds, closed systems, and ensured that
there were multiple filtration steps which can potentially move adventitious
agents.
Then
for tumorigenicity we again tested ourselves in both nude mice adult and
newborn. We have multiple very powerful
filtration steps in our process that remove intact cells. We showed you there was a risk factor or in
this case a safety factor on the order of 10 to the 15th that ensures we will
clear cells from our product.
For
oncogenicity again we tested the genomic DNA and we tested the lysate, we used
multiple rodents, newborn rodents for these tests, demonstrated there was no
evidence of oncogenicity. For
manufacturing we reduced the DNA size and the quantity, reduced the protein
quantity, and again our risk assessments would show we have on the order of 2
billion for the safety factor of DNA and over 700 billion for the infectivity
of a provirus in our material.
Coming
back and closing where we started on the left we have the cell substrate which
has no detectable adventitious agents, a very low profile for tumorigenicity
and no detectable oncogenicity. We
combine that with our manufacturing process which we demonstrated produces an
acellular vaccine, reduces the DNA quantity in size, reduces the protein
quantity. We combine that with our
routine vaccine testing. We put those
elements together to assure ourselves that we have a safe and reliable vaccine
produced in MDCK cells.
Cell
culture produced LAIV is a safe vaccine and it fills a need for influenza
vaccine. Safety is our primary
focus. We have used a scientifically
sound approach to produce a significant advance in influenza vaccine
production, the cell culture production system, increases the supply and the
reliability of the vaccine. That is
important both for the seasonal vaccine as well as for pandemic
preparedness. With that I will thank you
very much for your attention.
DR.
MODLIN: Dr. Kemble, thank you for a
very, very clear presentation. I think
it would be very appropriate for us to ask questions at this point in time and
again we will start yes ‑
DR.
KEITEL: I have two small technical
questions. The first is can you tell me
what the cell line is that is used for the plasmid rescue? The second is can you describe the effect of
benzonase digestion on the RNA virus?
DR.
KEMBLE: The cell line used for plasmid
rescue is VERO cell. It is a bank that
we have tested extensively. As we move
into and that was really for this first clinical trial materials and move into
further production of our cell‑based vaccine MDCK cells we would move it
into an MDCK cell bank.
The
second question benzonase on RNA viruses -- on an intact virus again benzonase
probably has limited access to nucleic acid. On pre nucleic acid it is an
endonuclease that should cleave the RNA as well.
DR.
COFFIN: I have two questions. One is a somewhat trivial one but neither I
nor my colleague here actually knows what benzonase really -- is could you
quickly explain that?
DR.
KEMBLE: It is an endonuclease. It is used routinely in manufacturing for degradation
of residual nucleic acids.
DR.
COFFIN: It is derived from what? Is it plant, animal ‑
DR.
KEMBLE: It is bacterial derived ‑
DR.
COFFIN: The second is although I think
your calculations are in general fine you I believe assume probably a partial
distribution of cuts for the DNA in doing your size calculations. There are any number of reasons of course why
that might not be true and there might be DNA that is in the sample that is
somehow protected for example from digestion.
It would seem not all that difficult to actually do PCR based
experiments where with great sensitivity you could go and directly assess the
survival of much larger sized DNA pieces.
Have you actually done that kind of experiment?
DR.
KEMBLE: We have not done it yet but as
I said and clearly as we move this program forward we continue to develop the
process and the assays especially to start to characterize and understand the
residual in that vaccine. We do not have
the evidence from those cells I showed you.
DR.
COFFIN: That is very insensitive. Ten percent of your DNA could be in larger
fragments. Half of your DNA could be in
larger fragments and you would not be able to see it.
DR.
KEMBLE: It is but although we cannot --
remember there are very, very small amounts of DNA approximately 100 decagrams
so it is getting difficult ‑
DR.
COFFIN: But experiment is always
preferable to argument if it can be done and I think in this case it can be
done.
DR.
KEMBLE: Understood.
DR.
ROMERO: My question again goes back to
the benzonase treatment. Wouldn't that
add another layer of safety with regard to cell viability because the reaction
conditions would be hostile to the cell?
The second question is what is your quantity of bank cells and how long
will that carry you out?
DR.
KEMBLE: With respect to benzonase on
the intact cells there is actually a number of steps in our manufacturing
process which quite honestly are quite hostile to the intact cell but we have
not experimentally determined a number so we did not give ourselves credit for
that but the affinity chromatography many of the other steps is likely going to
kill them if they are there. We simply
are not taking credit for them at this point but yes I would expect it to.
The
second question I am blanking already ‑
DR.
ROMERO: The quantity of your bank
cells ‑
DR.
KEMBLE: We laid down our traditional
bank which consists of a master bank and then working cell banks are derived
from that. Quite honestly we would
expect it to be virtually indefinite with that type of process.
DR.
HUGHES: I certainly understand why you
did the test with the lysate with large amounts of lysate in smaller animals
there was probably nothing else to do. I
wonder if you have done or would consider doing experiments in which you
deliberately add to that sort of concentrated lysate the kinds of positive
controls with active oncogenic DNA. I am
not suggesting I think this is necessarily the case but I think it is
conceivable that if you use a massive dose that actually in some sense
interferes with the uptake and would give you a different outcome than if you
had lots of small doses.
DR.
KEMBLE: Again a lot of these protocols
were developed within from CBER and they are standard models and we did not
consider particularly for those assays adding those types of controls in
because we were not sure one how to manage them and what really level of
information that they would give us that made us feel safer at this point.
DR.
HUGHES: I think in this case I would
suggest that with your colleagues from CBER you might want to revisit
that. I think showing that there is no
negative impact of using large amounts on detecting oncogenic potential in your
lysate would give comfort and really I do not think given the potency of some
of the models that are now available have to work really hard and I think we
would all feel much better if you showed that the property protocol that you
were using with the lysate was really capable of detecting a small amount of
oncogene if it was there. I do not think
that should be an onerous request.
DR.
SELF: My question is how you calculate
the capacity of your sequence of filters.
I am probably missing something very simple but I would have guessed
that you would add the capacity of each element in your filtration and you
multiplied to get this very large 10 to the 21 number. Could you explain why you are multiplying
those rather than simply adding them up to get the total capacity of your filtration?
DR.
KEMBLE: It is a traditional way of
doing a safety factor assessment when you have one step that clears 10 to the
12th and one that clears 10 to the 9th you get the combine effort of 10 to the
21 and I think you cannot experimentally validate that 10 to the 21 number I
think is the issue.
DR.
SELF: I understand that but I well
maybe John can explain this.
DR.
COFFIN: I mean that carries with it the
assumption that the two steps are completely independent of each other and these
are both filtration steps so that assumption is somewhat questionable
actually. They could both be passing a
little bit for the same reason. Some
cells are extremely long and extremely thin for example.
DR.
KEMBLE: We did take efforts to make either
different membrane, different pore sizes they are from a filtration perspective
they are very different types of filtration discussed.
DR.
MODLIN: It is kind of angels on the
head of a pin argument isn’t it?
DR.
DEBOLD: Slide 34 I am not understanding
something here. Can you explain how you
get the denominators here? This is the
slide where you are describing your results as it relates to oncogenic
components. You said here in your test
animals it was 3 out of 270 and I am trying to figure out how this computation
was done. That is my first
question.
DR.
KEMBLE: So that answer is it is 45
animals for each of those cells that you see there so it was 0 out of 45 mice,
0 out of 45 hamsters, 0 out of 45 rats a culmination of 3 out of 270 that is
how we got the denominator for those.
DR.
DEBOLD: My next question and excuse me
I am very consumer rep here. I am trying
to understand why that in the three test animals that developed tumors why
isn't it a concern that the tumors are rodent origin?
DR.
KEMBLE: I think the two pieces of
information we have that helps us conclude that one is and I will remind you we
use very, very high doses of DNA per animal.
The rates are similar and in fact statistically indistinguishable
between the control and the test group.
Secondly these are tumor types that are routinely seen in these model
systems. As I said we did the PCRs with
a fairly highly distributed MDCK probe that again no evidence of any
tumorigenicity and certainly consisted with animal studies of this type.
DR.
MODLIN: Dr. DeBold does that answer
your question?
DR.
DEBOLD: I think so. It just sort of jumps out a little bit that
you know I realize there is probably no statically significant difference here
but it just seemed to me that perhaps it was relevant that there were tumors of
rodent origin. I am just trying to
understand this thing.
DR.
COOK: In the CBER reading material that
was sent to us there are comments about some kind of a ‑ this
relates to your animal model testing.
The numbers of for example nude mice that you tested for tumor induction
and so you were measuring what you looked for which is tumor development or
not. In this material it talks about a
cell dose related systemic illness that occurred in animals challenged with
MDCK cells. In thinking about other
thing that you might have observed other than subcutaneous nodules or systemic
tumors how were the animal’s health throughout the course of the tumor
challenge and I guess it was did you say six month observation or so? Were they all perfectly healthy? Did any of them have anything else that was
not related to tumor development?
DR.
KEMBLE: I am going to let Dr. Lauren
address that. She is our pathologist at
MedImmune and she knows more about the clinical observations and those
tumorigenicity studies than I do.
DR.
RICHMOND: The general health of the
animals was recorded on a regular basis and the only things that came up were
associated with what you would expect in an animal in a cage to be exposed so
sometimes they would get abscess no not abscess but just abrasions on the skin
but nothing associated with MDCK cells.
No systemic illness that was described in the briefing package that CBER
sent before and that again was with a different ATCC derived cell line not our
clone.
DR.
ROMERO: On your briefing document I had
a question for you. On page 24 you state
that the MDCK cell protein in the vaccine is 0.47 micrograms per dose. Is that per individual component so that if
you add the three consulates(?) Do you now go to 1.41 micrograms?
DR.
KEMBLE: No that is a total protein of
all things considered.
DR.
MODLIN: Could I ask a couple questions
about ‑ you told us that the MDCK is obviously has considerably more
value than the other cells that you tested initially but you did not really
show us any data and guess specifically that yield.
It
looks like you are getting about 10 to the 8th PFU per mL of cell culture fluid
with the original isolate. How does that
compare to the other cell lines particularly let's say VERO cells which we have
as sort of an approved cell line for use of live cells for going even live
virus vaccine. I guess a related
question how does that compare to the yield in eggs?
DR.
KEMBLE: So let me take that into two
parts. This is our current subclone and
the productivity of I think 21 different strains anywhere from H1s, H3s, Ds(?),
and a small number of cold‑adapted live attenuated pandemic subtypes and
these are all ranging in the mid 8s to even up to 9s. So these are actually comparable, slightly
smaller than eggs but not at this point we do not have a statistical
comparison. But we would be happy with
most of these titers if it was egg production with the one outlier there.
As far as the other cell lines and
VERO is obviously one that we spend a lot of time thinking about the bottom
line was there were certain strains and we are unclear although we believe it
is partly due to our Ann Arbor backbone in our vaccines that these strains grew
very, very poorly titers of four or five logs which are unusable for us. That was true for a lot of those other types
that we would have had an easier path with MRC‑5, WI‑38. Some of these have been reported to grow wild‑type
flu. Certainly others have used VERO for
wild‑type flu and