1
This transcript has not been
edited or corrected, but appears as received from the commercial transcribing
service. Accordingly, the FDA makes no
representation to its accuracyY@
1
2
4 FOOD AND DRUG ADMINISTRATION
5 CENTER FOR BIOLOGICS EVALUATION AND RESEARCH
6
7
8
9 BIOLOGICAL RESPONSE MODIFIERS
10 ADVISORY COMMITTEE (BRMAC)
11 Meeting 36
12"This transcript has not been
edited or corrected, but appears as received from the commercial transcribing
service. Accordingly, the FDA
makes no representation to its accuracy..."
13
14
15
16
17
18
19
20
21 Gaithersburg, Maryland
22 Thursday, October 9, 2003
2
1
PARTICIPANTS:
2 BRMAC MEMBERS:
3 MAHENDRA S. RAO, Acting Chair
National Institute on
Aging
4
JONATHAN S. ALLAM
5 Southwest Foundation for Biomedical Research
6 BRUCE R. BLAZAR
University of Minnesota
7
DAVID M. HARLAN
8 National Institute of Diabetes and Digestive
and Kidney Disease
9
KATHERINE A. HIGH
10 University of Pennsylvania
11 JOANNE KURTZBERG
Duke University Medical
Center
12
ALISON F. LAWTON
13 Genzyme Corporation
14 RICHARD C. MULLIGAN
Harvard Medical School
15
ANASTASIOS A. TSIATIS
16 North Carolina State University
17 ALICE J.
WOLFSON
Wolfson &
Schlichtmann
18
TEMPORARY VOTING MEMBERS:
19
JAMES F. CHILDRESS
20 University of Virginia
21 LYNNE L. LEVITSKY
Harvard Medical School
22
3
1
PARTICIPANTS (CONT'D):
2 TEMPORARY VOTING MEMBERS (CONT'D):
3 ABBEY S. MEYERS
National Organization for
Rare Disorders
4
CAROLE B. MILLER
5 St. Agnes Healthcare
6 W. MICHAEL O'FALLON
Mayo Clinic
7
DANIEL R. SALOMON
8 The Scripps Research Institute
9 ROBERT S. SHERWIN
Yale University School of
Medicine
10
JANET H. SILVERSTEIN
11 University of
Florida College of Medicine
12 CONSULTANTS:
13 JOHN J. O'NEIL JR.
LifeScan, Inc.
14
CAMILLO RICORDI
15 University of Miami School of Medicine
16 GUESTS/GUEST SPEAKERS:
17 BERNARD J. HERING
University of Minnesota
18
JAMES SHAPIRO
19 University of Alberta
20 THOMAS L. EGGERMAN
National Institute of
Diabetes and Digestive
21 and Kidney Diseases
22
4
1
PARTICIPANTS (CONT'D):
2
GUESTS/GUEST SPEAKERS
(CONT'D):
3 JAMES BURDICK
Health Resources and
4 Services Administration
5 FRANCISCA AGBANYO
Health Canada
6
FOOD & DRUG ADMINISTRATION (FDA) PARTICIPANTS:
7
JESSE L. GOODMAN
8 Center for Biologics Evaluation
and Research
9
KATHRYN CARBONE
10 Center for
Biologics Evaluation
and Research
11
RAJ PURI
12 Center for Biologics Evaluation
and Research
13
CAROLYN WILSON
14
Center for Biologics
Evaluation
and Research
15
ANDREW BYRNES
16 Center for Biologics Evaluation
and Research
17
NANCY MARKOWITZ
18 Center for Biologics
Evaluation
and Research
19
STEVEN BAUER
20 Center for Biologics Evaluation
and Research
21
22
5
1
PARTICIPANTS (CONT'D):
2 FDA PARTICIPANTS (CONT'D):
3 AMY ROSENBERG
Center for Drug
Evaluation
4 and Research (CDER)
5 EMILY SHACTER
Center for Drug
Evaluation
6 and Research
7 GAIL DAPOLITO
Executive Secretary
8 Center for Biologics Evaluation
and Research
9
ROSANNA L. HARVEY
10 Committee Management Specialist
Center for Biologics
Evaluation
11 and Research
12 BRMAC #36 FDA PLANNING COMMITTEE MEMBERS:
13 PHILIP NOGUCHI
Center for Biologics
Evaluation
14 and Research
15 CYNTHIA RASK
Center for Biologics
Evaluation
16 and Research
17 DARIN WEBER
Center for Biologics
Evaluation
18 and Research
19 DWAINE RIEVES
Center for Biologics
Evaluation
20 and Research
21 KEITH M. WONNACOTT
Center for Biologics
Evaluation
22 and Research
6
1
PARTICIPANTS (CONT'D):
2 BRMAC #36 FDA
PLANNING COMMITTEE MEMBERS
(CONT'D):
3
NICHOLAS I. OBIRI
4 Center for Biologics Evaluation
and Research
5
RICHARD McFARLAND
6
STEPHEN GRANT
7
SUSAN LEIBENHAUT
8
JOHN ELTERMANN JR.
9
JOHN FINKBOHNER
10
SARAH KIM
11
SUSAN ELLENBERG
12
GHANSYAM GUPTA
13
ROBERT MISBIN
14
15
16
17
18
19
20 *
* * * *
21
22
7
1 C O N T E N T S
2
AGENDA SESSION: PAGE
3
Session 1:
4
Conflict of Interests
Meeting 11
5 Statement
6 FDA Introduction 16
7 Federal Oversight of Allogenic 37
Islet Transplantation
8
Moving from Investigational
Islet 74
9 Products to Licensed Islet Products
10 Islet Processing: Evolution 135
and Current Standards
11
Current Status of Islet 172
12 Characterization and Quality
13
Session 2:
14 Laboratory of Stem Cell Biology 412
15 Laboratory of Immunology 426
and Virology
16
Laboratory of
Biochemistry 458
17
*Proceedings of CLOSED SESSION
at pages
18
421-425 and 472-488 bound separately per
request
19
20
21 *
* * * *
22
8
1 P R O C E E D I N G S
2 (8:08 a.m.)
3 DR. RAO:
Good morning, everyone.
4
This is the 36th meeting of the BRMAC.
5 I'm going to just start by asking
6
everybody on the committee to introduce
7
themselves, and to point out that they are
8
going to use the microphone system like
9
you've done before, where you're going to
10
wait for the red light and wait to be
11
recognized by the chairman.
12 We'll start with introductions
13
from the left.
14 DR. SHERWIN:
On the left, Bob
15
Sherwin from Yale. I'm a
professor of
16
medicine there.
17 DR.
LEVITSKY: Lynne Levitsky.
18
I'm chief of the Pediatric Endocrine Unit at
19
Mass. General in Boston.
20 DR. CHILDRESS:
Jim Childress,
21
University of Virginia. I
specialize in
22 Bioethics.
9
1 MS. MEYERS:
Abbey Meyers,
2
President of the National Organization for
3
Rare Disorders, and I have diabetes.
4 DR. O'FALLON:
Michael O'Fallon,
5
biostatistician, Mayo Clinic.
6 DR. ALLAN:
I'm Jon Allan,
7
Southwest Foundation for Biomedical Research
8
in San Antonio, Texas. I'm a
virologist and
9
I study AIDS pathogenesis and animal model
10
systems.
11 MS. LAWTON:
Allison Lawton,
12
Genzyme Corporation, and I'm the industry
13
rep on the panel.
14 DR. KURTZBERG:
Joanne Kurtzberg.
15
I'm a pediatric hematologist at Duke
16
University and run the pediatric bone marrow
17
transplant program.
18 DR. BLAZAR:
Bruce Blazar,
19
University of Minnesota. I'm
involved in
20
pediatric organ transplantation and
21
immunology.
22 DR. RAO: I'm
Mahendra Rao. I'm
10
1
at the National Institute on Aging and I'm a
2
stem cell biologist.
3 MS. DAPOLITO:
Gail Dapolito,
4
executive secretary for the committee.
5 DR.
HIGH: Kathy High. I'm a
6
hematologist at the Children's Hospital in
7
Philadelphia.
8 DR. SALOMON:
Dan Salomon. I'm
9
the director of the Center for Organ and
10
Cell Transplantation, and I'm a transplant
11
physician. I'm also the chair of
the
12
NIH/NCRR Islet Cell Resources Steering
13
Committee.
14 DR. O'NEIL:
Jack O'Neil. I'm
15
with Johnson & Johnson, principal scientist
16
in the Center for Diabetes Advances.
17 DR. BURDICK:
Jim Burdick. I'm a
18
transplant surgeon, and I'm presently
19
director of the Division of Transplantation
20
in HRSA.
21 DR. AGBANYO:
I'm Francisca
22
Agbanyo. I'm from Health Canada,
which is
11
1
the agency that regulates therapeutic drugs
2
in Canada.
3 DR. RASK:
I'm Cynthia Rask. I'm
4
the director of the Clinical Evaluation and
5
Pharmacology/Toxicology at FDA CBER.
6 DR. WEBER:
Good morning. I'm
7
Darin Weber. I'm chief of the
Cell Therapy
8
Branch at the Division of Cell and Gene
9
Therapy in the Office of Cellular Tissues
10
and Gene Therapies.
11 DR.
NOGUCHI: I'm Phil Noguchi,
12
acting director of the Office of Cellular
13
Tissue and Gene Therapies.
14 DR. GOODMAN:
I'm Jesse Goodman,
15
Director of CBER. I guess I have
a conflict
16
of interest in that I used to be at the
17
University of Minnesota.
18 DR. RAO: I
guess, Gail, you need
19
to read the statement.
20 MS. DAPOLITO:
Good morning. The
21
following announcement addresses conflict of
22
interest issues associated with this meeting
12
1
of the Biological Response Modifiers
2 Advisory Committee on
October 9
3
and 10, 2003.
4 Pursuant to the authority granted
5
under the committee charter, the associate
6
commissioner for external relations, FDA,
7 appointed Drs.
Lynne Levitsky, Robert
8
Sherwin and Janet Silverstein as temporary
9
voting members.
10 In addition, the director of FDA's
11
Center for Biologics Evaluation and Research
12
has appointed Drs. James Childress, Michael
13
O'Fallon, Carole Miller, Daniel Salomon, and
14
Ms. Abbey Meyers as temporary voting
15
members.
16 Based on the agenda, it was
17
determined that there are no products being
18
approved at this meeting. The
committee
19
participants were screened for their
20
financial interests to determine if any
21
conflicts of interest existed.
22 The agency reviewed the agenda and
13
1
all relevant financial interest reported by
2
the meeting participants. In
accordance
3
with 18 U.S.C. 208, the following special
4
government employees were granted waivers
5
for their participation: Dr.
Bruce Blazar
6
was granted a full waiver that permits him
7 to participate
in the committee discussions.
8 In addition, a limited waiver is
9
granted to Dr. Camillo Ricordi so that he
10
may make a presentation and answer questions
11
regarding his presentation.
12 Dr. David Harlan recused himself
13
from the discussions on human allogeneic
14
islet transplantation.
15 We would like to note for the
16
record that Ms. Alison Lawton is
17
participating in this meeting as a
18
non-voting industry representative, acting
19
on behalf of regulated industry.
20
Ms. Lawton's appointment is not subject
21
to 18 U.S.C. 208.
22 She is employed by Genzyme
14
1
Corporation and thus has a financial
2
interest in her employer.
Genzyme has
3
associations with universities,
4
investigators and research foundations.
5 With regards to FDA's invited
6
guests and consultants, the agency has
7
determined that their services are
8
essential. The following
disclosures will
9
assist the public in objectively evaluating
10
presentation and/or comments made by the
11
participants for the discussions on human
12
allogeneic islet
transplantation.
13 Mr. Jack O'Neil Jr. is employed as
14
a principal scientist at Life Scan Center
15
for Diabetes Advances, Johnson & Johnson.
16 Dr. Francisca Agbanyo is employed
17
as a director, Biologics and Genetic
18
Therapies, Biologics and
19
Radiopharmaceuticals Evaluation Center,
20
Health Canada. Dr. Agbanyo is a
Canadian
21
government official involved in the
22
regulatory oversight of islet cell
15
1
transplantation.
2 Dr. Tom Eggerman, who just joined
3
us, is employed as the director, Islet
4
Transplantation Program, Division of
5
Diabetes, Endocrinology and Metabolic
6
Diseases, National Institute of Diabetes and
7
Digestive and Kidney Diseases.
8 Dr. Eggerman is
participating in
9
this meeting as part of his official
10
government duties. His division
is involved
11
in funding and monitoring a grant in islet
12
transplantation.
13 Dr. Bernhard Hering is employed as
14
the director, Islet Transplantation at the
15
University of Minnesota Medical School.
16
Dr. Hering is directly involved in islet
17
transplantation research.
18 Dr. James Shapiro is employed at
19
the University of Alberta, Clinical Islet
20
Transplant Program. Dr. Shapiro
is directly
21
involved in islet transplantation research.
22 The committee discussions of
16
1
Topic 2, relating to FDA's individual
2
research programs, present no potential for
3 a conflict of
interest. FDA's participants
4
are aware of the need to exclude themselves
5
from the discussions involving specific
6
products or firms that they have not been
7
screened for conflict of interest.
Their
8
exclusion will be noted for the public
9
record.
10 With respect to all other meeting
11
participants, we ask in the interest of
12
fairness that you state your name,
13
affiliation, and address any current or
14
previous financial involvement with any firm
15
whose products you wish to comment upon.
16
Waivers are available by written request
17 under the
Freedom of Information Act.
18 Thank you.
19 DR. RAO: We
have an introduction
20
by the FDA. Dr. Noguchi will
present it.
21 DR. NOGUCHI:
On behalf of CBER,
22
I'd like to welcome all of you, especially
17
1
those in the audience and our advisory
2
committee members to this, I guess it's
3 the 36th meeting
of the BRMAC committee.
4 We are now at a point, and really
5
reviewing a lot of data that has been
6
gathered, a lot of interest that has been
7
gathered since I guess it was March of 2002,
8
when we talked about this before.
9 We feel that its now time to
10
really have the critical discussion to be
11
provided today by all the members of the
12 committee, all the
members of the public and
13
our experts to really see what's going on.
14
It's always nice to really make a critical
15
assessment when it's an appropriate time.
16 I'd like to
also now take just a
17
little bit of time to introduce our center
18
director, Dr. Jesse Goodman. You
know, at
19
CBER, center directors tend to be rather
20
extensive in their stay. I've actually
been
21
here ever since the center has moved from
22
the NIH over to FDA, and I think Jesse is
18
1
now the fourth director.
2 The first one was Dr. Hank Meyer,
3
for about ten years, and Paul Parkman
4
another ten years, Kathy Zoon roughly ten
5
years. We certainly hope Dr.
Goodman will
6
continue in that tradition.
He'll have even
7
less hair when he's finished here.
8 But Dr. Goodman is exceptionally
9
well-qualified. He actually came
to the
10
government in the commissioner's office to
11 head up a program
on infectious diseases and
12
emerging infectious diseases, and has been
13
through a number of different capacities,
14
more recently as deputy director for
15
medicine at CBER.
16 Then as Dr. Zoon moved over to the
17
Cancer Institute, he came and graciously
18
accepted the title of director of CBER.
I
19
know for both him and the rest of us, it's
20
been a very steep and growing learning
21
curve, but we are very proud, myself, to
22
have him here today to give an opening
19
1
introduction of the vision of CBER and how
2
this particular area of cell and gene
3
therapy and tissues really plays into that.
4 Dr. Goodman.
5 DR. GOODMAN:
Good morning. I do
6
thank everybody for being
here. I thank
7
Phil for the kind introduction.
Actually, I
8
feel very privileged to be at CBER and to be
9
here today.
10 I'll use just a few minutes of
11 everybody's time in
a little bit of
12
opportunity to say a couple of things about
13
what's happening at FDA in general and at
14
CBER, and how working on this new technology
15
really fits into our vision of bringing safe
16
and effective products to people.
17 Also, I just want to share that
18
Dr. McClelland, the commissioner, is very
19
interested also in this subject and this
20 meeting, and
he's sorry that he couldn't be
21
here today with us to say hi at least.
22 Anyhow, as some of the pictures I
20
1
put on the opening slide show, this is
2
diabetes review, as a very important disease
3
and a very high priority.
Obviously, even
4
if you look at this week's "JAMA," we're
5
having a huge diabetes epidemic.
6 Those of us who do medical care
7
certainly understand the epidemic of
8
complications of diabetes and organ damage.
9
Of course, what we are here about is islet
10 cell transplantation and the potential for
11
that to address at least some of these
12
problems. Obviously, many of
these
13
problems, we could also prevent.
So that's
14
a whole other area.
15 I'd just
like to, for the members
16
of the committee and others, just sort of
17
give a little overview of just where this
18
fits in the spectrum of what we're facing at
19
FDA and CBER in general.
20 We are dealing with constant
21
related challenges, many of which are quite
22
acute and urgent at times; vaccine safety
21
1
and availability; blood safety and
2
availability. We have advisory
committees
3
that help us with this.
4 Many emerging infectious disease
5
issues. It seems that just when
we've done
6
what we need to do to have things moving on
7
one, SARS or Monkeypox comes along.
These
8
also, I think, as you think forward in
9
cellular therapy, the areas that the BRMAC
10 is concerned
on, are things that as these
11
things become incorporated more into medical
12
practice, you're going to need to be
13
thinking about as well.
14 Human tissue cell products and
15
gene therapy are very high priorities for
16
us. As you know, Kathy Zoon,
working with
17
Phil, moved this into the status of a new
18
office recently, and this signifies the
19
importance.
20 Again, our center has really borne
21
the brunt of dealing with bioterrorism
22
issues and counter-terrorism.
Again, all of
22
1
this is urgent. All of this is
24/7, and
2
all of it, I think, requires us to work in
3
better and new ways to try to bring products
4
along.
5 Well, Dr. McClelland has put forth
6
a vision through a strategic plan, and
7
again, these are very broad areas, the plan
8
is quite detailed, but I just thought, for
9
those who haven't heard them, I'd share them
10 with you: Science-based risk management,
11
which simply means exactly what we often do
12
in advisory committees, looking at the
13
information and making the best possible
14
decisions, being sure that we're paying
15
attention to the consumer and the patient in
16
terms of information, patient safety, that
17
the products are used wisely and safely.
18 Counter-terrorism has made it up
19
there, as you can see, for the reasons that
20
I said.
21 Then attention to a strong FDA,
22
and this has been a particular challenge for
23
1
everyone, of course, with the fiscal
2
restraints that we're all under.
But also
3
it's a challenge just for the government in
4
general to have personnel processes and
5 infrastructure
which enable us to deal with
6
important issues like this.
7 Part of the strength of FDA is in
8
our processes, whether it's review, science,
9
administrative processes to enhance the
10
availability of new technologies.
This is
11
something that I really believe in and
12
Dr. McClelland does as well. I
think where
13
all of these, and particularly these are
14 very pertinent
to CBER's mission, and our
15
actions do support this plan.
16 I'd just like to identify a few
17
other general areas which I think are
18
emerging as high priorities and ways of
19
getting there for our center in general, and
20
again it fits very well with what you're all
21
doing here today.
22 I think we really, as a center
24
1
actually within FDA, we've really been very
2
active in seeking outside collaboration and
3
input, whether it's in the science that
4
people do or whether it's in our review and
5
regulatory work. But we want to
do that
6
more. That involves doing things
like
7
bringing issues to groups like you, getting
8
input, perhaps sometimes even when there is
9
not a regulatory decision at that time, but
10
to help move a field along and help be sure
11
we're getting the best information to move
12
in the right direction.
13 Again, this doesn't just include
14
science, but includes the public, et cetera.
15
We need to really, as per a strong FDA,
16
strengthen the base for and performance of
17
CBER and its collaborative science.
Again,
18
I'm trying to extend the vision of science
19
not just to include laboratory science, but
20
epidemiologic, clinical science and
21
expertise, risk science.
22 We want to try to identify what
25
1
are the stumbling blocks to product
2
development and new technologies; you know,
3
which ones are fixable, which ones aren't,
4
and can we help be a partner in removing
5
those. Part of this is enhanced
6
interactions with all kinds of partners,
7
ranging from our colleagues at NIH, other
8
regulatory authorities. So we
have HRSA
9
here today on the organ transplant front,
10
and other partners.
11 Again, in all of this to the
12
degree that we can get input, look at what
13
we do, have it be focused, have a lot of
14
transparency, it's helpful.
15 Okay, just other major areas: As
16
you know, we have a new office and we really
17
appreciate what Phil and many of the others
18
in this room have done getting that going
19
under very challenging circumstances.
20 This is a key office in this
21
technology facilitation. I mean,
so much of
22
the promise of medicine is stuff that is
26
1
brought to your committee and are things in
2
these therapeutic areas.
3 I mentioned emerging infectious
4
diseases, but you may not realize we have
5
several issues. The protection
of blood
6
cell vaccine tissue safety is the most
7
obvious one, and it's the one we always deal
8
with every Friday afternoon at
5:00.
9 But also, we have a role in
10
products for prevention. We have
been very
11
active in trying to facilitate development,
12
for instance, of West Nile, SARS vaccines,
13
et cetera, as well as in treatment and
14
diagnosis, and tried to have a more systemic
15
approach to this also, not just in CBER, but
16
across FDA.
17 We have needs, based on what's
18
gone on in the world and what's gone on with
19
our products, to strengthen our emergency
20
response in crisis management.
We have some
21
external metrics that we work on with
22
industry that also involve increasingly the
27
1
support for our activities. So
we have the
2
User Fee Act's prescription drugs which
3 affects our
licensed biologic products, and
4
now a medical device User Fee Act.
5 Again, some people here may not
6
realize it, but quite a number of devices
7
involved in the safety of the blood supply
8
or in preparing cellular therapies such as
9
centrifuge and cell separators, et cetera,
10
are handled within CBER, and ideally, that
11
works by having us have a consistent view of
12
them as part of a product and a product
13
development and a system.
14 Then in all of this, we view our
15
primary function for the American public in
16
terms of safety and efficacy of our
17
therapies, and we want to see that we manage
18
our review process with high quality and
19
with consistency, and with the incredible
20
variation of products which FDA faces and
21 the incredible
creativity that's out there
22
in the academic and industrial world, this
28
1
is a very challenging thing to do.
2 So we do a
lot of innovative
3
technology in public health, and that's
4
really a big part of our vision at CBER.
5
These issues are uniquely focused within our
6
center. Looking at how we view
our mission
7
and our goals, certainly we want to protect
8
and improve public and individual health in
9
the U.S., and also where feasible, globally.
10
So we want to be good global partners.
11 We
want to facilitate development
12
approval and access to safe and effective
13
products and promising new technologies. We
14
want to strengthen CBER as a preeminent
15
regulatory organization for biologics, and
16
one that performs in an excellent manner
17
here and performs in an excellent manner
18
with our international partners.
19 So what about what we're here
20 today
about? It is about trying to find
21
safe and effective promising new
22
technologies. As a general goal,
we really
29
1
see ourselves having a role in assisting
2
product development in nascent fields across
3
industries. How can we help?
4 Examples are what we have had to
5
do in bioterrorism preparedness, gene
6
therapy, new areas like tissue engineering,
7
stem cells, cell therapies, new vaccine
8
technologies, and even in areas like blood,
9
which will seem to the clinician as quite
10
traditional and staid, there's a lot of
11
potential in terms of oxygen carriers,
12
pathogen inactivation, better pathogen
13
detection.
14 Well, how can we help? Most of
15
the ferment and most of the ideas are coming
16
from everyone, from industry, et cetera.
17
But our guidance, our standards, our
18
outreach, our policy can be creative.
19
Looking at safety and efficacy, we need to
20
find the best pathways to do that.
21 We need to work with the other
22
partners to improve our risk communication
30
1 to the public. We need
to be sure we have
2
the right internal expertise and the right
3
partnerships.
4 So based on those kinds of
5
priorities and that kind of vision, I think
6 we're here today to
learn about and get
7
input about, and also hopefully provide some
8
helpful thinking about islet cell
9
transplantation.
10 Again, you guys, hearing about all
11 of you and hearing about
some of our guests
12
here, you know, you truly are the experts.
13
I'm not really here to tell you any of this,
14
but what are some of the issues here?
15 Well, this is
the perfect example
16
of a promising technology, where if we can
17
work well with partners, we can perhaps play
18
a facilitating role at sorting out how is
19
that promise best directed and how is it
20
best evaluated.
21 This addresses a major unmet
22
public health need. We know that
even with
31
1
increasingly sophisticated treatments for
2
diabetes, we've got a long way to go, and
3
there are patients who nonetheless develop
4
serious complications.
5 There's a real need here to help
6 define and
guide product development in
7
regulatory pathways. We may not
have all
8
the information we need to know how to do
9
that right. But we have to do
our best.
10 We have to assure safety and
11
effectiveness, but hopefully not inhibit
12
availability of effective therapies, and
13
based on limited information, set up a
14
system or standards that inhibit future
15 improvements or
innovation. These are all
16
very fine balancing acts.
17 So what are some of the issues
18
that I think, I'm sure, are going to come
19
up? I've seen them in some of
the
20
materials. They are just the
ones that came
21
to me in thinking about this.
22 Organ availability, the variations
32
1 in short-term
outcomes that have been
2
observed in some of the studies in different
3
centers, centers, the quality of the
4
materials, the procedures themselves.
This
5
is the focus of a lot of today's discussion.
6
I'm not sure as much is known about patient
7
variables.
8 There are acute adverse event
9
outcomes that we need to consider in any new
10
therapy, and there are, as I said, all these
11
issues of product characterization:
The
12
quality, the quantity. And again
I think we
13
know less on a more sophisticated level
14
about the functionality of quality and how
15
to measure it.
16 Very important, as one thinks
17
about moving forward beyond some of the
18
really incredibly exciting promise that
19
we've heard about, is the issue of long-term
20
outcomes; to what extent do we restore
21
normal metabolic function.
Again, what are
22
the predictors of doing that or not doing
33
1
that? How can we do better? What are the
2
adverse effects of cell therapy?
3 We don't know much long-term.
4
Immunosuppression, I think we're starting to
5
have a pretty good database from lots of
6
other kinds of transplantation in terms of
7
long term and organ effects, benefits and
8
just the real outcomes that matter to human
9
beings in terms of morbidity, mortality,
10
quality of life.
11 Probably the people who do this
12
are thinking about it, but you don't see too
13
much thinking about who are the patients
14
most likely to benefit, early versus late
15
treatment? How do you assess
efficacy? We
16
deal with this frequently with exciting new
17
products: Are there ways to look
at
18
historical data? Are there
issues that
19 require control
groups?
20 What can be the pathways to
21
clinical success, and a somewhat different
22
question, to licensure? How much
data would
34
1
one want to see? How much
benefit does one
2
need to see to adopt a therapy like this?
3 The world of clinical medicine is
4
filled with examples, both where promising
5
new technologies are documented and
6
effective technologies have been adapted too
7
slowly.
8 But it's also filled with examples
9
where things that seem promising have been
10
adapted on a widespread basis, only to be
11
found not to really have the effect we
12
thought they would.
13 You know, autologous breast cancer
14
transplantation is a reasonable example of
15
that.
16 So I think no matter what you all
17
do and what this community does, we need to
18
think about how do you do long-term
19
assessment and how do you improve these
20
technologies as they move forward.
21 So anyhow, your input and the
22
broader community as well is very welcome
35
1 and
critical. We're going to hear it today.
2
I won't be able to be here for all of it,
3
but I'll find out about it.
We're really
4
going to work together with you and the rest
5
of the appropriate communities to try to
6
refine and develop this promising advance.
7 We'll try to do our part and
8
synthesize this information and provide the
9
best possible guidance to help this along.
10
I think the good thing is that we do have a
11
common goal here, which is to get therapy to
12
people. But a lot of our role is
to be sure
13
it's safe and effective, and we all want to
14
see better outcomes and quality of life.
15 Another point that really hit me
16
in thinking about this, and I think you
17
folks in this field should think very
18
carefully about, is what we learn, and
19
hopefully the successes that are achieved,
20
but also, as always, the things that we
21
didn't succeed in are going to be critical,
22
not only for these patients with diabetes,
36
1
but I think this a first event, potentially,
2
in the development of cellular therapies for
3
a variety of diseases.
4 There's a lot
of complexities here
5
from the clinical, the scientific and the
6
regulatory point of view that the more
7
thoughtful we are about this and the more we
8
learn from it, the better.
9 So with that, anyhow, I probably
10
took far too long and you knew this all
11
already, but I wanted to indicate how
12
important we thought this was and how
13
seriously we take outside input in this
14
area. So thanks.
15 DR. RAO:
Thank you, Dr. Goodman.
16
It's very useful to reemphasize the fact
17
that what we discuss today may have general
18
application related to all other stem cell
19
therapies as well.
20 We're going to have four speakers
21
from the FDA which will sort of set the base
22
for questions and issues. I'm
going to
37
1
request that if possible, unless it's a
2
burning question, that you hold it towards
3
the end because there will be overlapping
4
things which might possibly be
answered by
5
the subsequent speakers.
6 But if you need to ask a question,
7
just feel free to press the button.
8 DR. WEBER:
Good morning,
9 everyone. Thank you, Mr. Chairman and
10
members of the committee, and the support of
11
Dr. Goodman and Dr. Noguchi for supporting
12
this meeting this morning.
13 My task is to provide basically a
14
kind of an introduction to the topic as well
15
as some of the overview of FDA's regulatory
16
issues for allogeneic islet transplantation,
17
in about ten minutes or less. So
I'm going
18
to go fairly rapidly to try to give you a
19
sense of what we're trying to accomplish
20
today.
21 It's always helpful to know how we
22
got here. This is, of course,
how we got
38
1
here from the FDA's perspective.
We'll
2
discuss some of the goals of this meeting,
3
what we hope to get out of the meeting from
4 the FDA's side.
5 I'll briefly mention some of the
6
federal agencies who are involved in the
7
U.S., as well as provide an introduction to
8
the FDA's questions, and then set up the
9
stage, if you will, for the discussion
10
that's going to follow by introducing the
11
speakers.
12 Again, this is a cartoon. It
13
represents a timeline again from the FDA's
14
perspective. We certainly
acknowledge that
15
there's been a lot of research going on in
16
this field outside the purview of the FDA.
17
So I'll just recognize that. But
what I'm
18
trying to tell you here is that
for a
19
ten-year period between 1990 and 2000, the
20
FDA received a total of ten islet INDs for
21
allogeneic islet transplantation, which was
22
somewhat indicative of not a lot of rapid
39
1
progress being made in this field.
2 I think, as you all know, a lot of
3
that changed in the year 2000, for a variety
4
of reasons shown here. As Phil
indicated,
5
back in March of 2000, we had another
6
advisory committee, another BRMAC meeting on
7
the same topic. That meeting was
primarily
8
focused on some of the fundamental
9
regulatory issues for regulating this
10
therapy under IND, so we talked about
11
pre-clinical models; we talked about some of
12
the fundamental manufacturing information as
13
well as clinical issues that should be
14
included in an IND.
15 Of course, that was in many ways
16
an anticipation of the publication of
17
Dr. Shapiro's group, the Edmonton Protocol,
18
in "The New England Journal" in July
19
of 2000.
20 Then, subsequent to that, FDA sent
21
a follow-up letter; basically a Dear
22
Colleague letter to all the organ transplant
40
1
centers in the U.S., basically reminding
2
them that in fact, this therapy is regulated
3
by the FDA, and if you want to treat
4
patients, you would need to submit an IND.
5 So in many ways for the FDA, I
6
think this was a threshold year.
Of course,
7
subsequent to that and prior to this time,
8 there's been a lot of funding in this area
9
by many different organizations.
Obviously,
10
the JDRF is a major player, the Juvenile
11
Diabetes Research Foundation, and of course,
12
various institutes at the NIH.
13 I think it's fair to say it's
14
borne quite a bit of fruit. If
you can see
15
it here, the graphics show a little detail
16
here.
17 What I'm just trying to show here
18
is prior to 2000, which is right about here,
19
there is a low level of activity, and then a
20
real significant jump since that time.
We
21
have received about 28 islet INDs
22 since 2000, which
represents obviously a
41
1
significant work load for the FDA as well as
2
tremendous interest in the community for
3
this therapy.
4 So that brings us to today and
5
what some of the goals of this meeting are.
6
Honestly, we, from the FDA, have wanted to
7
talk about expectations, manufacturing data
8 and clinical
evidence that we would like to
9
see in a BLA, a Biologics License
10
Application, that would subsequently lead to
11
the approval for this therapy for Type One
12
Diabetes.
13 Of
course, in that context, we
14
want to get advice and perspectives from you
15
folks on the committee in terms of
16
discussing the data that you think should be
17
provided in a BLA.
18 Certainly, it's very important for
19
us to do this is in a public forum, to get
20
input and feedback from stakeholders who
21
obviously have a strong interest in this
22
therapy.
42
1 So this slide just gives you a
2
plethora of acronyms of various federal
3
agencies in the U.S. who are involved.
4 HRSA, of course,
is the Health Resources
5
Service Administration. They
are, of
6
course, involved in organ procurement and
7
allocation in the U.S.
8 Of course, the FDA is interested
9
in the regulatory oversight of chemical uses
10
of pancreatic islets.
11 Of course, our NIH colleagues, who
12
are involved with basic research as well as
13
clinical research for islet transplantation.
14
I think we have been very fortunate at FDA
15
to have a very good collaborative working
16
relationship with our colleagues.
Many of
17
them are here, and Dr. Eggerman's on the
18
committee as well.
19 Last, but not least, of course, is
20
the role that the Centers for Medicare and
21
Medicaid will play in terms of reimbursement
22
issues. I just wanted to point
out,
43
1
obviously, the whole issue of reimbursement
2
is something that's really beyond the scope
3
of the FDA and it's something we're not
4
going to talk about today at this meeting,
5
but we just want to acknowledge, obviously,
6
it's an important issue that is going to
7
have to be addressed in a different forum.
8 So now,
moving into the more
9
specific questions the FDA would like to
10
discuss in terms of islets as a license
11
product in terms of the manufacturing
12
issues; obviously, islets would need to be
13
prepared in a well-established manufacturing
14
process. We'd need a document
record of
15
manufacturing consistency to support a
16
license application.
17 Of course, the islets would have
18
to be prepared in a facility that is meeting
19
current GMP, or good manufacturing
20
practices, as well as, of course, complying
21
with the lot release test requirements for
22
these biological products.
44
1 So the FDA presentations this
2
morning will cover this. Mr.
Wonnacutt will
3
talk about the first and the third bullet,
4
and Dr. Obiri from the FDA is going to be
5
talking about the manufacturing issues.
6 So as a sneak preview, if you
7
will, in terms of the questions we're going
8
to ask, these are just paraphrased, and the
9
committee has a little more detail about the
10
background to these questions.
11 But, obviously, there's an
12
interest from the FDA in all aspects of
13
manufacturing as well as the delivery of the
14
product to the patients. So we
certainly
15
recognize that source organs are a
16
challenge, obviously, coming from a
17
cadaveric organ, coming from the organ
18
procurement system that's overseen by HRSA.
19 So we'd obviously like to have a
20
discussion concerning the use of basic
21
manufacturing experience data that's
22
currently being collected under IND that
45
1
would help establish pre-defined acceptance
2
criteria for these source donor organs, the
3
idea being only to include high-quality
4
organs while excluding unsuitable organs for
5
islet processing.
6 Moving down the manufacturing
7
scheme here in terms of dissociation
8
enzymatic and mechanical dissociation to get
9
islets; obviously, for a license
10
application, you're going to need a
11
well-controlled manufacturing process.
12 We also realize that the FDA is
13
going to have to be balanced by the need for
14
some flexibility in the manufacturing, and
15
again, recognition of the source material
16
here.
17 So again,
we would like to have a
18
discussion about the use of data being
19
collected under INDs that again can help
20
predetermine under what conditions various
21
reagents can be used in terms of helping to
22
optimize the yield of islets. A
lot more
46
1
detail about this will be discussed by
2
Mr. Wonnacutt in his presentation.
3 Of course, lot release testing,
4
making sure the product has a quality in the
5
safety characteristics before it's delivered
6
to the patient. One particular
type of lot
7 release in terms of
islet potency; basically
8
the idea is that an assay that can be
9
predictive of the ability of the islets, in
10
this case, to perform as expected.
11 So there's a variety of assays
12
being done. Many of them are
retrospective,
13
meaning the results are only available after
14
the patient receives it under the IND.
Of
15
course, for a licensed product, we need a
16
prospective assay, an assay that's available
17
prior to transplantation.
18 So I think there's a lot of
19
opportunity for discussion on this issue.
20 So the fourth manufacturing
21
question is dealing with comparability.
22
That deals with all different aspects of the
47
1
manufacturing process. Comparability,
2
product comparability in terms of
3
recognizing that at different academic
4
centers, islets are being prepared in
5
slightly different methods, in different
6
ways.
7 So how can we show comparability,
8
whether the product really is the same or
9
really is different, based on how it's
10
prepared? So what should be some
of the key
11
criteria, some of the key measures for
12
ensuring comparability?
13 So that could lead to a discussion
14
of various analytical assays, bioassays,
15
preclinical and even clinical studies that
16
would show comparability or would not show
17
comparability.
18 So transitioning into Day 2, the
19
clinical considerations, obviously, approval
20
of this product, of course, is going to be
21 based on data
from domestic or foreign
22
studies that are from well-controlled,
48
1
well-designed studies. They are
going to be
2
performed by qualified investigators.
And,
3
of course, conducted in accordance with
4
ethical principles.
5 So basically, good clinical
6
practices is what we are talking about here.
7 Of course, the
data has to be safe and
8
demonstrate efficacy.
9 So the clinical question is,
10
again, this is just a preview, now we have
11
an islet product and some of the questions
12
to come when you deliver that to the
13
patient, who are the right patients, what
14
are the right measures or outcomes?
15 So this is paraphrasing. So the
16
questions tomorrow will focus on outcome
17
measures, the basic importance and
18
limitations of various outcome measures
19
listed here. For example,
insulin dependent
20
hemoglobin, et cetera.
21 Then the
second question we'd like
22
to have discussed concerns a clinical
49
1
development plan as well as appropriate
2
risk-benefit assessments, things like safety
3
data, the nature and extent of long-term
4
clinical data, historical controls,
5
extrapolating results from a subset of
6
patients, et cetera.
7 All right, so in the last couple
8
of minutes of this talk, I just wanted to
9
remind you of the format.
Basically, it's
10
three parts. Day One consists of
the
11
discussion of manufacturing issues for
12
allogeneic islet transplant.
13 Then late afternoon, we're going
14
to provide an update on research programs.
15
I just wanted to point this out.
This is
16
something that is distinct and not part of
17
the discussion of islet transplantation, but
18
it's an open, public forum. If
you are
19
interested in hearing about it, please stick
20
around. There is a closed
session here.
21 Then tomorrow, of course, we'll be
22
focusing on the clinical issues.
50
1 So the speakers we've lined up:
2 Dr. Burdick from
HRSA is going to following
3
my talk and giving an overview of organ
4
procurement in the U.S., particularly with a
5
focus of the pancreas, of course.
6 Then from the FDA, the general
7
theme of talking about moving from
8
investigational to licensed islet products,
9
what do you need to do from the IND to a
10
license.
11 So Dr. Obiri will be speaking on
12
facilities and GMP issues, and Dr. Wonnacutt
13
about the processing and product quality.
14 Then we're real fortunate to have
15
some experts in the field. Of
course,
16 Dr. Ricordi from
the University of Miami is
17
going to talk about, I think, a historical
18
perspective as well as the current standards
19
for this therapy, for preparation of the
20
product.
21 Of course, Dr. Hering from the
22
University of Minnesota is going to talk
51
1
about characterization and quality
2
standards.
3 Again, the idea here is to provide
4
information that will help provide context
5
for the questions we're asking the committee
6
to discuss.
7 So again, switching to Day Two,
8
Dr. Shapiro is here from the University of
9
Alberta and will give us a talk on clinical
10
islet transplantation and his experience at
11
Edmonton as well as a part of the immune
12
tolerance network, the multi-center study,
13
as well as other studies that he has
14
information on.
15 Then Dr. Burdick has kindly agreed
16
to come back and talk more about allocation
17
issues for pancreas in the sense of how that
18
might impact distribution of islets.
19 Then Dr. Childress from the
20
University of Virginia is going to give us a
21
discussion on ethical considerations for
22
this therapy.
52
1 Then finally, we'll have a
2
presentation from Dr. Dwayne Rieves about
3
the clinical development of islet products
4
from the FDA.
5 I think that's all I had to say.
6 DR. RAO:
Thank you, Dr. Weber.
7
Our next speaker is going to be Dr. Burdick
8
from HRSA.
9 (Pause)
10 DR. BURDICK:
Thanks. That's far
11
and away the best way to get the obligatory
12
PowerPoint delay down to a minimum.
13 Good morning.
Thanks for the
14
invitation. It's nice to be
here. What I'm
15
going to do today is talk about the initial
16
process of having a pancreas available for
17
what we're all talking about for these two
18
days, and that is the regulatory background,
19
and then a bit about what happens in the
20
actual process of retrieval.
21 Tomorrow, as was said, we'll
22 address some of the
more specific issues
53
1
about islets.
2 By way of perspective, I'll try to
3
go through these slides pretty quickly.
4
They are in your packet, but I'll emphasize
5
a few things. I think it's
important to
6
note an important time in the history of
7
transplantation, which is 1984.
I tell
8 people it was the dawning of the age of
9
Aquarius.
10 It was the year that the drug
11
cyclosporine was approved, and it absolutely
12
revolutionized transplantation.
It's hard
13
to describe what an amazing difference that
14
drug made. There have been many
15
improvements in immunosuppression and in
16
control of infection and other things,
17
preservation, et cetera, since.
18 But that was really the time point
19
at which kidney transplantation became
20
absolutely routine, and transplantation of
21
many other organs became relatively
22
feasible.
54
1 The government clearly saw the
2
fact that this was going to require national
3
activity. The National Organ
Transplant Act
4 was passed in
1984. The other major
5
relevant statutory area deals with specific
6
areas of CNS that I'll mention, in the
7
Social Security Act, 1138.
8 There have been amendments in the
9
usual legislative process, but the overall
10
situation hasn't changed much in concept
11
over the past 20 years. It
created a
12
taskforce which reviewed the situation, made
13
recommendations and is no
longer active.
14
That was a transient thing.
15 Then it established that there
16
would be an organ procurement network and a
17
scientific registry of transplant
18
recipients. These were arranged
as
19
contracts to a non-profit bidder for doing
20
the actual work.
21 The way this was put together
22
involved a very strong join between the
55
1
actual clinical process and the allocation
2
of the organs and retrieval of pre- and
3
post-transplant information, so that it has
4 put in force a
situation in which I think
5
arguably there's more complete national
6
reliable information about this little area
7
of medicine than in anything else in
8
medicine.
9 It has warts and blemishes, but
10
it's very important to know how complete and
11
national that information is.
There's also
12
work on public and professional education,
13
and this is the point at which it became
14
illegal to purchase transplantable organs.
15 The organs you see listed here,
16
the usual ones being transplanted.
But it
17
left this open to the Secretary, and we may
18
need to return to that.
19 The nonprofit entity was the
20
United Network for Organ Sharing.
It is the
21
OPTN contractor. It also
established OPO
22
participation, which is important for us
56
1
today, because the organ procurement
2
organizations are the major technical area
3
where the process of retrieving the pancreas
4
is run, and then obviously, the transplant
5
team, surgeons and nurses, do the actual
6
surgical procedure.
7 OPTN has two or three major
8
activities. In the first place,
it's a
9
membership organization. Members
are the
10
institutions that do transplantation.
And
11
the essence of it is that there is a
12
national system with policies that are
13
generated and evolved, because it's a
14
continuing change from time to time in what
15
the policies will be.
16 They run the organ center. They
17
keep a system and a list, and when an organ
18
becomes available, they are the source of
19
the information about where that organ
20
should go and how to arrange it, as well as
21
the source of the rules about how that will
22
be done.
57
1 Then they also work on improving
2
the supply, improving public and private
3
education, and oversee the collection of
4 data which is
analyzed through the
5
scientific registry.
6 The Social Security Act stipulates
7
that hospitals must have written protocols
8
for identification of donors.
The hospital
9
must be part of the OPTN. The
organ
10
procurement organizations in the hospitals
11
are tied together by regs, and obviously,
12
this deals with the reimbursement ultimately
13
to both; in this case the OPOs and also
14
transplant centers, which is under the CMS
15
process.
16 I should apologize perhaps in this
17
setting slightly for this somewhat
18
HRSA-centric slide to some of the people
19
from other agencies. Clearly, if
we were
20
doing it on the basis of the fraction of the
21
total $500 billion that Secretary Thompson
22
oversees, then this block would be here and
58
1
we'd all be down in the corner, and it's not
2
really fair in any way to have these here.
3 But at any rate, from our point of
4
view, so you understand where things sit,
5
we're in HRSA. We're in the
Office of
6
Special Programs, and that's the Division of
7
Transplantation. Our office
oversees the
8 contracts for the
OPTN, which presently is
9
held by UNOS for the data registry, which is
10
held by a group presently in Michigan,
11
URREA. They are formerly the
contractors
12
for the Dialysis and End Stage Renal Disease
13
Registry.
14 Also, we oversee the contract for
15
the National Bone Marrow Registry, in
16
coordination with the Secretary's wonderful
17
recent initiatives on improving public
18
education about the need and value of being
19
in favor of donating organs.
20 The regulation includes a lot of
21
specifics and a lot of delegation to the
22 contractor. This is ultimately,
59
1
fundamentally, and dominantly a community
2
activity, but it is very clearly done with
3
input from and ultimate regulatory authority
4
exercised by the Secretary and the DOT and
5
the HHS in general.
6 Some rather specific things are in
7
the final rule. The
configuration
8 membership
requirements at least are
9
partially specified. How
transplant
10
programs behave once they become members;
11
the necessity for data collection; are all
12
things that we work together from the
13
government and with the community, which
14
essentially is the OPTN contractor by proxy,
15
to optimize the process.
16 You can see the contractors have
17
the responsibilities, including an important
18
website, public information, which is a very
19
valuable place to go for anyone for some of
20
the background information that you might
21
want.
22 Now, the only
procurement
60
1
organization is the technical source of the
2
pancreas in terms of much of the process.
3
They are the ones that are in the hospitals,
4
involved with the patient families, arrange
5
for where and when the retrieval will happen
6
and how it will go. This is
through the
7
organ center arrangements for what organs
8
will be transplanted where.
9 We could go into the process in
10
more detail, but I don't think it's
11
necessary for these purposes.
12 Both the oversight of all of the
13
technical things and the actual logistic
14
arrangements are the Organ Procurement
15
Organization, and these are 60-some
16
nonprofit organizations throughout the
17
country generally representing several
18
transplant centers. And their
activities
19
are closely overseen and specified both
20
within the final rule by implication from
21
NOTA and also by the CMS oversight as the
22
pair.
61
1 The OPOs are looking at all the
2
organs, and I think that's important to keep
3
in mind. This process of having
a pancreas
4
for islets is inextricably tied to the
5
process for having a heart and a liver and
6
kidneys et cetera to be transplantable as
7
well.
8 So a fairly general medical
9
assessment; most important perhaps are the
10
infectious disease things, which work
11
remarkably well. The big issue,
12
particularly with these things, is the
13
timeframe, because as you probably know, a
14
heart needs to be transplanted,
15
revascularized in the recipient within about
16
five hours of cessation of blood supply in
17
the donor.
18 Kidneys,
especially with some
19
aids, can go up to 48 hours, but it's
20
preferable to have a liver in within 8 to 12
21
or 14 hours.
22 The pancreas, there's some
62
1
discussion, but generally a vascularized
2
pancreas graft, the results are a bit better
3
if they can go in within 12 hours or a
4
little more.
5 So you
need to be able to get all
6
of this stuff back quite quickly, and it's
7
done. In transmission of
infectious
8
disease, with some terrible, very prominent
9
disasters notwithstanding, it has been
10
extremely uncommon and the control of that,
11
I think, works very well.
12 Again, given the process, it's
13
probably not as close to 100 percent as it
14
would be if you had months to
deal with each
15
individual organ before it were
16
transplanted. But it does work
very well.
17 This is something again that's
18
national. It's uniform. We have the data
19
across the country, so it's really there for
20
study and evaluation and thought, the
21
processes we are talking about.
22 Obviously, pancreas function is
63
1
checked, although the pancreas transplant
2
surgeon, considering a whole organ, pays
3
little attention to any of these, because in
4
general, there are all sort of reasons they
5
don't correlate very well with what's going
6
to happen in the recipient.
7 Well, there is a shortage, but
8
with pancreas transplantation, it's an even
9 more complex
shortage, because it's not only
10
the number of patients on the list versus
11
the number of pancreases available, but it's
12
the general status of the field.
13 Pancreas transplantation has not,
14
although it is quite successful and it is
15
something that's done on a regular basis,
16
well over 1,000 per year in recent years
17
being done in this country, it's not worked
18
out with quite the same reliability and
19
success as some other organs, for various
20
physiological reasons.
21 So there's less general sense of
22
purpose and mission and need and value for a
64
1
pancreas transplantation, I guess I would
2
summarize, across the country than there is
3
for the kidney or heart or liver.
4 This means that decisions are
5
often made paying less attention to whether
6
the pancreas will be transplanted or not.
7
This is not to say that there isn't a
8
perfectly available and robust process for
9
achieving just the right removal techniques
10
for a process such as a very excellent islet
11
cell transplant treatment that works very
12
well on almost all patients and for which
13
there's a great need for very good
14
pancreases to be removed in just the right
15
way. That process is available
to be
16
facilitated.
17 We are going to talk more about
18
things related to allocation.
There's a lot
19
of research going on with the pancreas right
20
now, because the field in general is still
21
early. So there are some
financial issues
22
that probably we'll just put off until
65
1
tomorrow.
2 But at any rate, one of the issues
3
for islet transplantation right this minute
4
is that there are not as many pancreases
5
retrieved as might be. Again, I
think that
6
this is going to be driven by the results.
7
As the results become clearly better, and
8
it's clear that the OPO will be able to have
9
the pancreas as an organ that gets a
10
reimbursement for the clinical process, as
11
is true for other organs, et cetera, that
12
will drive excellent and more complete
13
retrieval from donors that become available.
14 We're still left with the main
15
problem of the disparity between donors and
16
recipients. About a fifth are actually
used
17
for organ transplantation.
18 Of course, there are other
19
problems. This is one of the
biggest
20
problems. Presently, this is the
21
allocation. It involves two
things. One is
22
first, it will go to a whole organ if there
66
1
is an appropriate recipient identified in
2
the center that feels that that client is
3
appropriate for whole organ transplantation.
4 If that isn't the case, it goes
5
for islet transplantation if possible, and
6
if that doesn't happen, which is quite
7 uncommon now as yet, then it goes perhaps
8
for research or is discarded.
9 That situation, with a large
10
fraction not going to treat a patient in any
11
way, or perhaps even going for research, is
12
one of the reasons for the relatively lower
13
retrieval rate.
14 It's also true that facilitated
15
placement, for one purpose or another, is
16
stipulated in the OPTN policies if initial
17
placement of the organ isn't fairly rapid.
18 It's important to remember that
19
the OPO begins the placement process
20
actually hours before the organ is taken out
21 of the donor, in almost
all cases. In fact,
22
facilitated placement can start if it looks
67
1
like retrieval will be starting within about
2 an hour. That's part of the policy for how
3
to best find a place where this will be an
4
effective donation.
5 I think the procurement process is
6
unlikely to turn out to be a big issue for
7
people, but it certainly is going to be an
8
issue if it changes either the logistics or
9
the cost.
10 I think the criteria are going to
11
have to be studied and developed for exactly
12
what the limits are for an appropriate
13
organ, and I'm going to talk more about
14
what's going on in the OPTN about that right
15
now.
16 Because of its intrinsic
17
relationship to organ transplantation in
18
general, as you can guess, it's very much a
19
major interest in the OPTN, and it's
20
something that the OPTN will continue to
21
play a major role in.
22 The preservation method, I think,
68
1
is interesting. Presently, we
use what's
2
called the University of Wisconsin solution,
3
which is designed particularly for liver
4
preservation, but seems to work well for all
5
the abdominal organs; again, something that
6
will come up as an issue in the specifics of
7
how the glands are taken care of.
8 I think probably I should
9
summarize with a couple of points for
10
tomorrow's discussion, and for further
11
thoughts about how this is going to play out
12
from the point of view of the FDA regulation
13
particularly.
14 One is that the process of getting
15
the pancreas to the point at which the
16
consideration for preparing islets is made,
17
and the oversight of what happens afterwards
18
in the recipient, which is something the
19
OPTN will continue to be part of, because of
20
the involvement with organ transplantation
21
in general, are robust. There's
a long
22
successful history now in the country of
69
1
dealing with this, and it has very clear and
2
active federal oversight.
3 It does, however, involve deeming,
4
if you will, some of that process to the
5
individual medical approaches worked out
6
through the contractor's policies, and
7
that's something that people thinking about
8
the details of regulatory language should
9
understand.
10 Now, the second thing I'll say,
11
and probably more importantly for tomorrow's
12
discussion, just for people to be thinking
13
about, is the concept of product or device,
14
which to some degree gets down to the issue
15
of ownership.
16 I'm sure Jim
Childress is going to
17
give us the real word on this, so I don't
18
want to go too far on this. But
I think
19
it's important to understand that an organ,
20
a kidney or a heart, is in an interesting
21
situation after it's been removed from the
22
donor.
70
1 While it's still in the donor,
2
it's owned by the donor. I don't
think
3
lawyers would have a problem with that.
4
Once it's been revascularized in the
5
recipient, the recipient owns that organ,
6
and again, I don't think there's too much
7
trouble with that.
8 In the meantime, the OPO that
9
packs it up and carries it off, the OPO that
10
receives it to be taken to the hospital
11
where it'll be transplanted, is exercising
12
stewardship. That organ is not
really owned
13
by anybody.
14 Now, if you're talking about
15
products or devices, you're thinking about
16
something that essentially can be owned, and
17
I think that's an issue to be dealt with in
18
this interface question of the islets, which
19
is one of the sort of interesting parts of
20
this.
21 So I leave you with that for
22
today. Thanks for the attention.
71
1 DR. RAO:
Thank you, Dr. Burdick.
2
Any burning questions?
3 MS. MEYERS:
I'm somewhat
4
confused, because when we talk about organ
5
transplantation, the FDA really doesn't
6
regulate organ transplantation, you know.
7
There's no GMPs that a facility has to live
8
up to, et cetera. It's really
run by these
9
contractors to the government.
10 Now here, we're talking about
11
these cells, and we're talking about what
12
appears to be setting up GMPs and the whole
13
manufacturing process, and I'm wondering
14
under what legal authority FDA has to
15
regulate these -- while it looks like it's
16
going to end up regulating these cells
17
whereas it doesn't really regulate the whole
18
pancreas.
19 DR. NOGUCHI:
Abbey, thank you for
20
that question. That is at the
heart of some
21
of the continuing discussions that we have.
22
But I think the way we would look at it is
72
1
that products can be manufactured from a
2
variety of different sources.
3 This is an unusual situation,
4
where we are getting material from one
5
source that has really had a large degree of
6
oversight by another agency. We
are looking
7
at this as, once it has been delivered to a
8
place where it's manufactured, that is where
9
FDA oversight begins for that particular
10
process. We are not directly
addressing the
11
question of ownership. I think
the question
12
that has just been raised is an important
13
one, for which, quite frankly, we don't have
14
a good answer at this time.
15 But part of it is trying to say
16 that we believe that
in some cases, a whole
17
organ may or may not be used for
18
transplantation, but that does not
19
necessarily mean that then it should be not
20
looked at as a source of material that could
21
be further manufactured for a product that
22
may be beneficial to a human recipient.
73
1 I'm not sure if I'm quite getting
2
to the question you're asking, but what we
3
are saying is that we believe that cellular
4
therapies, and this is a part of that,
5
albeit that the islet is a collection of
6
cells, is something that we have
been
7
regulating actually for quite a number of
8
years, as Dr. Weber noted, for just
9
about 15, perhaps 20 years on a very
10
irregular basis, but on a more regular basis
11
since the year 2000.
12 The other question you're asking,
13
though, is a more complicated one.
It's
14
about how does the government really deal
15
with something where different things come
16
into something that eventually is used in a
17
human in a way that we consider at FDA that
18
to be a manufactured product.
19 MS. MEYERS:
What you're saying is
20 that these cells
will be delivered through
21
some manufacturing site that will then
22
process it or do something to it, and that
74
1
manufacturing site will then send the cells
2
out to whatever facility; rather than a
3
whole organ moving from one hospital to
4
another hospital and no manufacturing.
5 DR. NOGUCHI:
Essentially, right.
6
Many of them may not be shipped further than
7
the place of manufacture, and some will.
8
You know, that's still to be developed.
9
Most of the experience has been they are
10
shipped to a place where that facility then
11
for that university or for the hospital will
12
prepare the cellular islet transplant.
13 DR. RAO:
That's a really
14
important topic, and we should hold it
15
because that might be a segue into the
16
discussion as we begin.
17 Our next speaker is going to be
18
Dr. Nicholas Obiri.
19 DR. OBIRI:
Good morning. I'm
20
Nicholas Obiri. I'm with the
division of
21
manufacturing and product quality over at
22
CBER.
75
1 It's my role today to provide an
2
overview of the facilities and good
3
manufacturing practices and expectations for
4
a biologics license application to
5
manufacture allogeneic islets.
6 I'll begin
with a quick overview
7
of the regulatory authority that FDA has to
8
regulate this product. I'll
review the
9
general design principles for a facility
10
that manufactures the product, and I will go
11
over measures that should be in place to
12
maintain control of the facility.
13 Because of the particular
14
relevance of aseptic processing to
15
allogeneic islet manufacture, I'll talk
16
about a few aspects of aseptic processing.
17 FDA's authority to regulate this
18
is product actually is rooted in this act,
19
the Public Health Service Act, Section 351
20
of it, which basically says that FDA shall
21
license biological products when certain
22
conditions are met.
76
1 Another regulation that we need to
2
refer to, which is particularly relevant
3
here, is the Title 21 of the Code of Federal
4
Regulations, Part 601 of its Section 3(d).
5 It basically says that in order to
6
license a product, a biologic, it has to
7
have these attributes in terms of quality,
8
but in addition to that, it has to be
9
manufactured in a facility that meets
10
certain standards. Now the standards are
11
defined in these regulations here.
12 Just to illustrate what I'm trying
13
to say, if we say that the qualities that a
14
product should have would make it acceptable
15
to FDA, or we should just refer to it as a
16
quality product, then in order to get a
17
biologics license to make allogeneic islets
18
or any other biologic, you'd have to
19 demonstrate ability
to manufacture a quality
20
product; i.e., a product that meets those
21
attributes I showed in the previous slide.
22 Then, not only that, but a
77
1
facility in which you made that product has
2
to meet such design standards.
In addition
3
to the design standards for the facility,
4
you also have to have the manufacturing
5
operations reflect these attributes.
In
6
other words, the operations within the
7
facility should have these characteristics.
8
I will return to this slide at a later
9
point.
10 But for right now, I do want to
11
start with a discussion of the facility. In
12
other words, we are talking about a facility
13
that would be a compliant facility.
It
14 would be one that
meets all of the
15
requirements that are defined by the
16
regulations I referred to earlier, and these
17
regulations are otherwise known as the good
18
manufacturing practices, or GMPs.
19 Such a facility should be
20
appropriately designed. It
should have a
21
controlled environment within the facility
22
where the operations are occurring.
It
78
1
should have provisions for using only
2
equipment that's properly qualified, and
3
there should be adequate measures to control
4
cross-contamination, and there should also
5
be adequate measures to ensure that there is
6
no mix-up of patient material.
7 There should be provision for
8
controlling incoming raw materials, and
9 there should also
be an independent quality
10
assurance or quality control staff.
11 There should also be appropriate
12
provision for keeping up with the records
13
and making sure that all documentation is up
14
to date.
15 So just to spend a little more
16
time on the design attributes of the
17
facility, the design should be influenced by
18
the nature of the source material.
For
19
example, if we were starting with a solid
20
material like a pancreata, for example, then
21
one would expect that a facility would be
22
designed in such a way that there's
79
1
appropriate receiving area to receive the
2
source material, and also to do all the
3
processing and documentation that would be
4 required.
5 Now if, on the other hand, we were
6
talking about static materials that would be
7
a vial of frozen cells, then we probably
8
wouldn't need to have an elaborate design
9 for a receiving
area.
10 The design should also be
11
influenced by what the purpose of the
12
facility is. Is it going to be a
single
13
product facility or is it going to be a
14 multi-product
facility?
15 Critical manufacturing areas
16
should be designed into the facility.
For
17
this kind of product, one would anticipate
18
that aseptic processing would occur.
19
Therefore, the facility should be made of
20
materials that will be appropriate for that
21
kind of processing. Just as an
example, the
22
interior surface of the material should be
80
1
made up of materials that are smooth and
2
solid, being able to resist the cleaning
3
agents that would necessarily have to be
4
used to maintain a high level of sanitation
5
within the facility.
6 For this class of products in
7
particular, I think it's well-agreed that
8
the processing should be well-defined and
9
well-characterized. So the
manufacturing
10
process, would need to be such that it is
11
built into the facility so that the design
12
of the facility should provide for a proper
13
flow of personnel and the process.
14 Just as an example, one can
15
anticipate that the initial processing of
16
the material would occur in a particular
17
part of the facility. While the
design
18
should provide for a situation in which
19
there is a defined location for that kind of
20
processing; not only that, but any
21
subsequent processing of the materials that
22
are likely to be more refined should occur
81
1
in a segregated area from the more crude
2
initial processing.
3 This scheme would allow for a
4
situation in which not only
does the
5
manufacturing process flow from the upstream
6
manufacturing areas to the downstream
7
manufacturing areas, but you would also have
8
a natural flow of personnel.
This
9
arrangement would ensure that there is very
10
minimal chance for cross-contamination.
It
11
would also make it very unlikely that you
12
would have product mix-up.
13 So I want to
switch gears just a
14
little bit. I've talked
primarily up to
15
this point about the facility, the physical
16
facility. As I indicated at the
beginning,
17
especially when we are looking at that
18
regulation, we not only have to have control
19
of the physical facility; we also have to
20
have control of the environment within the
21
facility which governs the manufacturing
22
operations.
82
1 So I'm going to go on now and talk
2
about the measures that should be taken to
3
ensure that we maintain environmental
4
control within the facility.
5 The single most important system
6
that would ensure control of the
7
environmental conditions within the facility
8
would be the heating, ventilation and the
9
air conditioning system, commonly referred
10
to as HVAC. That system is
critical,
11
because it has to be able to provide a
12
HEPA-filtered air in the manufacturing
13 areas. By that I mean high efficiency
14
particulate filtered air.
Because the
15
environment in the facility has to be of the
16
cleanest standard, has to meet the highest
17
standard of cleanliness with regard to the
18
air quality, as much as possible.
19 It should provide then for the
20
ability to control the air supply to the
21
area, and also be able to create conditions
22 or areas within
the facility where critical
83
1
operations can occur.
2 It should provide for the ability
3
to use a pressure cascade to protect the
4
product, and this should be possible by
5
allowing very critical operations, such as
6
situations in which we have to open the
7
exposed product to the environment; that
8
kind of operation should be performed in an
9
area of high pressure surrounded by an area
10
of low pressure.
11 The HVAC system should be able to
12
provide for the ability to use a pressure
13
sink to protect all the manufacturing areas
14
and personnel.
15 Because of its importance, we
16
should have a very well-defined process or
17
procedure or program to qualify the HVAC
18
system, because we need to be able to
19
confirm that the equipment itself, that's
20
the hardware of the HVAC, its control and
21
the circulation system, we have to be able
22 to ensure that
they meet expected
84
1
performance quality.
2 This is usually done by monitoring
3
the environment. Again, remember
that the
4
environment is provided by the HVAC.
So in
5
order to do the qualification of the HVAC,
6
we would have a program of environmental
7
monitoring where we monitor conditions
8 within the
facility under non-operational
9
conditions as well as under operational
10
conditions.
11 So in addition to having a clean
12
environment within the facility, we also
13 want to ensure
that all of the materials,
14
the reagents that come into the facility,
15
meet a minimum standard. For
example, we
16
would expect that pharmaceutical-grade
17
reagents and supplies, such as water,
18
processed air, and utility gases would meet
19
these minimal standards.
20 Again, the manufacturing process
21
has to be validated. This
validation would
22
be based on data. Actually, data that is
85
1
gathered by the manufacturer. We
would
2
expect that there would be demonstration of
3
the ability of the manufacturer to make this
4
product on a consistent basis.
5 Because of the nature of this
6
particular class of product, we would expect
7
the manufacturer to demonstrate ability to
8
carry out aseptic processing.
9 So the validation of the process
10
would encompass not only the ability to
11
manufacture the product on a consistent
12
basis, but it would also be expected that
13
qualified equipment would be used in that
14
manufacture. At the same time,
we would
15
expect the manufacturer to demonstrate an
16
ability to maintain control over other
17 facility systems while
making the
18
consistency lots.
19 In order to achieve this, of
20
course, you would also expect that all of
21
the staff that would be used in this process
22
would be properly trained and qualified.
86
1 So I've talked about several
2
different things that should be in place in
3
order to make a quality product.
One might
4
wonder how does one keep track of all of
5
these things. That's the reason
there
6
should be a quality system. A
quality
7
system is a system that should be in place
8
in the facility that should have these
9
attributes.
10 There should be provision for
11
vendor audit, and this would be the
12
suppliers of the reagents and the materials
13
that are used for manufacture.
The vendors
14
should be qualified with regard to integrity
15
as well as with regard to the quality of the
16
materials that they supply.
17
There should be
provision for
18
material qualification. All
materials that
19
are going to be used for manufacture should
20
be properly qualified. There
should be
21
provision for an oversight of the process,
22
and there should be provision for a change
87
1
control. By this I mean a
well-thought-out
2
procedure that would be used for making
3
changes in the manufacturing process or with
4
regard to equipment after the initial
5
qualification or validation has occurred.
6 I should also mention that after
7
licensure, FDA does not expect changes in
8
the manufacturing process or some of the
9
critical processes. But when
those changes
10
are necessary, there is provision or
11
requirements for how the agency should be
12
notified about those changes.
13 There should also be provision for
14
personnel training. I think it's
obvious
15
that people that are going to be involved in
16
the manufacture of this kind of product
17
would need to be properly trained and be
18
well-versed in the use of complex equipment.
19 However, what should not be
20
overlooked is the need for these personnel
21
to have GMP training also. So
usually one
22
would expect that there would be initial
88
1
training of all staff on GMP issues, but
2
also a regular updating of that training.
3
It's also very important that that training
4
be documented.
5 There should be provision within a
6
quality system for investigation of
7
deviations, recalls, product complaints and
8
the Med Watch Program.
9 So I said I would return to this
10
slide and I think from the previous things
11
that I have said, you can see what I mean by
12
saying that to make a quality product, which
13
would meet all of the attributes that are
14
required by the law, then it is FDA's
15
expectation that these features would be
16
obvious in the manufacturing facility.
17 So we would need a facility that
18
meets certain minimal design standards, but
19
in addition to that, that the manufacturing
20
operations within the facility also manifest
21
these features.
22 We would expect them to use
89
1
qualified equipment. We would
expect that
2
the manufacturing process be validated.
We
3
would expect that the components and the raw
4
materials would be qualified, and we would
5
expect that the environment within the
6 facility would be
under control.
7 We would also expect that there
8
would be a quality unit that assures that
9
all of these standards are met.
In some
10
places, it would be a quality assurance
11
unit. In some places, it would
be a quality
12
control unit. In many places,
you'd have
13
both of them; you'd have both a quality
14
control unit as well as a quality assurance
15 unit.
16 One might look at it as that the
17
quality control unit would carry out all the
18
tests that need to be performed on the
19
product, as well as on the intermediate
20 reagents.
21 Whereas, the quality assurance
22
unit would be the unit that uses the results
90
1
generated by the QC unit to reject or accept
2
the product.
3 So with this setup in place in a
4
well-designed facility, we would expect to
5
be able to make a quality product.
6 I did say that I would say a few
7
words about aseptic processing.
I am
8
putting up this definition, which I think
9
talks about what I have in mind here.
It's
10
a processing approach in which a product
11
manufacturer goes under environmental and
12
processing conditions that assures minimal
13
opportunity for contamination from the
14
environment or personnel.
15 Because of its nature, terminal
16
sterilization would not be a physical option
17
for allogeneic islets.
Therefore, the final
18
product has to be assembled by introducing
19
the aseptically-produced or processed final
20
formulation of islet cells into a sterilized
21
container and then filled with a sterilized
22
closure system in a high quality
91
1
environment.
2 In order to do that, it would be
3
necessary for all open manipulations and
4
connections that have to be made, they have
5
to be made under aseptic conditions.
So
6
aseptic processing would involve trained
7
personnel and qualified personnel.
It must
8
be validated.
9 Typically, aseptic processing
10
would be validated through media challenges.
11 Basically what this
means is that we would
12
simulate the entire manufacturing process,
13
except that we would substitute media for
14
the product, and then we would incubate that
15
media, and hopefully there will be no
16
microbial growth after a period of
17
incubation.
18 So aseptic processing typically
19
occurs in a Class 100 environment.
A
20
Class 100 environment is just the highest
21
quality environment that one would expect to
22
see in a manufacturing facility.
The
92
1
conditions in a vial safety cabinet fulfill
2
that requirement, but when operations are
3
being performed in this kind of environment
4
there should be appropriate environmental
5
monitoring.
6 For example, viable and non-viable
7
airborne particulates should be monitored.
8
Aseptic processing may also occur in a
9
closed system. A manufacturer
may define a
10
system as closed, but we would expect that
11
such a claim would be supported by
12
validation data.
13 An example of a closed system
14
would be, for example, a system of
15
fermenters or a system of bags that are
16
aseptically put together, for example, using
17
sterile connecting devices. So
as I said
18
again, it's very important that there should
19
be data that supports the claim of a closed
20
system.
21 So I think I can
leave you with
22
this point as my take-home message.
We are
93
1
saying that we should design compliance into
2 the facility
plans. It's advisable to seek
3
CBER input prior to construction.
I also
4
think that might be cost-effective in some
5
instances.
6 We should establish a thorough
7
qualification or validation program, and we
8
should maintain an effective quality
9
assurance or quality control unit to assure
10
maintenance of quality standards and
11
regulatory compliance.
12 We should maintain an aggressive
13
approach to compliance with aseptic
14
processing requirements.
15 I just want to leave this up as an
16
additional resource that may be helpful.
17
The division of manufacturing and product
18
quality would welcome an opportunity to
19
answer questions that manufacturers may
20
have.
21 We would entertain a request for
22
Type C meetings where we would discuss
94
1
facility issues. It's usually
helpful if
2
the manufacturer has specific questions that
3
they want to ask.
4 In the preparation for obtaining a
5
biologics license, one of the things that
6
have to be done would be a pre-operation
7
inspection. We would also be
very happy to
8
discuss details of those inspections with
9
you.
10 I'd like to acknowledge my
11
colleagues at the Division of Manufacturing
12
Quality, as well as John Eltermann, the
13
director, Dr. Finkbohner, the deputy
14
director, for useful discussions and
15
contributions that they made to this
16
presentation.
17 Thank you very much.
18 DR.
RAO: Thank you, Dr. Obiri.
19
Before we go on to the next speaker, I'd
20
like to take this opportunity to welcome two
21
additions to the committee. I'd
like to ask
22
them to just briefly introduce themselves,
95
1
Dr. Eggerman and Dr. Mulligan.
2 DR. EGGERMAN:
I'm Tom Eggerman.
3
I'm a program director for Islet
4
Transplantation in the Diabetes Institute,
5
and I'm very glad to be here.
Thank you.
6 DR. MULLIGAN:
I'm Richard
7
Mulligan from Harvard Medical School.
I'm a
8
member of the BRMAC and I'm a researcher in
9
the area of stem cells and gene transfer.
10 DR. RAO: Our
next speaker is
11
Dr. Wonnacutt, who will sort of carry on on
12
the next aspect of looking at quality
13
control and looking at some of the product
14
issues.
15 DR. WONNACUTT:
Thank you,
16
Dr. Rao. My name is Keith
Wonnacutt, and
17
I'm in the Office of Cellular Tissue and
18
Gene Therapies, and I'll be talking about
19
processing and product quality issues.
20 As an overview for what I'm going
21
to be talking about, first I'll talk a
22
little bit just about the FDA regulation of
96
1
islets, and then go into specific issues
2
related to islet product quality, how source
3
materials relate to that manufacturing
4
process, and product testing all contribute
5
to product quality, and the questions that
6
the FDA has surrounding these areas.
And
7
then conclude with issues related to islet
8
comparability.
9 So the first part, FDA regulation
10
of islets. As Darin mentioned in
his talk,
11
in September of 2000, the FDA issued a Dear
12
Colleague letter which stated, "The purpose
13
of this letter is to inform or remind you of
14
how the Food and Drug Administration
15
regulates allogeneic pancreatic islets for
16
transplantation. These cellular
therapies
17
are regulated as biological products subject
18
to licensing under Section 351 of the Public
19
Health Service Act."
20 So what is actually licensed? The
21
FDA licenses products. In this
case, what
22
would be licensed is the final islet
97
1
cellular product. The
manufacturing process
2
is not licensed; however, a licensed product
3
is dependent upon a specific manufacturing
4
process.
5 In the absence of extensive
6
product characterization and manufacturing
7
process, the manufacturing process helps to
8
define the product. So in order
to obtain a
9
license for a biological product, the
10
sponsor has to submit a biologics license
11
application.
12 The data that's needed to support
13
this application is prescribed in the regs,
14
which state "the manufacturing shall submit
15
data derived from non-clinical laboratory
16
and clinical studies which demonstrate that
17
the manufactured product meets prescribed
18
requirements of safety, purity and potency."
19 So how do we ensure safety, purity
20
and potency? The way we do that
is by
21
applying quality standards to the products.
22
During pre-clinical development, before we
98
1
start in humans, although not necessarily
2
required, good laboratory practices or GLPs
3
contribute to product quality.
4 During investigational stages in
5
the clinic of drug development, GMPs are
6
required, but they are applied in a way that
7
allows for development of the product early
8
and control of the product late in
9
development. Full compliance
with good
10
manufacturing practices are required for a
11
licensed product.
12 During this process development,
13
we expect that the characterization of the
14
product will be constantly improving,
15
although we expect that a threshold of
16
product characterization would
be met prior
17
to beginning pivotal studies so that the
18
manufacturer understands and can control
19
what is being given to the patients.
20 So specifically, how do we control
21
product quality, and what goes into current
22
good manufacturing practices? In
the
99
1
squares here, I've just listed some of the
2
major points involved in the current good
3
manufacturing practices, as outlined in the
4
regulations. They would include
things such
5
as organization and personnel, buildings and
6 facilities,
packing and labeling, control of
7
components, manufacturing controls,
8
laboratory controls, records and reports,
9
holding and distribution.
10 All of these things ensure the
11
safety, purity and potency of the product.
12 Now, LaVelle Edwards, who was the
13
BYU football coach at my alma mater, used to
14
say practice doesn't make perfect, but
15
perfect practice makes
perfect. When we
16
talk about controls, I would apply this to
17
that and say controls don't make quality
18
products, but quality controls make quality
19
products.
20 So my talk will focus on what are
21
the aspects of quality control that will
22
lead to a quality product, especially in
100
1
terms of the source material, the
2
manufacturing, and the product testing.
3 So this leads me into the second
4
major section of my talk. The
control of
5
the source material relates directly to the
6
first CMC question that we're proposing to
7
the panel today. It deals with
the quality
8
control of the source material, or the
9
cadaveric organs.
10 Source material control is
11
difficult in the case of islets because the
12
islet source material is variable.
The
13
source material for islets are cadaveric
14
organs and cannot be controlled in a
15
traditional way, because each
organ is
16
unique in terms of organ size, donor age,
17
extent of fibrosis and autolysis.
18 Also, organ procurement procedures
19
may vary. This was part of what
Dr. Burdick
20
was talking about: Ischemia
time, transport
21
media, organ core temperature, these are all
22
things that may vary with the organ, and
101
1
it's not within the FDA's purview to control
2
some of those things.
3 So a key component for ensuring
4
the control of a validated islet
5
manufacturing process is the use of
6
predefined acceptance criteria for the
7
source material. The acceptance
criteria
8
should ensure that only suitable donor
9
organs, or organs with maximal potential for
10
yielding adequate numbers of
islets are used
11
for islet manufacturing, while unsuitable
12
organs are excluded from further
13
manufacture.
14 Of course, as Dr. Burdick points
15
out, we have to balance this
with the idea
16
that we want to be as inclusive as possible
17
because of the shortage of the organs.
So
18
what could go into the acceptance criteria
19
for the source material or the donor organs?
20
Donor suitability determination, such as
21
viral testing, organ characteristics,
22
harvesting conditions and transport
102
1
conditions, are all things that may
2
contribute to source material control.
3 So our first question is what
4
would be appropriate for a license
5
application? Please discuss the
data needed
6
for developing predefined acceptance
7
criteria for source organs.
8 The second area I'd like to talk
9
about is manufacturing controls.
10
Specifically, the quality control of the
11
manufacturing process. First,
our
12
expectations. In order to
produce a product
13
that is consistent in safety, purity and
14
potency, the manufacturing process should be
15
standardized and validated.
16 In-process testing should confirm
17
the consistency of this process, and for
18
licensed products, the process by which they
19
are made should not be experimental and
20
should be shown to produce a safe and
21
effective product.
22 I'd just like to point out that
103
1 experimental procedures
result in
2
experimental products as far as the FDA is
3
concerned.
4 So what about manufacturing
5
changes? We know that
manufacturing changes
6 can impact product
safety, identity, purity,
7
potency, consistency and stability in
8
unforeseen ways. Therefore, the
product
9
used in pivotal trials should be
10
representative of the product that is
11
intended to be licensed.
12 In the case of allogeneic islets,
13
there is a lot of change, or differences, in
14
the processing. Investigators
frequently
15
customize an islet isolation procedure based
16
on a given donor organ's characteristics to
17
optimize the yield of islets.
18 There are many variations in islet
19
isolation methods, both within centers or a
20
single manufacturer and across centers.
21
Examples of manufacturing variations include
22
digestion time and temperature; the use of
104
1
additives such as DNase and protease
2
inhibitors; issues with critical digestive
3
enzyme, which is usually liberase; and
4
culturing islets prior to transplantation.
5 All of these things can be varied
6
and do vary from manufacturer to
7
manufacturer. Also, all of these
are many
8
times used to optimize the yield of the
9
islets. So here at the FDA, we
agree that
10
some flexibility in the manufacturing
11
process is acceptable, if it's conducted
12
using predefined criteria or algorithms
13
within a validated manufacturing protocol.
14 These predefined criteria would
15
establish conditions that would allow for
16
processing variations based on the
17
characteristics of each donor organ.
18 So in terms again of a licensing
19 application or for
a license, is it
20
reasonable to expect that criteria or
21
algorithms can be developed based on data
22
collected during IND studies to predetermine
105
1
under what conditions the use of a specific
2
reagent, reagent concentration or processing
3
method is appropriate?
4 The third CMC question that we
5 have deal with the quality
control of the
6
final product, or product testing.
First,
7
for expectations, release testing of the
8
final product should be performed on a
9
sample of the final product. The
test
10
methods are prescribed by regulation.
Some
11
test methods are prescribed by regulation
12
and some are proposed by the applicant.
13 Each test result should contribute
14
in a meaningful way, or contribute
15
meaningful information about the safety,
16
purity and potency of the product.
Now,
17
this is just a few of the prescribed methods
18
that are required, and these are biologic
19
product standards that are outlined in the
20
regulations in 21 C.F.R. 610, and they
21
include sterility and purity, identity and
22
potency.
106
1 You can see that sterility and
2
purity are specified; the test is specified.
3
Identity and potency are not specified and
4
proposed by the sponsor, and the
5
specifications that are set are also
6
proposed by the sponsor and should ensure
7
product quality.
8 So specifically, I want to focus
9
on potency because it has important
10
implications for islets. First,
to define
11
potency, which as defined in the
12
regulations, it says "the word potency is
13
interpreted to mean the specific ability or
14
capacity of the product, as indicated by
15
appropriate laboratory tests, or by
16
adequately controlled clinical data obtained
17
through the administration of the product in
18
the manner intended, to affect the given
19
result."
20 So what are some of the challenges
21
for potency assays in islet manufacturing?
22
First, the results of the potency assay
107
1
should be available before the product is
2
released. The results should
show the
3
ability to affect a given result.
4 Potency, ideally, although not
5
explicitly, should correlate with biological
6
activities that provide the intended
7
therapeutic effect of the product in vivo.
8
You want to correlate what you are testing
9
to what is happening in the patient in the
10
ideal world.
11 So examples of current potency
12
assays used for islets:
Glucose-stimulated
13
insulin release or injection of islets under
14
the kidney capsule of diabetic mice to
15
restore proper control of blood sugar in
16
those mice.
17 However, in both of these cases,
18
these results are not available prior to the
19 release of the
product. I should say that
20
other assays are under development.
I think
21
Dr. Hering will talk about that.
22 So an acceptable lot release
108
1
potency assay is required for a biologic
2
license. If bioassays aren't
feasible for
3
lot release, the applicant may provide
4
rationale for other approaches to ensure
5
product potency. An example may
be
6
viability plus other characteristics such as
7
dithizone staining, and correlating those
8
measures with the clinical data.
9 So our question is, "please
10
discuss any assay or assays that are
11
currently or could be performed on the final
12
islet product before patient administration
13
which may be predictive of the ability of
14
islets to perform as expected after patient
15
administration."
16 So the last area that I'd like to
17
cover is comparability.
Comparability is
18
demonstrating that critical product
19
characteristics, including safety, purity,
20
and potency, have not changed even when the
21
manufacturing process has changed.
22 So products manufactured with
109
1
different processes, in the FDA's eyes, are
2
considered to be different products until
3
comparability is demonstrated.
How can you
4
demonstrate comparability?
Testing to
5
demonstrate comparability may include
6
analytical assays, bioassays, pre-clinical
7
studies, or it may require clinical studies.
8 Why does this matter? In general,
9
each license is for one product produced by
10
one manufacturing process. It is
unclear
11
how difference in methods to prepare
12
allogeneic islets by various groups impact
13
the characteristics of the
final allogeneic
14
islet product.
15 Data from different manufacturing
16
facilities, or even the same facility using
17
different processes, cannot be used to
18 support the same
license application unless
19
comparability is demonstrated.
20 So our final question in terms of
21
manufacturing is "what should be the key
22
criteria or measures for demonstrating
110
1
allogeneic islet product comparability?
2
Please discuss appropriate analytic assays,
3
bioassays, pre-clinical studies and clinical
4
studies that may be required, whatever you
5
deem to be appropriate."
6 So in summary, for licensure, a
7
well-controlled, validated manufacturing
8
process is needed to assure the safety,
9
purity and potency of the final product.
10
This will require control of all starting
11
materials used, control and consistency of
12
the manufacturing process, and testing of
13
the final product to verify it meets
14
predefined product safety and quality
15
standards.
16 I'd just like to reemphasize what
17
Dr. Goodman had said, that although we are
18
specifically talking about islet cells
19
today, many of these issues expand and can
20
be applied to many of the other products
21
that we deal with in the Office of Cellular
22
Tissue and Gene Therapies.
111
1 That's all.
Thank you very much.
2 DR. RAO:
Thank you. We have a
3
short question-and-answer session
4
specifically related to topics that they
5
have raised rather than the questions that
6
the FDA wants us to address.
7 MS. MEYERS:
I hate to go back to
8
the same issue on the manufacturing.
Do
9
they have the same requirements in bone
10
marrow transplant labs? I mean,
the bone
11
marrow has to be processed. It
has to be
12
grown, and yet FDA doesn't regulate that.
13 DR. NOGUCHI:
Well, actually we
14
do, if in fact they do the things such as
15
been demonstrated here where they are
16
isolated and then purified, manufactured.
17
They may be grown. Genes may be
added to
18
them. In fact, we have been
regulating
19
cellular therapies of that nature.
20 It is true; there are certain
21
aspects of bone marrow transplantation,
22 which is
regulated currently and directly by
112
1
HRSA, which we don't oversee. We
have tried
2
over time to really distinguish between an
3 area of
clinical medicine where we feel our
4
oversight does not add value; that is in
5
standard bone marrow transplantation, to
6
that area which is a zone that we don't know
7
how to define precisely.
8 But clearly at some point in time,
9
when you are manipulating cells, there comes
10
a point, which is still being debated, at
11
which we no longer are assured that they are
12
safe, let alone that they are effective.
13
Safety and effectiveness is a FDA regulatory
14
oversight responsibility. That
is what we
15
are talking about.
16 As an example, if we were talking
17
about a pancreas or a pancreas is cut in
18
half in an attempt to try to use it for more
19
than one patient, we would not consider that
20
to be manufacturing. That would
raise the
21
same level of safety and effectiveness that
22
FDA traditionally sees.
113
1 For these types of products here,
2
we do feel that, as an example, how the
3
cells are liberated, how the islets are
4
liberated from the whole pancreas; they are
5
done by mechanical and enzymatic methods.
6
They can be subject to additional concerns
7
about cross-contamination during that
8
manufacture.
9 As you isolate them, do they have
10
the same functional characteristics that a
11
normal pancreas would have, more and more
12
manufacture that takes place -- and this is
13
a debatable item -- is where we want to
14
exert more oversight to make sure that the
15
end user, which is ultimately all of our
16 goal, receives as
best a product as they
17
can.
18 In these cases for islets, we
19
consider that question of safety and
20
effectiveness one to be answered under IND.
21 For many
areas of bone marrow
22
transplantation, even there, even the term
114
1
bone marrow transplantation is undergoing an
2
evolution where one may say I received a
3
bone marrow transplantation and not realize
4
that actually, they already received an
5
FDA-regulated product or GCSF for
6
mobilization of your bone marrow, which is
7
then removed from your arm and then may be
8
processed further and given back to you. So
9
it is quite a bit of a moving target.
10 But our concern is really to say
11
we want to exert oversight where we have
12
questions of safety and effectiveness.
13 DR. RAO:
Maybe I can follow up on
14
Abbey's question, because it's, as you
15
pointed out, quite correct that certain
16
things which are directly harvested from the
17
tissue are under HRSA's control, but even if
18
you take bone marrow as an example, you
19
don't regulate the fact that if you take
20 bone marrow itself and you enrich for stem
21
cells and you do it over a short period of
22
time, and you don't grow the stem cells in
115
1
culture, it's not regulated as a product,
2
right?
3 So if you take pancreatic islets,
4
for example, and you disassociate them and
5
you were to transplant them in a short
6
period of time in an identified donor, would
7
that be considered under the purview of the
8
FDA, or would there have to be additional
9
manipulation before it would be considered
10
under the purview of the FDA?
11 DR. NOGUCHI:
There are several
12
questions there and I don't want to spend
13
too much time directly on this issue, as
14
there are many other questions.
But
15
specifically, early days of islet
16
transplantation were very much simply taking
17
a pancreas and disassociating them and
18
attempting to see whether or not they were
19
safe and/or effective.
20 Those have already been under
21
regulation for, I think actually the first
22
one was well before 1993. So
yes, we would.
116
1
Part of the discussion
that also
2
has to be brought in is what is the
3
indicated use. Let's go to bone
marrow. As
4
an example, if you take bone marrow out, put
5
it back into the individual to be used as
6
bone marrow, that is certainly something
7
that has been widely practiced and has a
8
long history of clinical acceptance.
9 If you take bone marrow out and
10 you inject it into the arm and you are
11
attempting to grow muscle, for example, that
12
certainly doesn't really make a whole lot of
13
sense from conventional wisdom, and we would
14
have concerns about safety and
15
effectiveness.
16 Likewise, if you took bone marrow
17
out with minimal manipulation and injected
18
it into the heart in an attempt to grow new
19
muscle, or in an attempt to revascularize
20
the heart, those are questions that simply
21
at this point in time have not been
22
demonstrated scientifically, medically,
117
1
clinically at all, and raise questions of
2
safety and effectiveness.
3 So part of the problem that we
4
need to really consider in the overall
5
scheme of things is it's not only what is
6
the product and the degree of manipulation
7
manufacture, it is what is the intended use
8
and where is it going to be used.
9 All those things together, in a
10
perhaps more dramatic example, is if we have
11
something that is a wound covering, like a
12
piece of gauze or an artificial membrane
13
that's used on your hand, that's one thing.
14
The same cleared FDA-approved product, if
15
you used it as an aortic patch for an
16
aneurysm, clearly in that indication does
17
not work, would not work.
18 No one here would agree at all
19 that just because
it's an FDA-cleared
20
product, it can be used medically as an
21
aortic patch. We are facing
similar
22
conditions here, where you have seemingly
118
1
minor differences between one process and
2
another, and at some point, it becomes to a
3
state where no longer safety and
4
effectiveness can be assumed. It
is very
5
dependent on the intended use and the
6
clinical use.
7 So again, these are obviously very
8
wide-ranging studies or conditions and
9
questions, but in the interim, I think we
10
would like to make sure that we acknowledge
11
that it's a very broad and complicated
12
question, but also want to make sure if
13
there are any specific questions here for
14
the presentations, we can address those.
15 But I mean, this is not to
16
discount Abbey's question at all.
17 DR. RAO: Dr.
Burdick, would you
18
like to comment?
19 DR. BURDICK:
Just to maybe go a
20
little bit more into this, I think part of
21
the mischief here is the concern about the
22
use of the term "product," because the
119
1
community, I think, has a tendency to think
2
about islets much more like an organ.
3 Now, in the case of a kidney, it
4
can be put on a pulsatile perfusion machine,
5
as is often done for deceased donor kidneys,
6
for a period of time, and all of the
7
machinery and the solutions and everything
8
are FDA-approved products, but what comes
9
out, that kidney that gets transplanted, is
10
not a product. No argument.
11 A piece of skin graft from a
12
deceased donor can be processed and
13
freeze-dried and then sold by the company,
14 or passed on for
the appropriate fee by the
15
company that does that. That
seems to fall
16
well into the definition of a product.
17 But whether what is being looked
18
at, whether the regulations that are
19
proposed, what they are dealing with is a
20
product or not is presumed by the regs
21
specified in 2000. We may be
beyond that.
22
But I think we have to recognize that that's
120
1
where the problem is.
2 The biggest single concern I think
3
that I have about that is what happens once
4
it's ready to go. We'll talk
about
5
allocation tomorrow.
6 But it comes back to what the
7
community really thinks about it.
I think
8
it does cause some problems for the FDA's
9
viewing this as a product that
are going to
10
have to be addressed.
11 DR. RAO: I
have one more
12
question, if nobody else does.
It's this
13
issue that you raised about ownership.
I
14 think the minute
you define something as a
15
product, the issue of ownership then becomes
16
really important.
17 What is the consensus between,
18
say, HRSA and the FDA? You have
an organ
19
which belongs to a donor, and you have a
20
recipient where you put this in.
You have
21
taken the cells through a procurement
22
organization. You have
disassociated the
121
1
cells and you've maybe processed them.
Do
2
we have to worry about who that product
3
belongs to before any decision can be made?
4
And who takes responsibility for it now that
5
it's a product?
6 DR. NOGUCHI:
I think what FDA
7
would say is we think the question of
8
ownership is a very important issue, but we
9
do not think that has to be addressed as
10
we're moving forward in this area.
11 I think we have several
12
alternatives; one of which is that there is
13
documented data that some patients with
14
severe diabetes are benefiting both by whole
15
organ transplant as well as by islet cell
16
transplant.
17 Should we be in a position of
18
trying to choose between the two?
That's
19
certainly not our intent at FDA, not the
20
intent at HRSA, and I am quite sure, nor
21
internationally.
22 We are striving to find the best
122
1
balance of approaches that will yield the
2
most benefit for the most people.
Ownership
3
per se, as far as FDA is concerned for a
4
product, does not necessarily enter into
5
questions about whether or not quality
6
should be built in as an example or that
7
yes, we need to know that things are safe
8
and effective.
9 In terms of the actual
10
distribution of that, in a traditional
11
setting, that is a more clear-cut case.
If
12
you are manufacturing a drug, the person
13
that develops that, there are questions of
14
ownership and intellectual property that are
15
quite straightforward.
16 In this arena, where much of what
17
we are talking about goes far beyond the
18
traditional pharmaceutical model and it's
19 talking about
distribution of -- let's call
20
them entities -- that can be useful to
21
patients, many of which are done not in a
22
profit-making mode, but in a nonprofit mode
123
1
or even in a distributive mode that is far
2
beyond what we traditionally have for a
3
pharmaceutical model, we think that the
4
principles that are being used to
5
manufacture a quality product by a
6
pharmaceutical, not all of which can be
7
translated to this situation; however
8
certainly, the principles can be translated
9
to make sure that every time you make a
10
product or that you make an islet
11
preparation, let's say, that it's the best
12
quality that can be used and it offers the
13
patient the best opportunity for
14 amelioration of
the disease.
15 So I agree with everyone that it
16
continues to be, these questions such as
17
ownership, such as distributive oversight
18
between two or more agencies here, including
19
the public sector, are very complicated, but
20
it should not inhibit us from really trying
21
to move forward and get the best therapy out
22
for our patient, which is our joint
124
1
constituency.
2 MS. LAWTON:
Could I just add a
3
comment now? I think when we
think about
4
the ownership issue, to me, the key issue
5
that we'll have to discuss is about the
6
control of the source material.
That's
7
going to be a key component that we have to
8
understand what's important in the source
9 material.
10 Ultimately, we may need to sort
11
out who owns the material and how do we
12
control for what happens to that source
13
material. So it may end up
coming back to
14 being an important
point as far as how the
15
organs are procured and stored, et cetera,
16
as part of a key component of the source
17
material itself.
18 DR. CHILDRESS:
The ownership
19
issue is certainly an important one,
20
philosophically and legally. But
I think I
21
agree that there are ways to break it down.
22
Ownership is a very broad category, and
125
1
often lawyers talk about it in terms of a
2
bundle of rights. So some of the
issues
3
would have to do with control and a variety
4
of other things.
5 So I think that we can, for
6
instance, think about ownership in ways that
7
exclude the possibility of sale.
We have
8
quasi-property rights recognized.
So I
9
think it is important to focus as broadly as
10
possible on the kinds of concerns that Phil
11
and others have raised, knowing that this
12
ownership question will linger, that we need
13
to break it down into component parts.
14 DR. RAO: Dr.
Salomon?
15 DR. SALOMON:
I want to return to
16
this issue of product again. I
think that
17
for the discussions that follow in the next
18
day or so, this is a central issue.
What is
19
the role of the FDA in regulating this as a
20
product?
21 I don't think it's the only issue,
22
though. I think the problem here
is for the
126
1
committee to realize there's a bit of an
2
awkwardness in the way this flows.
3 You start off with a whole organ
4
pancreas, and you have transplant programs
5
in the United States that are transplanting
6
about 1,000, 1,200, 1,500 whole organ
7
pancreases a year.
8 They take that organ. It's
9 procured as a whole organ and transplanted
10
into a patient as a whole organ, and none of
11
that is touched by the FDA now.
12 Here's where the awkwardness
13
comes. We are also saying that
you are
14
going to take the same whole organ pancreas
15
and procure it in sometimes similar ways,
16
but not exactly the same, because the
17
procurement of a pancreas today for optimal
18
islet isolation is a very
different process
19
from the procurement of a pancreas for whole
20
organ transplantation. That's
something
21
that is an organizational challenge of major
22
proportions for the whole community.
127
1 Then you take that organ and you
2
take it to basically a facility that has its
3
own unique characteristics, and you process
4
it into a cellular product.
5 I personally hope we don't spend
6
hours and hours obsessing about whether it's
7
a product, because as far as I'm concerned,
8
from everything I've known in the last ten
9
years working with the FDA, that's going to
10
be a difficult argument to make other than
11
the fact that, guess what, it's a product.
12 Then you take the product with
13
certain lot release criterion and you put it
14
back into a patient, and from then on, it's
15
a transplant patient again.
16 So I just think that we ought to
17 realize that
there's this really interesting
18
challenge of flow from transplantation to
19
something that's very comfortable in the
20
FDA's eyes as a product, and then back to
21
transplantation.
22 Along that path are all kinds of
128
1
interesting challenges.
2 DR. RASK:
That's really well-said
3
and we should keep that thought
in mind as
4
we look at criteria.
5 Dr. High?
6 DR. HIGH:
Can I just ask one
7
point of clarification? Is an
approved BLA
8
required for reimbursement? This
is linked
9
to the issue of transplantation and product
10
and who owns things. I mean, who
benefits
11
from improved BLA, or could this sort of
12
thing proceed as a sort of subsection of
13
transplantation without an approved BLA?
14 DR. BURDICK:
Was that question
15
for me? I have to know what a
BLA is before
16
I answer.
17 DR. HIGH: Dr.
Noguchi.
18 DR. NOGUCHI:
Quite frankly, the
19
question of reimbursement is very largely
20
tied in with CMS. They are not
represented
21
here. I don't know the actual
conditions
22
under which reimbursement for something such
129
1
as this procedure would be granted.
2 DR. BURDICK:
This is, I think,
3
not something we need to spend a lot of time
4
on now. CMS doesn't regulate
islet
5
transplants and in fact does not have
6
certification for providers for pancreas.
7
It's on an individual basis for the local
8
intermediary.
9 There is no BLA involved in any of
10
the organ transplant processes.
11 DR. RAO:
Maybe I can ask this as
12
one additional question, just to try and
13
frame this clearly in my mind as well.
What
14
happens with blood products? You
know,
15
blood is donated. Now, you
separate blood
16
into components. Is this
regulated as a
17
product? And if it is, then
maybe we have a
18
model which we can consider conceptually at
19
least.
20 DR. NOGUCHI:
Yes, the answer to
21
that is another extremely broad discussion.
22
We have a product office within CBER, the
130
1
Office of Blood Research and Review, which
2
regulates blood and blood components,
3
including what is today considered to be
4
simple taking of red cells from one person,
5
typing it, screening it and beyond, and
6
giving that to another person.
It has its
7
own separate history.
8 But
much of what we are dealing
9
with certainly draws upon that history as
10
well as the history of viral products, viral
11
vaccines. So yes, we do have
that model as
12
well.
13 DR.
EGGERMAN: Yes, I just wanted
14
to get back to the issue of reimbursement.
15
Certainly, CMS has their own process for
16
consideration of reimbursement.
When you
17
deal with an approved product, it flows
18
somewhat in parallel. But when
you take the
19
case of, for instance, pancreas
20
reimbursement, this was an issue that they
21
took up and deliberated upon.
22 For instance, they decided to
131
1
reimburse in the case of where there was a
2
kidney transplantation involved, but not in
3
the case of solitary pancreas
4
transplantation. So CMS does
have a process
5
for considering it. But
certainly if it's
6
approved by the FDA, it's almost a certain
7
thing that will go through and be approved
8
by CMS. But they'd still
consider it.
9 DR. RAO: I
think that was very
10
useful. Then I think the point
you made
11
was, it's true, guess what, it's going to be
12
a product, right? If it is going to be --
13 MS. MEYERS:
Bone marrow
14
transplants. Bone marrow is not
a regulated
15
FDA product, and yet CMS reimburses for bone
16
marrow transplants for a very narrow group
17
of cases. So I don't think a BLA
has
18
anything to do with whether CMS will pay for
19
this.
20 DR. RAO:
That's absolutely
21
correct.
22 DR.
KURTZBERG: But, you know,
132
1
they reimburse the procedure.
They don't
2
actually pay for the product.
3 DR. EGGERMAN:
But one other point
4
is that if it's being considered by FDA for
5
product licensure, then it's somewhat
6
unlikely that they would take it up for
7
consideration of reimbursement, because they
8 would like to have the FDA
involvement in
9
consideration of that product.
Certainly in
10
the cases of organs, for instance, the FDA
11
doesn't have oversight, so they know they
12
have to evaluate whether or not they are
13
going to be involved with consideration of
14
reimbursement.
15 MS. MEYERS:
I think that if we
16
looked at the beginning of bone marrow
17
science, when it was moved out of the
18
experimental stage and moved into the useful
19
stage, if we had put regulations around it,
20
the GMPs, like I just saw on these slides,
21
there wouldn't be any bone marrow
22
transplants because it would have been too
133
1
expensive.
2 You know, could an academic lab
3
have set up these types of things to prove
4
that every batch of bone marrow would be
5
perfect? I mean, the cost has to
be
6
enormous.
7 DR. NOGUCHI:
Well, I think some
8
of our further discussion will be on that.
9
Granted, while at another point in time,
10
costs may have had one particular outcome or
11
another, we are really dealing right now
12
with the current situation.
13 I think
you will find some of the
14
further discussion to be quite useful in
15
addressing the question of can this be done
16
now.
17 DR. RAO: Dr.
Ricordi.
18
DR. RICORDI: Camillo Ricordi from
19
the University of Miami. Just
two quick
20
comments: One is that in the
previous
21
context that we have in discussion on CMS,
22
it was clear that until islet
134
1
transplantation is under IND status, and
2
therefore is experimental, there is no
3
chance for reimbursement or to even begin
4 talking about
reimbursement.
5 So then I don't know whether there
6
is another pathway besides the BLA.
But I
7
think that that will become a condition,
8
like to get after IND status and have
9
something that is approved, maybe.
10 One thing that I hope will be
11
discussed is that if the indication is for
12
only a maximum or 1,000 or 2,000 transplants
13
a year, because that is the
limitation both
14
for the organ source and is also the
15
limitation of patient population, it is as
16
much a group of Type 1, patients with Type 1
17
diabetes, that could be maybe is a product.
18 I agree with them that maybe we
19
should consider it a product, but it's not a
20
product in the classical term, because it's
21
something that could become like a
22 biological
license application for an orphan
135
1
biologic.
2 The challenge would be to set what
3
are the parameters determined to define the
4
safety and potency of this product that are
5
probably very different from the standard
6
product release characteristics that you
7
would expect for a pharmaceutical.
8 So that
will be the challenge.
9 DR. RAO: I
think that's very
10
clear. Hopefully, with your
talks, we which
11
will hear a little bit more about what's
12
feasible and what's not.
13 On that thought, I move that we
14
take a break and then listen to specific
15
talks on islet transplants.
16 (Recess)
17 DR. RAO: We
would like to resume.
18
Our first speaker is going to be
19
Dr. Ricordi, who will talk about islet
20
processing, evolution and current standards.
21 DR. RICORDI:
Thank you. It is
22
indeed a pleasure to be able to give this
136
1
talk on behalf of also Dr. Hering and
2
Dr. Shapiro. This is a
demonstration that
3
this is a uniform product that follows the
4
same rules at the three leading
5
institutions.
6 The challenge for us to do
7
pancreas and islet transplantation has been
8
for the last 20 years I thought that the
9
challenge was for islets because it is a
10
more simple procedure, safer, that can be
11
done as an outpatient procedure, and much
12
easier than the more complex organ
13
transplant.
14 From what we heard in the
15
introduction, it seems that it is the
16
opposite. It is much more
complex to do an
17
islet transplant than an organ transplant.
18
So I hope that by
the end of these
19
two days, we will not be all convinced to go
20
back to the organ transplantation.
21 I eliminated some of the slides
22
because some of these issues have been
137
1
addressed by the wonderful introductory
2
talks. But indeed islet
isolation and
3
processing start with pancreas procurement
4
and donor selection.
5 This is just to show you that
6
there are several criteria, most of them are
7
in common with other fields of
8
transplantation, but there is a fairly
9
well-standardized way to select donors with
10
very well-defined inclusion criteria that
11
cover like all the viral testing, the fact
12
that there is adequate in situ hypothermic
13
perturfusion of the organs;
maximum cold
14
ischemia time for the organs and age rate
15
between 15 and 65.
16 There as well very well-defined
17
exclusion criteria for which we don't accept
18
a pancreas if any of this element is met,
19
including warm ischemia, before close
20
clamp(?) or any positivity in this testing,
21
or cardiac arrest -- when stable circulation
22
cannot be achieved in the following two
138
1
days.
2 So I would not spend a lot of time
3
in the detail of this, but this is just to
4 show that the
scientists working in this
5
field are following a very well-defined
6
selection of which donors are suitable for a
7
pancreas to then enter the isolation
8
process.
9 There are also some studies that
10
have addressed, like this one from Edmonton,
11
variables that can predict success of the
12
isolation or predict failure of an
13
isolation. And these are also
studies that
14
we can use to try to identify which ones are
15
the best potential donors.
16 An ideal pancreas donor does not
17
exist, as they usually say, because the
18
donor would be still alive. So that if he
19
died, there is generally some element of
20
disturb to the organ, and there is a study
21
emerging on how brain death is one of the
22
major problems also of instability of an
139
1
organ.
2 But if we would define an ideal
3
range, it would be like between age 15
4
and 45, and that the pancreas is kept cold
5
during procurement. This is
essential,
6
because we may receive an organ in the
7
isolation facility that is already
8
compromised. Generally, higher
BMIs are
9 favorable. Short hospital stay is a good
10
predictor.
11 Also, we look at creatinine levels
12
below 1.5, because generally, it's a sign of
13
ischemic damage. If the pancreas
takes a
14
ischemic hit, islets are very sensitive to
15
oxygen levels and to hypoxia. So
we know if
16
the kidney was damaged, probably your islets
17
were as well.
18 The pancreas procurement -- and
19
this is to show you how the first
20
significant portion of the islet isolation
21
is like a surgical procedure. So
if
22
pancreas procurement is not done
140
1
appropriately, especially maintaining a lot
2
of iced saline around the pancreas and
3
cooling the cut of the pancreas while other
4
organs are removed generally before the
5
pancreas, you can compromise the ability to
6
purify and separate enough valuable cells.
7 This has been shown in several
8
studies. This one is from the
Edmonton
9
group, monitoring temperature during
10
pancreas procurement, and showing how if
11
you're not very careful packing iced saline
12
around the pancreas, you can reach
13
temperatures that are around 18 degrees for
14
a prolonged time instead of going rapidly
15
towards 5 or 6 degrees Centigrade.
16 So this slope could be responsible
17
for yields that are half of what you would
18 obtain if the
pancreas is procured
19
correctly. This is just to say
that just a
20
little detail even before you start any
21
variability in processing may have an
22
effect.
141
1 Now we know also that because the
2
islets are so sensitive to hypoxia and low
3
oxygen levels, it is important also to
4
preserve the pancreas before processing with
5
the so-called two-layer pancreas
6
preservation solution.
7 These are all recent studies from
8
groups both in Minnesota, Miami, Edmonton,
9 confirming the
initial data in animals from
10
the Japanese groups of Professor Koroda and
11
Matsumoto that adding oxygenated
12
fluorocarbons to the AW (?) solution
13
increased energy level of the pancreas,
14
increase ATP, and improved islet isolation,
15
even from marginal donors.
16 Then finally the pancreas arrives
17
to an isolation facility, and the process is
18
still based pretty much on a concept that
19
was introduced in the late '80s; that is a
20
continuous digestion process after an enzyme
21
is injected in the pancreas, in which the
22
pancreas is progressively disassembled into
142
1
smaller and smaller fragments.
2 After the size of cluster of cells
3
of the islet that can then be freed from
4
this digestion chamber by a constant flow
5
that passes the particles through a screen,
6
and then you save the islets from any
7
further enzymatic digestion by cooling and
8
dilution.
9 So some advances in the procedure
10
have included the introduction of endotoxin.
11
That's the reagents like the new lot of
12
enzyme blends currently produced that allow
13
us now to have all the
products that there
14
can be using an islet processing, being
15
FDA-compliant, or possibly FDA-approved and
16
safe.
17 So this is a major point, that
18
everything we use now for islet
processing,
19
I think, is adequate.
20 Then we have another substantial
21
part in isolation lab for surgery in the
22
pancreas, separation from the duodenum,
143
1
cannulation of the ducts division and
2
trimming of all the fat and membranes from
3
the organ so that there is still something
4
that is very far away from a controlled cell
5
processing, as you would have in a
6
methodologic (?) cell processing.
7 Then the processes I mentioned,
8
you infuse an enzyme blend by controlled
9
perfusion, and here you can control the
10
pressure of the fusion with the flow of the
11
time pressure. You perform this
continued
12
digestion process and then you have your
13
final product, I mean, this
first pancreas
14
and the fibrous network of ducts and vessels
15
is retained in the chamber.
16 So this whole system can be
17
monitored throughout the digestion to see
18
when the islets are liberated and then
19
finally the whole product is purified
20
through density gradients using the
21
preferred ways this COBE 2991 synthesizer.
22
The final product is then infused by gravity
144
1
in the liver.
2 There are also more recent
3
introduction of multiparametric monitoring
4 systems during
processing that are even
5
further increasing our ability to constantly
6
check different parameters like pH, PC02 and
7
temperature throughout the digestion in the
8
chamber. This will be useful to
introduce
9
further elements of in-process monitoring
10
and will allow in the future maybe also to
11
introduce in process a correction, if you
12
see like the pH going in the wrong direction
13
or other parameters.
14 So that this is becoming, compared
15
to how we were doing islet isolation ten
16
years ago or five years ago, it is becoming
17
a much more controlled and regular and
18
reproducible system.
19 Now, things that have been
20
introduced besides the fluorocarbon,
21
pre-isolation preservation of the pancreas,
22
is also the pre-transplant culture, which is
145
1
equivalent or compatible, as you say it, at
2
the three major institutions, at Edmonton,
3
Miami or Minnesota.
4 We all do pre-transplant culture
5
with similar systems. The
importance is not
6
just to allow us to transfer a patient from
7
a remote site or to ship the product to a
8 remote site, but
also that will allow us to
9
perform those product release criteria,
10
including the biologic system in nude mice,
11
or the stimulation index before the islets
12
are infused, because we would have two,
13
three days time lapse between isolation and
14
transplant.
15 We may have also some inflammatory
16
immunologic benefits, last but not least the
17
fact that you can start the
18
immunosuppressive therapy and achieve
19
therapeutic levels of anti-rejection drugs
20
used before the islets are infused.
21 So finally, so the final product
22 is introduced by
gravity, using blood
146
1
transfusions, which is a very simple
2
procedure. Following
percutaneous
3
catheterization of the portal vein, and is
4
just infused by gravity with the islets just
5
mixed in the bag, because the islets have a
6
tendency to settle very fast.
7 The facility, and this is just a
8
picture of the Minnesota facility, but they
9
are similar at Edmonton and at Miami,
10
meeting all those, we hope at least, we are
11
meeting most of the regulatory requirements
12
for cell manufacturing, and we have to scale
13
up these requirements, especially now in
14
Miami, but also Minnesota, with the fact
15
that we are shipping this product across
16
state barriers, like the processed islets in
17
Miami and transplantations like at Baylor in
18
Texas.
19 But we follow all the general CG&P
20
and GTP regulation in maintaining clear flow
21
of personnel, material, product in ways with
22
all the controls for equipment,
147
1
environmental, and a huge series of general
2
standard operating procedures that are now
3
available at the centers just addressing
4
general procedures for the facility, vendor
5
approval, sterility follow-up and
6
investigation, process validation and so on,
7
equipment validation, and then the
8
human-specific processing facility, SOPs for
9
every single step of the process.
10 So this has been a tremendous
11
exercise. We are tremendously
grateful to
12
the FDA to allow us to improve through this
13
exercise so that now we are indeed able to
14
deliver a more consistent, reproducible and
15
better quality product.
16 We have preclinical safety and
17
efficacy testing. This is only
data from
18
Miami, but we had 188 papers published using
19
this as method. The first in all
those
20
papers were related to the first attempt to
21
perform an isolation assessment and
22
introduce the first criteria for defining a
148
1
good from a bad islet preparation.
2
Still now, this method
that was
3
introduced in the '80s that is just
4
introduced an adequate of islets beneath the
5
renal capsule of a diabetic mouse, a nude
6
mouse that cannot reject the tissue, but
7
that is made diabetic before the transplant.
8
You see, you can follow the normalization of
9
glucose levels and then perform a refraction
10
of the kidney, put in the graft and a return
11
to the diabetic state will demonstrate that
12
where indeed the transplanted islets
13
responsible for the cell biological effect.
14
And this, besides giving you a physiologic
15
readout, will give you also the opportunity
16
to perform a section on the kidney
17
transplanted and analyze the islets for
18
composition, both insulin glucatose (?) and
19
other.
20 We have several studies now in
21
nonhuman primates using the similar
22
compatible islet processing system, both in
149
1
studies in collaboration with Dr. Hering's
2
group in Minnesota and others.
This has
3
been used already to provide preclinical
4
testing for the introduction in clinical
5
trials of new or novel immunosuppressive
6 agent.
7 We have papers in common islet
8
transplant models, a human islet, which of
9
course is still the best animal model
10
available for efficacy, and that percent (?)
11
islet papers as well. I think
that the
12
clinical data really will, I hope, will
13
convince this panel of the efficacy of this
14
procedure and the way that islets are
15
manufactured so far at the different
16
institutions.
17 So current product, at least
18
criteria, address both cell identity, yield,
19
viability, lack of contamination and
20
reducing the toxin contained in levels
21
acceptable by FDA, and post-release assays
22
complement also the release criteria include
150
1
mycoplasma, sterility and potency in the
2
cases especially where the islet product is
3
infused fresh without any culture.
4 But in the last years, all our
5
centers are performing two to three days
6
culture before transplant, and that should
7
enable us to address better potency tests.
8
This is a consistency and reproducible of
9
the product and the product release criteria
10
of the University of Miami. It
shows, you
11
know, the stimulation index that must be
12
over one. This is the incidence
of
13
secretion in response to glucose challenge
14
of the islets in vitro.
15 We have viability that must be
16
greater than 70 percent, and is so in all
17
cases. We have all the sterility
tests and
18
endotoxin tests, both aerobic and anaerobic
19
fungal (?) mycoplasma, gram stain and
20
endotoxin all within acceptable range, and
21
the purity of the final product as well as
22
the islet number, the kilogram of recipient
151
1 body weight that has to be
above like 4,000
2
islet equivalent per kilogram.
3 These are all the products that
4
have been transplanted in patients.
They
5
all met these requirements, both of potency,
6
viability and identity. This is
including a
7
transplant that we have performed at the
8
Baylor College of Medicine in Houston.
All
9
similar results. The product was
shipped
10
there.
11 These are the same summary tables
12
for the University of Minnesota, with
13
virtually 100 percent of products that were
14
within the specification for viability,
15
potency, safety and purity, with the islet
16
yield that allow enough of the process ÄÄÄÄ
17
to be sufficient for transplantation.
18 This is added detail on the
19
aerobic and anaerobic fungal microplasma,
20
confirmed within specification.
This is the
21
viability group of stimulation, purity and
22
islet yield. This is the same
test that the
152
1
University of Alberta, always very much
2
within the range specified for acceptability
3
of the product, for safety, older culture
4
and aerobic and anaerobic fungal also all
5
negative.
6 So comparability of the products,
7
I would be unfair if I pretended we have
8
exactly the same process, but I think the
9
differences are minimal and acceptable.
10 We constantly compare, and every
11
time one of us has a product or a process
12
step that we think is improved, we compare
13
it head-to-head in a controlled environment,
14
and then we agree or we agree to disagree,
15
like in this case, where these two gradients
16
show in seven consecutive pancreas
17
processing equivalent results with a quality
18
score around eight.
19 But this is a minimal difference
20
in the composition of the solution used for
21
density separation, and we would very well
22
agree to all use one or the other if this
153
1
would constitute a problem.
2 Now, this method, the final
3
potency test, is reversal of diabetes
4
following transplantation. Dr.
Shapiro will
5
address this tomorrow extensively.
Just as
6
an introduction, since the introduction of
7
the automatic method in islet processing,
8
the results in islet auto-transplant where
9 you don't have
problems of overlapping
10
autoimmunity or toxicity of
11
immunosuppressive drugs that had been quite
12
remarkable, with 80 percent of the patient
13
long term. The longest cases
now, around 15
14
years, post-transplant, with the safe
15
procedure that has been indeed reproducible.
16 This is in the allo transplant
17
field. This was the first trial
in
18
Pittsburgh. It was the first series of
19
successful islet allographs in patients
20
receiving a liver transplant that were
21
performed at the University of Pittsburgh
22
in 1990.
154
1 Then the well-known challengers
2
were islets transplanted together with
3
keetners (?) in the presence of steroids in
4
conjunction with calconeural inhibitor, and
5
then since the introduction of the Edmonton
6
protocol in rappamycin (?) with steroid
7
immunosuppression, much better results in
8
the clinical trials are reported so far.
9 So I'll conclude here this brief
10
introduction.
11 DR. RAO:
Thank you. We have time
12
for questions.
13 DR. BLAZAR:
Could you make a
14 comment on the exportability of the SOPs and
15
the processing to centers that are not quite
16
as large as the ones you illustrated?
17 DR. RICORDI:
Well, that is a very
18
important question. It raises
the issue of
19
where this processing should be done, which
20
facility had the most likelihood of success
21
and which environment would be the most
22
cost-effective.
155
1 At the University of Miami, we
2
selected the path that is not justified just
3
to have a facility with the degree of
4
complexity just to do islet cell transplant,
5
because you need to perform 150 to 100
6
procedure a year to become cost-effective,
7
justify keeping teams on call, available 24
8
hours 7 in shifts and all the process of
9
validation facility requirements.
10 So I think you have right now a
11
very strong requirement of institutional
12
commitment. You need to have a
lot of
13
support from all of the agencies.
Juvenile
14
Diabetes Foundation is helping the NCR and
15
NIDDK and NIH. But still that
would not be
16
sufficient if you don't keep up a volume of
17
operation that will allow your team to be
18
constantly trained and very fast.
19 I think that it's a combination.
20
You know, we inherit the complexity of
21
manufacturing of cellular product in a
22
regulated environment, with the challenge
156
1
and the complexity of experience like would
2
be developed through organ transplant
3
programs that perform over 100 transplants a
4
year versus a transplant program that
5
perform less than ten.
6 DR. MULLIGAN:
I have two
7
questions. On the question of
potency,
8
there is the numbers from one to nine or
9
something. So in cases where
there is
10
comparable viability of the preps (?) what
11
does it mean that one gets a potency score
12
of nine and one gets a potency score of two?
13
Is that a significant thing?
14 DR. RICORDI:
Yeah. This potency
15
score is actually something that Dr. Hering
16
may want to comment on, because it is a
17
complex system that takes in account all the
18
islet shape, viability, and there's enough
19
time to create a score, but is not what we
20
actually use as product release criteria.
21 That was just to compare two
22
separate processing, two separate separation
157
1
techniques. What we use as a
potency
2
release criteria for the product is just the
3
islets have to be viable; they have to be
4
more than 70 percent viable based on
5
inclusion-exclusion dyes.
6 We have to provide the gram stain
7
for sterility and we have to provide, we are
8
performing a stimulation index that is
9
insulin release following a glucose
10
challenge, which in the case of fresh islet
11
transplant product is available after the
12
islets are transplanted, but in the case of
13
a cultured product could be available
14
before.
15 Then the best potency test remains
16
today the reversal of the diabetes in the
17 nude mouse. But that also can be done only
18
if the product is transplanted a few days
19
after.
20 DR. MULLIGAN:
I was curious, you
21
know, what actually makes up that score
22 exactly. Maybe you can talk about that.
158
1
The issue is essentially as it appears when
2
you see that, that you have two preps that
3 are processed in a very comparable way, yet
4
they have this very different score.
So
5
that's the first part of it.
6 The second was the issue of, is
7
there a property of the cells for
8
transplantation that influences their
9
capacity to amplify after transplantation?
10
I guess a piece of this is, I would think in
11
principle, since the cells can regulate
12
insulin secretion based on glucose
13
concentration, the dose of transplant cells
14
may not make all that much difference if
15
you're above some threshold.
16 But I'm curious whether it does
17
make a difference. Is there an
optimal
18
number of cells to end up in the recipient
19
to get the optimal glucose sensing, and is
20
there some way that you can test to see if
21
there's a measure, a property of the cells
22
that gets transplanted that influences
159
1
whether they have the capacity to amplify in
2
the host?
3 I guess the analogy is to, you
4
know, field cells for Parkinson's, and the
5
question of whether or not there's any
6
safety considerations due to some property
7
of the cells that you wouldn't easily be
8
able to tell, but after transplantation may
9
be important, like they grow to a level
10
that's too high or something.
11 DR. RICORDI:
Yeah, in a way.
12
Well, this is just a table. The
quality of
13
score difference was 7.9 or 8.0, so it's
14
just a .1 difference. But
Bernard can get
15
into more detail, and he will address all
16
the potency testing and the new quality
17
control that we are studying to move to the
18
next level of complexity.
19 But you are perfectly right. You
20
know, viability may not be sufficient.
In
21
our studies, compared to Parkinson's, we are
22
lacking in a way because the readout in the
160
1
patient is much more in need.
Like, we
2
know, you know, within a week or four weeks
3
exactly how the islets are functioning, how
4
they are recovering, and if the patient is
5
on insulin or off insulin, you can test the
6
peptide as a measure of the function of the
7
transplanted islets.
8 What is the challenge in the test,
9
the potency test pre-transplant, is that
10
these cells have seen hell; they pass from
11
brain-dead to the organ procurement to
12
preservation. They go from four
degrees,
13
then suddenly they are exposed to 37 to
14
activate the enzymes, and they are freed in
15
an environment that has multiple, you know,
16
collagenous ripzyne (?) and other enzymes
17
and they are saved again by dilution and
18
cooling.
19 So sometimes you have score as far
20
as potency in insulin release that are close
21
to one to no stimulation, where the patient
22
is still going off the insulin after
161
1
transplantation.
2 So the
challenge now that we're
3
trying to combine is this with other tests,
4
and maybe will not be a single one.
We have
5
very strong evidence, for example, that if
6
you combine insulin release to oxygen
7
consumption, the mitochondria membrane
8
potential remained to have much more
9
predictive tests.
10 DR. MULLIGAN:
Here's an animal
11
model question. If you
transplant
12
sufficient cells in normalized glucose and
13
then you compare the way in which the
14
insulin secretion is regulated by glucose,
15
if you give ten times that number of
16
glucose, is there any difference in like the
17
time in which it takes to normalize glucose
18
levels?
19 That is, is there any difference
20
in the way you regulate secretion based on
21 the total load of cells that you have?
22 DR. RICORDI:
Definitely. Like we
162
1
are dealing in most of the cases there is
2
still a reduce beta cell mass that we
3
infuse. So we didn't have to
deal with this
4
problem of having an excess
5
insulin-producing mass because we are taking
6
a hit as far as recovery coming from a
7
pancreas and then separating the islets
8
preserving, et cetera.
9 But in models, like in animal
10
models like non-human primates, where we try
11
to transfer like four times the dose
12
necessary to reverse diabetes, still you
13
indeed have much faster normalization of
14
glucose levels after a meal.
15 There are luckily protective
16
mechanisms against hypoglycemia so that
17
severe hypoglycemia has never been observed
18
after an islet cell transplant.
19 So that at the beginning of the
20
clinical experience several years ago, one
21 of the worries like
in the operating room,
22
there were people ready with glucose
163
1
solutions, but the reality has been that so
2
far, it has been a very safe procedure as
3
far as the risk of overtreating.
4 DR. MULLIGAN:
But in terms of the
5
therapeutic efficacy, it sounds like that,
6
you know, tests to see how quickly you
7
normalize glucose are obviously important.
8 I guess my point is just this
9
issue --
10 DR. RICORDI:
As you can see the
11
curve, how fast you normalize glucose as an
12
index of the potency of the product.
We do
13
it in a way like we have a rodent model in
14
the nude mice, and in other rodent models
15
that we use to take is like the minimal mass
16 models, where you
transplant a subliminal
17
sub-therapeutic dose of islets.
18 Then you measure time to normalize
19
glycemia so that you know that if you have
20
an average product, it should take 40 days
21
to become completely normal glycemic, and
22
then any experimental condition that improve
164
1
your product should shift the curve to the
2
left or if you are doing worse, shifting it
3
to the right.
4 But this has been so far at
5
experimental level. We are not
applying it
6
to the clinical set.
7 DR. ALLAN: I
have sort of a
8
general question. Is there a
criteria
9
that's used to decide whether to use a
10
pancreas as a whole organ transplant versus
11
to process it into islets? I
mean, does the
12
University of Miami still do whole organ
13
pancreas transplants?
14 DR. RICORDI:
Right now, and this
15
is something we'll discuss hopefully more
16 tomorrow, but one of
the major challenges of
17
islet transplantation is that we have to
18
demonstrate comparability with pancreas
19
transplant results using the organs that
20
have been refused or second-class or in any
21
way discarded from utilization as whole
22
organ transplants.
165
1 So at the University of Miami,
2 every pancreas
that we have used in our
3
trials has been a pancreas that have been
4
turned down locally, regionally or
5
nationally by all the organ transplant
6
programs.
7
This will be a very
important
8
issue to discuss, because, you know, the
9
probability of success and more importantly
10
long-term results will most likely be also
11
proportionate and parallel the quality of
12
the initial organ.
13 So we have a double challenge. On
14
one side, how we can use more and more and
15
make every pancreas count and try to develop
16
strategies that are successful to retrieve
17
islets from any organ that is not used for
18
whole organ transplant. We also
have to
19
learn from the whole organ transplant field
20
that there are reasons why you don't use an
21
organ if there has been cardiac arrest or
22
hypoxia or a drown victim. There
are some
166
1
things that have been painfully learned in
2
the clinical experience in whole organ
3
transplants, and we shouldn't have the same
4
mistakes made.
5 DR. RASK: I
just have a quick
6
question on the assessment of cell
7
viability. I was wondering if
the dyes that
8
you use are able to distinguish between
9
intact life cells and cells that have
10
undergone apoptosis. Are those
dyes able to
11
make a distinction between
those two cell
12
types?
13 DR. RICORDI:
That's a very
14
important question. As a matter
of fact, it
15
is more sensitive if these same tests are
16
combined with an assessment of apoptosis,
17
because sometimes, to have a viable product
18
today doesn't mean necessarily that you have
19
a viable product tomorrow.
20 If you have a loss of cell, you
21
generally see it in the first day in
22
culture, and this is why we don't like to
167
1
transplant cells fresh, because we have a
2 higher risk of
primary non-function if you
3
have a product that was damaged or already
4
doomed to fail or maybe the apoptosis was
5
already triggered at the time of organ
6
procurement.
7 Then if you transplant
8
immediately, you will not see it.
That's
9
why we repeat this test also after the day
10
of isolation. So far, we have
done, what,
11
over 60 transplants clinically, the three
12
institutions, with culture products, and
13
there have been not a single failure of
14
function post-transplant.
15 Our exercise is the opposite. We
16
are doing all this
post-transplant release
17
criteria trying to find the correlation with
18
failure and something that would have
19
predicted failure. But the good
situation
20
where we are right now is that so far is if
21
you follow this very basic criteria that are
22
similar in stem cell transplantation in a
168
1
hematapoietic progenital for positive cells,
2
you can base this release criteria on flow
3
site dometry (?) viability and very simple
4
based on count and how many ÄÄÄÄ do you
5
have, how viable is the product.
And so
6 far, we have
been in the same situation with
7
the clinical trials at our institutions.
8 DR. BLAZAR:
You know, I wanted to
9
just push this point I asked you before
10
about the different-sized centers.
In terms
11
of patient outcome data, is outcome
12
different in centers that process more
13
islets than ones that don't, and what's your
14
feeling for, is there a critical number in
15
terms of islet cell processings for patient
16
safety that has derived from these outcome
17
data in the field so far that would limit
18
this to certain sites, or is this a training
19
issue or financial issue?
20 DR. RICORDI:
I think it will be a
21
common sense issue. I think NIH
has made a
22
major step forward creating the Islet Cell
169
1
Resource Center that will be a regional
2
center for distribution of islet cell
3
product to other facilities in the region.
4 How many do you need in the United
5
States of such facility?
Probably, you
6
know, you may need only one to three
7
initially until the number of transplants is
8
up to 1,000. But there will be a
natural
9
selection that is also defined by the degree
10
of institutional commitment and volume that
11
you need to put into these kind of efforts.
12 I have a personal bias that it is
13
as important, the volume that you have as
14
the technology of the collagens, or the
15
quality of the pancreas. You
know, there
16
are preliminary results of multi-center
17
trials that have been reported showing this
18 degree of
difference between some center
19
with close to 100 percent success rate and
20
then others with close to zero, so that our
21
real issues is not just an hypothetical
22
concern.
170
1 DR. RAO: One
last question, of
2
all the cells that you have transplanted,
3
has there been any one where you have frozen
4 the cells and then
thawed them again? Is
5
there any time when you pooled the cells?
6 DR. RICORDI:
Yeah. We don't like
7
to perform cryo-preservation at this stage,
8
even though we don't exclude this in the
9
future, because it would be a very good way
10
to pool suboptimal preservation of
11
processing so that you can use pool donors
12
into a single transplant.
13
We have occasionally
combined a
14
preparation in which there were good quality
15
but not enough numbers. But this
is
16
something I would not advise to do, and we
17
don't want to do it prospectively in the
18
future because it would confuse the
19
interpretation of which were the
20
characteristic of that donor that can
21
parallel success.
22 So then if you start pulling
171
1
together more than one, at this stage, it
2
would be better not to do that in the
3
future. Of course, it could be
banking
4 preservation and
pooling of different
5
product could be -- one word of caution is
6
if you did pool several donors, the more
7
donors you use, the more could be the
8
concern of having that sensitivity to donor
9
and to antigens, and especially moving
10
toward tolerance induction strategies where
11
we try to induce donor-specific tolerance to
12
that islet product.
13 I think the
way to go is to do the
14
single donor islet transplant, maybe
15
sacrificing insulin independence as an end
16
point, but going for good metabolic control
17
in the absence of hypoglycemia, but trying
18
to get this islet to survive without
19
immunosuppression.
20 DR. RAO:
There are no more
21
questions.
22 DR. RICORDI:
Thanks.
172
1 DR. RAO: Our
next speaker is
2
Dr. Hering.
3 DR. HERING:
Mr. Chairman,
4
Dr. Goodman, Dr. Noguchi, members of the
5
committee, it is my task on
behalf of my
6
colleagues, Dr. Ricordi and Dr. Shapiro and
7
also Dr. Pappas (?) to address the important
8
issue of islet potency testing.
9 The objectives are the following:
10
To determine whether product release testing
11
results are within a pre-defined
12
specification range; to determine whether
13
the manufacturing processes are controlled
14
and consistent; and
importantly, to prevent,
15
detect and correct deficiencies in the
16
manufacturing process that may compromise
17
product integrity or function, or may lead
18
to the transmission of infectious agents.
19 Assessing the final islet product
20
involves assessing product safety and
21
product quality. Safety
assessments involve
22
sterility, endotoxin, mycoplasma, and
173
1
quality addresses identity, purity and
2
potency. Potency has presented
significant
3
challenges, and I will focus therefore on
4
potency testing.
5 Potency testing is important in
6
particular for two reasons: A
reasonable
7
measure of product potency must be developed
8
and demonstrated prior to the approval of a
9
BLA. That is, we need a
predictive potency
10
assay. And the second: Results must be
11
available prior to patient administration;
12
that is, we need a real-time potency assay.
13 How can we
approach this
14
challenge? We will give high
priority to a
15
development of very robust assays for
16
measuring viable beta cell mass, and in
17
addition, fractional beta cell viability.
18 Then we will ask the question
19
whether these assays will predict
20
post-transplant islet function in a diabetic
21
nude mouse bioassay.
22 Finally, we have to ask the
174
1
question whether the nude mouse bioassay
2
will predict post-transplant islet function
3
in a relevant pre-clinical large animal
4 model, and also in Type 1
diabetic islet
5
allograft recipients.
6 What are important assessment
7
parameters? With respect to
viable beta
8
cell mass, we have reviewed and tested a
9
number of candidate assays. We
have decided
10
to pursue the development and
11
standardization and implementation of beta
12
cell oxygen consumption rate assays, or in
13
addition, beta cell ATP content.
14 With respect to fractional
15
viability, we will focus on oxygen
16
consumption rate per DNA, or ATP per DNA.
17 So the first question is can we do
18 this, can we predict
outcome? Can we do
19
tests? Or in other words, can
human islets
20
be cultured to facilitate free transplant
21
islet product potency assessment?
22 The answer is yes, and as pointed
175
1
out before, 65 islet transplants have been
2
performed at our three institutions, using
3
cultured islets. Islets are
isolated on day
4
minus two, kept in culture and transplanted
5
on day zero, and induction immunosuppression
6
is typically initiated on day minus two.
7
This is a general approach followed at an
8
increasing number of institutions.
9 So we have at least two days to
10
perform islet potency testing and islet
11
safety testing.
12 Now, the question is do islet
13 product testing
results obtained in the
14
diabetic nude mouse bioassay correlate with
15
post-transplant islet function in Type 1
16
diabetic recipients or in nonhuman primates.
17 I have listed here results
18
obtained in 15 consecutive nonhuman primate
19
islet allo transplants. You can
see in all
20
but two recipients, normal glycemia and
21
insulin independence have been restored
22
after islet allo transplantation.
176
1 Now, this slide shows the
2
concordance rate of primate and mouse
3
recipients of primate islet grafts from the
4
same donor with insulin independence or
5
euglycemia as readout.
6 Here you see outcomes in nonhuman
7
primates, and you see all but two animals
8 achieved insulin
independence. And mice
9
that received islets from the very same
10
product became euglycemic, except one mouse.
11
And here we have a concordance.
We
12
encountered problems during pancreas
13
procurement, technical problems, and we were
14
not able to reverse diabetes in mice that
15
received an islet aliquot.
16 Here, this particular monkey
17
received an islet transplant after a stem
18
cell transplant from the very same donor
19
animal and was sensitized at the time of
20
transplantation. So here, we
encountered an
21
immunological, but not a technical problem.
22
Therefore, there was no concordance.
177
1 I have reviewed outcomes in 19
2
single donor islet allografts at our
3 institution. In 16 patients, insulin
4
independence was restored. Two
resulted in
5
partial function and one showed primary
6
non-function.
7 Now, again, what is the
8 concordance of human
and mouse recipients of
9
human islet grafts from the same donor with
10
insulin independence/euglycemia as readout?
11
Three patients failed to achieve insulin
12
independence. In all three
instances, islet
13
aliquots from the very same pancreas failed
14
to restore euglycemia in diabetic mice,
15
indicating that the sensitivity is 100
16
percent. The specificity is not
100
17
percent, but nevertheless, this is
18
encouraging and promising.
19 Now, I will focus on two assays.
20
The first: Cellular composition,
and beta
21
cell mass. Now, techniques are
available to
22
determine the cellular composition and the
178
1
beta cell mass of an islet product.
Islets
2
are disassociated into single cell
3
preparations without a substantial loss of
4
cells. Based on DNA, cell
recovery is 95
5
percent or higher. Cells are now
exposed to
6
a panel of primary antibodies directed
7
against insulin glucagon
somatstatin NPP
8
amalayse and cytocarotine 19, and then
9
analyzed using image analysis techniques.
10 If you do this, you obtain images
11
like this. And now if you
analyze the
12
results, you see here in 55 consecutive
13
human preparations, 32 percent of the cells
14
were beta cells. In nonhuman
primate
15
preparations 34, porcine islet
16
preparations, 42 percent.
17 This is very well established at
18
all our institutions. You can,
of course,
19
calculate now purity as a sum of beta cells,
20
alpha cells, delta cells or PP cells divided
21
by the total cell number.
22 You can calculate the total number
179
1
of beta cells based on DNA content and
2
percent of beta cells in the preparation.
3 Now, we asked the question, is
4
this information important? Will
this be
5
helpful? If you look at 42
consecutive
6
human islet products transplanted into a
7
diabetic nude mouse and 62 percent, we were
8
able to achieve euglycemia; now if after
9
transplanting 2,000 islet equivalents, if we
10
analyze data based on beta cell numbers, we
11
see that 89 percent of mice that received
12
more than 2.25 million beta cells within
13
the 2,000 islet equivalents achieved
14
euglycemia.
15 We also asked the question whether
16 there's a dose
response and whether there is
17
a beta cell number that would consistently
18
reverse diabetes in this model.
If you
19
transplant more than 100 million per
20
kilogram recipient, you see normal glycemia
21
in all animals.
22 Then we asked the question, is
180
1
this information important in predicting
2
sustained insulin independence in Type 1
3
diabetic recipients of islet allografts, and
4
in red, you see patients that failed to
5
achieve or maintain insulin independence for
6
more than one or two or three years.
7 Here in yellow are patients that
8
remained insulin-independent.
Those
9
patients who failed to maintain insulin
10
independence long-term received a lower beta
11 cell mass, again
indicating that this is
12
important information.
13 Now, we also asked the question,
14
is there a beta cell mass that would
15
consistently result in sustained insulin
16
independence, and if patients receive more
17
than five million beta cells per kilogram
18
recipient body weight, then it is very
19
likely that sustained insulin independence
20
can be achieved.
21 I want to turn now to oxygen
22
consumption rate. Most of the
results that
181
1
I represent have been generated by
2 Dr. Glycos
Pappas, who is here in the
3
audience, and Dr. Clark Colton.
4 This is the apparatus, a
5
dual-chamber oxygen consumption rate
6
apparatus that is used and that has been
7
customized for use in islet assessment by
8
Dr. Pappas and Colton.
9 Now, the principle is illustrated
10
here. Isolated islets are
transferred to a
11
chamber, and oxygen consumption is measured.
12
Oxygen consumption is proportional to the
13
viability. That is, there is
very low
14
oxygen consumption with low viability, and
15
higher viability clearly is associated with
16
very fast oxygen consumption.
This is then
17
a readout of the number of viable islet
18
equivalents.
19 The slope from the OCR measurement
20
correlates with the number of viable islet
21
equivalents. So now we ask the
question:
22
Is it possible to predict outcome
182
1
post-transplant based on OCR and OCR per DNA
2
information.
3 So the assessment parameters are
4
the following: OCR is
proportional to the
5
number of viable cells, or the volume of
6
viable tissue, and the DNA, of course,
7
indicates the number of cells.
And OCR per
8
DNA is a measure of viable tissue volume, or
9
quality of the tissue.
10 Then we also assess the fractional
11
viability; that is OCR per DNA divided by
12
OCR per DNA of a completely viable islet.
13 Again, we perform nude mouse
14
transplants. Three principal
outcomes can
15
be recorded. Animals show a Type
A outcome.
16
Animals become normal glycemic and remain
17
normal glycemic after transplantation of
18
islets. Also, partial function
or failure
19
to achieve euglycemia, which would be
20
Outcome C.
21 Then we ask the question whether
22
we can use this information in order to
183
1
predict outcomes. On the X axis,
we have
2 plotted normalized
oxygen consumption rate
3
or OCR, per DNA, and on the Y axis, oxygen
4
consumption rate, OCR, as a measure of the
5
viable islet equivalents. You
can now
6
calculate the number of viable islet
7
equivalents based on this information.
8 Now, if you do a number of
9
transplants using rat islets transplanted
10
into diabetic Bulb/C mice immunosuppressed
11
with a depleting anti-CD-4 antibody, and if
12
you now, for example, transplant islets with
13
a very high fractional viability of 350
14
nanomole per minute per milligram DNA, then
15
with a low number of islet equivalents, you
16
will see a Type A response in six of six
17
recipients.
18 However, if your viability with
19
the very same number of viable islet
20
equivalents is reduced, you will not be able
21
to reverse diabetes in any animal.
And you
22
can also follow this line and you look at a
184
1
very low number of viable islet equivalents.
2
Now, the fractional viability is very
3
important.
4 So to put this into different
5
words, if you transplant a high proportion
6
of dead islets or dead cells, this clearly
7
compromises outcome, even though you may
8
have the same number of viable islet
9
equivalents.
10 Now, you can plot this in a
11
different way. Now, let's
consider dealing
12
with a low islet, no number of viable islet
13
equivalents, 68 per mouse. This
corresponds
14
to 2500 viable equivalents per kilogram in a
15
Type 1 diabetes recipient, a low number.
16 You will be able to reverse
17
diabetes in all recipients if the fractional
18
viability is 85 percent. If,
however, the
19
fractional viability drops to 15 percent,
20 diabetes reversal is not possible, even if
21
the number of islet equivalents transplanted
22
is almost ten times higher.
185
1 Basically, the message is that the
2
sensitivity of this assay with respect to
3
islet potency is much higher compared with
4
the sensitivity of assays based on membrane
5
integrity, like FDA Pi or Atirium
6
promate (?) or Tripan Blue and many others.
7 Now, we have analyzed results in
8
rat, porcine and human systems.
The
9
fractional viability based on live/dead
10
staining was always between 80 and 100
11
percent, suggesting very good viability.
12 However, based on OCR per DNA
13
assessments, the same tissue may have a
14
viability of 5 percent or 100 percent,
15 clearly indicating that
the sensitivity of
16
this assay is significantly different.
17 We have started to analyze our
18
results obtained with porcine islets and
19
with isolated human islets, and the very
20
same principles apply. If you
transplant an
21
islet preparation with a fairly good
22
fractional viability, let's say 300
186
1
nanomoles oxygen per minute per milligram
2
DNA, you will be able to reverse diabetes
3
with a low islet mass. If the
fractional
4
viability drops to 100, you cannot reverse
5
diabetes with any islet mass possible.
6 Now, we have also analyzed other
7
measures, and I will very briefly mention
8
two. ATP. Now if you address the
9
fractional viability based on ATP per DNA
10
and you look at 1,000 human islet
11
equivalents and you ask the question
12
whether 1,000 human islet equivalents with a
13
high ATP content, that is 110 or higher
14
picomole per microgram DNA, will it reverse
15
diabetes, and the answer is yes in almost 80
16
percent of all animals, as opposed to 15
17
to 20 percent if the ATP per DNA content is
18
less than 110.
19 We also asked whether indication
20
of stress or markers that may predict
21
apoptosis could be helpful in predicting
22
outcomes. Here, we look at human
islets.
187
1
If the P-38 kinase activity is low, reversal
2
of diabetes is a consistent finding, both in
3
human islets and in porcine islets.
And
4
with higher P-38 kinase activity, diabetes
5
reversal is less likely.
6 So what we propose is a new
7
approach that would take into consideration
8
the oxygen consumption rate in a perfect
9
setting, the beta cell oxygen consumption
10
rate, expressed as a nanomole per minute per
11
kilogram body weight, indicating the viable
12
islet mass transplanted, and also the
13
fractional viability defined by OCR per DNA.
14 Now if one would have 4,000 islet
15
equivalents available per kilogram body
16
weight, it would be safe to release this
17
product for transplantation as long as your
18
fractional viability is 150 or higher.
So a
19
release for transplant, high likelihood of
20
success.
21 Then there is an intermediary zone
22
where additional and alternative assays
188
1
should be considered, and if the fractional
2
viability is below 125, no islet dose will
3
reverse diabetes in Type 1 diabetic
4
recipients, and therefore,
this other
5
product should not be released for
6
transplant.
7 Can we implement this in a
8
real-life setting? Now, islets
are prepared
9
at day minus 2 relative to transplant on day
10
zero. What we suggest is
determining the
11
islet yield in terms of islet equivalents,
12
as we have done for many years, but also
13
obtain information on OCR and OCR per DNA
14
and the packed tissue volume.
15 Now, this will determine whether a
16
patient is asked to come and whether you
17
would admit a patient for transplantation.
18
Now, the next morning, after overnight
19
culture, starting at 7:00 a.m., we determine
20
the cellular composition. This
will provide
21
information on islet identity, on beta cell
22
identity, beta cell mass and also purity.
189
1
This assay is completed in four hours.
2 Then we will perform again OCR per
3
DNA measurements; OCR measurements can be
4
completed in 15 minutes, and DNA may take
5
two hours. But we have new
assays now
6
available based on nuclei counting using
7
flow cytometers, which can provide the same
8 information basically in minutes of the ATP
9
per DNA, which is a readout of potency.
10
This will really determine whether we should
11
initiate immunosuppressive treatment.
12 Then, on the morning of the
13
transplant, and the transplant is typically
14
performed around noon, so in the morning, we
15
will now obtain information again, OCR, OCR
16
per DNA, and then also obtain samples from
17
the final product to address safety issues.
18 This is the approach we like to
19
propose. Increasing evidence
indicates
20
close correlation between the proposed
21
potency assay and post-transplant rat,
22
porcine and human islet function.
190
1 Robust clinical data is being
2
generated at the Universities of Minnesota,
3
Alberta, Miami and other sites participating
4
in the NIH/NCRR Islet Cell Resource Program
5
Initiative.
6 We believe it is fair to say that
7
predictive islet potency testing is within
8
reach for clinical implementation.
9 I would like to acknowledge the
10
support provided by the Juvenile Diabetes
11
Research Foundation, the Immune Tolerance
12 Network, NIAID,
NIDDK and NCRR. I would
13
also like to acknowledge the instrumental
14
contributions made by both Glycos Pappas and
15
Clark Colton at MIT, and Jack O'Neil, Maria
16
Cremander (?) and Gordon Weir at the Joslin
17
Diabetes Center and Mass. General Hospital,
18
and the support provided by the staff of our
19
Diabetes Institute for Immunology and
20
Transplantation.
21
Thank you for your
attention.
22 DR. RAO: Any
questions?
191
1 DR. BLAZAR:
Bernard, you did a
2
nice job outlining the potential potency
3
assays you could use. I guess
this will
4
require, obviously, validation on clinical
5
data for multiple sites.
6 My question relates in part to as
7
you change the immunosuppressive protocols,
8
if you have these strict criteria for
9
cut-offs with potency, these are all based
10
upon standards being conducted at these
11
centers.
12
But there are
obviously other
13
centers involved in islet cell
14
transplantation that may be using different
15
protocols, et cetera.
16 How much flexibility should there
17 be in using these
potency assays as you
18
change the immunosuppressive protocols which
19
may result in different criteria for cure
20
rates?
21 DR. HERING:
I agree this is an
22 important
consideration, and the ability to
192
1
reverse diabetes is really determined by a
2
number of factors; islet potency being
3
perhaps the most important
one. But insulin
4
resistance, insulin action on the recipient
5
and the use of diabetogenic
6
immunosuppressive agents clearly, on the
7
side that you chose, are clearly relevant as
8
well.
9 At this point in time, most
10
centers, if not all, perform intrapo (?)
11
islet transplants. Most centers,
if not
12
all, use a corticosteroid, free
13 immunosuppressive
protocols, and an
14
increasing number of centers are using also
15
caltinew (?) inhibitor free protocols.
16 So you are no longer dealing with
17
diabetogenic side effects
18
post-transplantation. You are
still dealing
19
with a possible insulin resistance state
20
that could be corrected pre-transplantation,
21
insulin resistance due to hyperglycemia.
22
I think we will
learn to address
193
1
these issues, but I think the message is we
2
can measure potency at a level of
3
sophistication that was not possible before,
4
and it does not take hours and hours and
5
days to have results available.
Those are
6
real-time assays.
7 I think there should be a
8
threshold that clearly defines whether you
9
will be able to reverse diabetes regardless
10
of immunosuppression, yes or no.
There is
11
clearly a cut-off, and you should not
12
release this product for transplantation.
13 DR. BLAZAR:
I guess that's part
14
of my question. Is this the
state of the
15
art in the field such that these are going
16
to be fixed potency level assays that
17
absolutely have to be applied to all lot
18
release criteria at all sites, or are you
19
still collecting clinical data to know under
20
what venues certain potency assays in lot
21
release criteria should be applied?
22 I couldn't quite get the flavor as
194
1
to whether you are really ready to say that
2
these were the strict --
3 DR. HERING:
I think it is
4
correct. This is not
definitive. I think
5
this is an assay that we propose, and we
6
would like to continue collecting
7
information on this assay.
8 What is very compelling at this
9
point in time is information obtained in
10
small animal models, and we have yet to
11
document that this is a consistent finding
12
in the clinical setting as well.
13 I agree, this is not completely
14
ready, but I have confidence in our ability
15
to show this at some point in time.
16 DR. RAO: Dr.
Ricordi?
17 DR.
RICORDI: Just a comment.
18
This is to show that we are fully prepared
19
to go to any level of complexity in the
20
definition of the potency test as it may be
21
requested by FDA. But at the
same time, we
22
do hope that there will be a simpler way to
195
1
demonstrate the efficacy, as we did like in
2
the last 60 transplants, the genetic success
3
or results after transplantation.
4 We didn't apply necessarily this
5
level of complexity. Maybe if it
will be
6
sufficient to show viability maybe combined
7
with the ÄÄÄÄ evaluation to make it more
8
significant, there could be an intermediate
9
step that is closer to what we do for a
10
hematopoietic product that will not require
11
the same level of complexity.
12 DR. O'FALLON:
My interpretation
13
of the figures suggested that most of the
14
numbers that I saw on the screen, most of
15
the points that I saw on the screen were
16 based on very small
numbers of animals that
17
had experiences that were either positive or
18
negative.
19 So I would certainly argue that
20
you are not ready for making definitive
21
statements. Much of the way you
phrased
22
your findings, I think, were perhaps a
196
1
little orally optimistic in what they might
2 be if you had put
confidence intervals, for
3
example; if three out of three failed,
4
that's hardly a definitive piece of evidence
5
that failure is a certainty in 300, or
6
something like that.
7 This is the direction that has to
8
be taken, but I think we haven't reached
9
home plate quite yet. We may not
even be
10
past first base.
11 DR. RAO: Dr.
High.
12 DR. HIGH: I
wanted to ask a
13
question about potency assay
14
reproducibility. So can I start
by asking
15
out of a lot or a prep, what's the absolute
16
number of cells that are assayed in the
17
potency assay and what's the percent of the
18
prep that is assayed in the potency assay?
19 If you took, you know, five
20
different aliquots of cells to do the
21 potency assay on,
do you get highly
22
reproducible data in the OCR and the other
197
1
assays?
2 DR. HERING:
Yes, I think with
3
this diabetic nude mouse bioassay, the
4
intra-assay variability is very low.
I
5
think the first question that you asked is
6
what is the proportion of islets that we use
7
for quality testing. It is in
the order one
8
to five percent of the final product will be
9
used to islet product testing.
10 DR. HIGH: Is
the prep
11
sufficiently homogeneous that that's always
12
representative of --
13 DR. HERING:
This is the case. I
14
think sampling errors always offer concern,
15
but I think we do typically run assays at
16
least in triplicate, and I think, again,
17
intra-assay variability is at a lower level.
18 DR. MULLIGAN:
I have a question
19
about whether or not the ability of the
20
cells to effectively transplant can be a
21 completely
independent parameter from their
22
functional behavior in vitro.
198
1 I like the OCR/DNA test. I think
2
that's an excellent test. But
then the way
3
you do the kind of measure of success in
4
transplantation is essentially you set it
5
that, you know, if you go above a certain
6
number of cells being transplanted, all the
7
animals do okay.
8 The question is, if you don't set
9
the dose that high, reduce the dose, not all
10
the animals get there. Doesn't
that suggest
11
that although they may be functional,
12
potentially there might be some other
13
characteristic of the cells that would
14
influence whether or not they take at the
15
same efficiency, they galvanize (?) the
16 right niche
effectively.
17 So it seems to me you need some
18
sort of potency. So I would say
actually
19
the OCR/DNA test, I would call it more a
20
functional test. A potency test
would be
21
something that really tries to look at how
22
reproduceful a certain number of cells will
199
1
give you normal glycemia at a certain time
2
after transplantation.
3 I know ultimately you want to set
4
something such that everyone will do okay.
5
But isn't that an important way to look at
6
it; that, you know, you could have preps
7
that differ very significantly in their
8
transplantability potential, and you'd never
9
really assess that if you just used a dosage
10
of cells that was up high enough that, you
11
know, whatever variation you have from prep
12
to prep, you can always make sure you have
13
sufficiency.
14 DR. HERING:
I think what is
15
really important to emphasize is what we
16
want to avoid is primary non-function in
17
recipients. Based on product
testing, it
18
was not as sophisticated as the one that we
19
proposed for further investigation.
20
As Dr. Ricordi pointed
out, we did
21
see primary non-function in about one
22
percent of our recipients. That
was already
200
1
a very acceptable outcome, I would think.
2 So now we can take it, and the
3
results need to be analyzed, at least to the
4
same level. I think this would
be an
5
acceptable achievement. I agree
with you,
6
we need to improve the efficiency of the
7
process.
8 There are many more lessons to be
9
learned in how we can make effective use of
10
islets from expanded or marginal donors.
11
This will clearly become a topic, because we
12
have to learn how to make this treatment
13
available to more patients, how can we
14
maximize utilization of suboptimal donors,
15
and then questions that you raised become
16
critically important, because maybe then,
17
assays that we have here may not be
18
sufficient and we have to look at cell
19
injury. We have to look at cell
stress. We
20
have to look at reversibility of our
21
findings and we have to look at kinetics and
22
so forth.
201
1 I think this
is meant to be a
2
real-time potency assay that can be
3
performed at every single institution.
4
Results are available.
Investments are
5
absolutely minor. This would
provide an
6 assay really for transplant
centers that
7
have an interest in providing islet
8
transplantation for patient care purposes
9
under license.
10 This may not meet the standards
11 for clinical investigation
and clinical
12
research, and the questions may be entirely
13
different.
14 DR. MULLIGAN:
So the time to
15
normal glycemia, I assume that there are
16 doses in the animal
systems that will give
17
you ultimately normal glycemia, but it will
18
take a different amount of time to get to
19
that point. That's the dosage,
right? The
20
longer it takes, I guess that means that you
21
have less of the cells. Is that
what you
22
think?
202
1 DR. HERING:
I have not shown this
2
data. We have done numerous
transplants in
3
immuno-incompetent mice, and we have done
4
dose responses. We have
transplanted close
5
to 100 consecutive human islet preps.
Seven
6
hundred and fifty, one thousand and two
7
thousand are the equivalents per mouse.
8 Clearly, the time to normal
9
glycemia was dependent on the cell dose.
10
There was a clear-cut relationship.
I
11
couldn't show all the data, but I take your
12
point. This is time to normal
glycemia
13
using a given dose of islets.
It's very
14
important. What we anticipate
that we
15
should do, we should define precisely the
16
cell dose that is infused in such a mouse
17
model, not only based on islet numbers, but
18
based on beta cell numbers.
19 Then we can analyze whether there
20
is an impact of OCR or ATP or any other
21
measures. Then we can further
validate the
22
assays. But we keep the beta
cell number
203
1 within a very
narrow limit.
2 DR. MULLIGAN:
I think for where
3
you are at this point, that's very, very
4
sensible. But I was thinking way
ahead;
5
that as this begins to be a clinical
6
practice, if the precise characteristics of
7
glucose sensing are going to be important
8
therapeutically in the long run, which
9
probably will be the case, then you want to
10
begin to think early of whether
or not
11
indeed there's a way to measure the actual
12
absolute numbers of functioning islet cells
13
in a transplant recipient, so you can see
14
whether the absolute number of those cells,
15
which would be different than what a normal
16
pancreas has, will influence the way sensing
17
does occur.
18 DR. SALOMON:
My concern here is
19
that the discussion may tend to confuse two
20
very important things. The first
is that
21
the field needs a potency assay that can be
22
used in a sense as a lot release criteria.
204
1 Then a second issue is going to
2
be, based on this potency assay, what is the
3
prediction that an islet preparation
4
released will cure diabetes in the patient?
5 I would say that both are
6
critically important, but they are very
7
different. This conversation has
tended to
8
confuse the two, I think.
9 The point that Bernhard wants to
10
make, and I think he's shown us a very good
11
proof of principle, Dr. Fallon's comments
12
about numbers and statistical significance
13
are very well-taken, but as a proof of
14 principle, here
he's showing that you can
15
develop lot release criteria that at least
16
demonstrates the physiological functionality
17
of an islet preparation based on curing
18
diabetes in several different animal models,
19
both nonhuman, primate and small model
20
murine.
21 I think that's what a potency
22
assay should be defined as.
Certainly
205
1
someone in a big drug company making an
2
anti-hypertensive drug doesn't want to hold
3
up the lot release criteria until he cures
4
the hypertension on 2500 patients.
5 I think that's not unfair, to ask
6
for the same sort of criteria for islet
7
transplantation. So certainly
after you've
8
done 150, 200, 1,000 islet transplants, you
9 go back and look at
the data and begin to
10
ask some of the questions.
11 I think Richard was good at
12
pointing out how there may be a transition
13
to going back to some of these data and
14
refining the potency assays.
15 I think it's really important,
16
because Bruce, when you start talking about
17
immunosuppression, that's not got anything
18
to do with this part of the conversation.
19
It's important, but it's not part of a
20
potency assay. Because if you
don't
21
transplant a functional mass of islets, then
22
you don't get function. It
doesn't have
206
1
anything to do with the immune response.
2 DR. BLAZAR:
The potency assays
3
are not directly measuring glucose.
It's
4 oxygen
consumption. So the presumption here
5
is that that measurement relates to curing
6
animals and in humans, and that therefore,
7
it's a valid criteria to use to say that
8
this is a good preparation.
9 It's not really measuring glucose
10
metabolism by this OCR assay. So
then it
11
becomes a surrogate assay that is meant to
12
correlate with clinical outcomes.
So I
13 think even the
assay itself brings up your
14
point, but it is confusing because it's a
15
surrogate.
16 DR. SALOMON:
I just want to point
17
out that what I was referring to, Bruce, is
18
what we want to say is if I transplant an
19
islet mass of a defined functionality based
20
on refining the kinds of things that Camillo
21
and Bernhard and James have been working on
22
in their three big institutions,
then that
207
1
is where you begin to define the quality and
2
success of my immunosuppressive regimens.
3 That was kind
of what I was
4
talking about earlier; this is where you
5
transition from transplantation into product
6
and then back to transplantation.
We have
7
to be really clear where those transitions
8
occur.
9 DR. RAO: A
point well-taken. I
10
want to add a couple of questions to this
11
and then maybe we can go on to that.
12 One thing that seemed to become
13
clear from your presentation was that dead
14
cells are bad, right; even if you have the
15
same number of viable islets, you still
16
don't get the same end result if the total
17 number of dead cells present or some
18
non-viable cells is there.
19 To me, that implies that the
20
OCR/DNA test might be a better marker than
21
using a viability assay which uses dyes, for
22
example. You alluded to
that. In your
208
1
opinion, do you think that that's absolutely
2
true, that one could not or should not use a
3
simpler test, or that's too strong a
4
generalization based on the amount of data
5
that's present, taking into account the
6
small "n" in testing, et cetera?
7 DR. HERING: One
slide showed the
8
comparison, fractional viability as assessed
9
by OCR for DNA and dye exclusion tests,
10
live/dead staining assays.
11 I think the lesson that I took
12
from this state is that really the
13
sensitivity is remarkably better with OCR
14
per DNA. If I take into
consideration that
15
it takes 15 minutes to obtain this
16
information, and that the investments are
17
really fairly minimal, I think I would
18
prefer this assay, and would think that one
19
should definitely include this in the
20
assessment.
21 I think when we look at islet
22
transplant outcomes in animal models and
209
1
clinical transplants around the world and we
2
look at the registry results, live/dead
3 staining or tripad
blue or even
4
glucose-stimulated insulin release have
5
never predicted outcome in any one setting
6
anywhere around the world.
7 That is why I did not present
8
this. So this is what we know
already.
9
That is why we understand that we have to
10
develop and standardize and implement better
11
assays with better sensitivity and better
12
specificity as well.
13 We understand we have not
14
completed our homework, but we believe we
15
should continue and should collect the
16
information and really analyze this.
But I
17
think the preliminary data seem to suggest
18
that this is a much more sensitive assay.
19 DR. RAO: Can
I extend that
20
question to sort of follow up on what Bruce
21
said? Is this criterion in some
sense too
22
rigid? That means it would
change things if
210
1
it turned out that protocols increased or
2
that, you know, the disassociation protocol
3
that you used for cells would be better, or
4
if some other institute had better
5
immunesuppression protocols so that the
6
viable mass, with an end readout which is
7
really, you know, normalizing glucose, would
8
be different; would this in some sense alter
9
by having some kind of absolute criteria?
10 DR. HERING:
My approach would be
11
the following: I would like to
use the most
12
sensitive assay, but the specification range
13
could be discussed, so that at least you can
14
obtain and collect valuable information that
15
you would otherwise not obtain with an assay
16
that has more or less zero sensitivity.
17 So I think since we really don't
18
know the lower limit, I would agree with
19
you, the specification range should then be
20 defined in view of
the fact that we don't
21
know the answer yet, but at least then we
22
would obtain very meaningful clinical
211
1
information, and we'll really learn to call
2
it outcomes in the assay with outcomes in
3
patients.
4 DR. RICORDI:
Just a comment that
5
this kind of testing, if you consider the
6 correlation between the OCR or any of this
7
other testing with the reverse set of
8
diabetes in the nude mouse model, in which
9
we don't have immunosuppression, that that
10
would exclude the confusing variable with
11
the immunosuppressive arrangement could be
12
more or less diabetogenic, because there you
13
have no drug as a starting point.
14 But I would be very concerned. I
15 would hope that
the discussion could be more
16
like since we have three institution culture
17
islets, 60 cases of successful transplant
18
responding to defined criteria for product
19
release, shouldn't this be considered a
20
fairly good readout, the fact that you
21
reversed diabetes or you have a definite
22
effect after transplantation?
212
1 Bernhard didn't mention, but to
2
have the negative results, he has to kill
3
the human islets with cytokines, or exposing
4
them, because it is actually in a way a more
5 artificial system,
the way we are using,
6
trying to demonstrate.
7 It is a real challenge. What you
8
want to do is how you exclude the
9
preparation that past the point of no
10 return, that we
never recover in vivo and be
11
able to have a definite positive biologic
12
function.
13 DR. NOGUCHI:
I just wanted to say
14
we appreciate this very discussion and
15
getting at the heart of the matters for how
16
we can make progress.
17 I'd just point out, one of the
18
purposes of the potency test, if we go back
19
to the original 1902 law, in the legislative
20
history, part of this is to assist the
21
person, the physician who would be
22
administering the -- we will call it product
213
1
at this point.
2 A potency assay is meant to be
3
used for many things. One of
them is we
4
hope you will be able to come up eventually
5
with a minimal number such that above this,
6
we believe, is a useful thing.
7 But it's also used by the
8
clinician in terms of how that product is
9
being used. A person may be a
little bit
10
sicker or your potency may be a little less
11
than what you expect.
12 In the case where you have enough
13
to actually use, you might use a little bit
14
more, or you may be more careful on how you
15 do other parts of the
treatment such that
16
you may predict that it will take longer to
17
engraft.
18 We should not be looking at
19
potency as being some magic; if you don't
20 hit this
number, we won't let you use it.
21
It is true, this is transplantation.
Some
22
of the concerns we would not allow in other
214
1 settings, we will
allow here.
2 The donor screen, as an example,
3
you may get to the point where you want to
4
be able to transplant to HIV-positive
5
individuals HIV-positive islets.
We would
6
not object to that because that's a type of
7
situation for which human tissues are being
8
used.
9 So think of this as being very
10
rigorous kinds of ideals to approach.
We do
11
realize transplantation is different.
If
12
you need to transplant something in the case
13
of a bone marrow, and you have nothing else
14
to use, even if they are 90 percent dead,
15
you will still use it in an attempt to
16
effect a favorable outcome.
17 We aren't unreasonable, but we are
18
saying let's make sure we do the very best
19
collectively together as we move forward.
20 But it is true, the patient
21
dictates as well as the physician treating.
22 DR. RAO: Mr.
O'Neill.
215
1 DR. O'NEIL:
Just one comment on
2
the viability assay, the live/dead staining.
3
One thing to appreciate is that is a very
4
subjective measure of viability.
It's
5
usually performed by an individual under
6
observation under a fluorescent microscope,
7
and it's really a very subjective
8
determination of viability, where the OCR is
9
very compelling data and it's very
10
quantitative.
11 The question I have, Bernhard, is
12
although we would all appreciate
13
standardization throughout assays and
14
throughout centers, is it realistic to
15 expect every
center to perform the same
16
assay, or would you leave it open to -- you
17
presented data on OCR, ATP and P-38 kinase
18
inhibitor as all correlating to subsequent
19
function and transplant setting.
20 Would it be up to an individual
21
center to determine their own potency assay
22
and then validate it basically the way you
216
1
have with OCR?
2 DR. HERING:
I think this is a
3
question that Dr. Noguchi probably could
4
answer much better than I. I
think it is
5
conceivable that you would submit a BLA
6
application and you would really indicate
7
what type of potency assets you have
8
validated in your facility.
9 Then you have compelling evidence
10
suggesting that this is all valid and very
11
well worked out. This would
probably be
12
sufficient and meet the requirements of 21
13
CFR 600.3, a reasonable measure of product
14
potency.
15 I think if
you have this
16
information and you use different assays, I
17
would think this would address the point.
18 DR. SALOMON:
I guess my comment
19
to that would be, you certainly don't want
20
to do anything to inhibit innovation.
I
21
don't mean my comments to be taken in that
22
respect. But I would think at
this early
217
1
time, particularly given the very clear
2
evidence we have at this point, it is very
3
center-dependent and that there's a big
4
learning curve for any center wanting to get
5
into this, that I would argue for a
6
relatively narrow group of well-accepted
7
potency measurements.
8 Then everybody can innovate around
9
it and you can define how you would validate
10 this or that new
potency measurement. I
11
think everyone, you know, would want to see
12
that happen as the days turn into years.
13 But I think at the starting point
14
here, we ought to have a very specific
15
group, not that every center coming down the
16
pike has their own.
17 DR. O'NEIL:
Oh, I agree with you
18
absolutely. One of the concerns
is even for
19
something like DNA assays or insulin assays
20
or immune histochemistry, those same assays
21
done at different centers, even if they are
22
following the same procedure, can sometimes
218
1
lead to very disparate results.
2 So the interpretation of the data
3
sometimes is subject to the individual
4
center that's performing all the tests and
5
may not transfer directly from one to
6
another. But I do agree
standardization is
7
the way that we would all like to go.
8 DR. RAO:
Well, if there are no
9
more questions, thank you.
10 MS. LAWTON:
Can I just ask one
11
question in follow-up to that discussion?
12
Earlier on, you were asked a question about
13
assay variability, and you talked about
14
intra-assay variability.
15 But I wanted to know whether you
16
have done actually any studies on
17
inter-assay variability, for that very
18
reason.
19 DR.
HERING: Well, I think more
20
work needs to be done, but if you're talking
21
about inter-assay, I think since we keep
22
islets for so many days in culture, then you
219
1
would also need to address the issue of
2
product stability, which could be a
3
confounding factor when you like to assess
4
inter-assay variability, because there may
5
be a time when the product is no longer
6
stable.
7 Of course, you could set it up
8
such that you could test inter-assay at the
9
very same time during one day.
10 I must admit this information is
11
not complete. I think this
presentation was
12
meant to indicate that we have a concept
13
that we'd like to pursue, and we understand
14
that more work needs to be done.
15 DR. RAO:
Thank you. We now move
16
into the public phase of this meeting.
We
17
have two people who asked to be recognized.
18
It's Dr. Michael Millis from the University
19
of Chicago, and Dr. Robert Goldstein from
20
JDRF. They will each speak for
about five
21
minutes.
22 Before we begin that, I need to
220
1
read our open public hearing announcement
2
for general matters meetings.
Both the Food
3
and Drug Administration and the public
4
believe in a transparent process for
5
information-gathering and decision-making.
6 To ensure such transparency at the
7
open public hearing session of the Advisory
8
Committee meeting, FDA believes that it is
9
important to understand the context of an
10
individual's presentation.
11 For this reason, FDA encourages
12
you, the open public hearing speaker, at the
13
beginning of your written or oral statement,
14
to advise the committee of any financial
15
relationship that you may have with any
16
company or any group that is likely to be
17
impacted by the topic of this meeting.
18 For example, the financial
19
information may include the company's or
20
group's payment of your travel, lodging or
21
other expenses in connection with your
22
attendance at the meeting.
221
1 Likewise, FDA encourages you at
2
the beginning of your statement to advise
3
the committee if you do not have any such
4
financial relationships.
5 If you chose not to address this
6
issue of financial relationships at the
7
beginning of your statement, it will not
8
preclude you from speaking.
9 MR.
MILLIS: I'm Michael Millis.
10
I'm representing the University of Chicago,
11
and to a far lesser extent, a company that
12
was developed to help move some of the
13
technology to help other centers.
14 From the University of Chicago, I
15
am the section chief in transplantation. I
16
am not the PI on any islet transplant
17
protocol. The section is very
interested in
18
pursuing islet transplantation, so I do have
19
an interest in that that way.
20 I would first, before I do
21
anything, like to thank the big three over
22
here who have really pushed this whole field
222
1
forward. In no way are any of my
comments
2
to anything but help with the progress that
3
they have made.
4 We came, at
the University of
5
Chicago, to this issue a little bit
6
differently than the big three and others
7
have done in the sense that of the
8
scientific movement towards; and that was we
9
wanted to build a facility that provided
10
multiple users the access for cellular- and
11
tissue-based products.
12 Within that framework, we wanted
13
to help facilitate islet transplantation.
14
As the section chief in transplantation, and
15
ultimately the director of this facility, I
16
wanted it to be flexible enough to provide
17
the islet transplant group a facility to
18 work as well as
other investigators such as
19
cellular oncologic investigations, et
20
cetera.
21 But it had to obviously be
22
practical from an academic institution's
223
1
financial constraints. It was
particularly
2
noted that the comments this morning about
3
can an academic institution do this.
We are
4 in the process. I don't
have a definitive
5
answer to that yet, but we are trying.
6 So we saw really what we were at
7
here at the time a couple of years ago, a
8
lot of open bridges that had very little
9
ways of crossing that we could see.
We had
10
all the protocols that were being
11
anticipated based on standard laboratory
12
practice, the reminder letter that's been
13
referred to from the FDA a couple of times,
14
and then in March of 2000, a guidance
15
regarding the validation of processing for
16
human tissues for transplantation.
17 What we wanted to get to is there
18
on the left side of the slide, a procedure
19
that encompassed CGMP for the manufacturer
20
of biologics that would be available to
21
provide the support for BLAs, which would
22 ultimately help in
the reimbursement.
224
1 We decided what we needed to do
2
since none of us at the institution, an
3
academic institution, had experience in
4
this, we hired consultants, and through that
5
consulting arrangement, a company, CGMP, was
6
formed to help move some of the technology
7
that we were developing, that had already
8
been developed, but actually appeared to be
9
commercially acceptable forward.
10 So the principles of biologic
11
manufacturing that we wanted to focus on
12
were control and validation of
facility and
13
the process, and the goal specifically in
14
regards to islet transplantation was to have
15
it conform to the principles of biologic
16
manufacturing.
17 Early
on, we thought that it was
18
critically important to match the existing
19
assets of the institution to the anticipated
20
regulatory compliance. We needed
to form a
21
good plan that matched what we had and what
22
we wanted to get to and what we thought the
225
1
FDA was going to require by the time we got
2
there.
3 We wanted to make sure that the
4
processes that were being defined were
5
robust enough that whatever the process was
6
two years, five years from now, would be
7
able to be accomplished in the facility, and
8
those processes would necessarily need to be
9
validated. We had to make sure
that we
10
could control costs.
11 That was obviously part of the
12
multi-user concept of the facility.
So we
13
wanted to control the critical processes of
14
the environment that the cells and tissue
15
were exposed to, temperature, bio-burden and
16
cross-contamination, eliminate all the
17
equipment that couldn't be validated to
18
conform to those above three aspects, and to
19
eliminate the components that required the
20
use of undesirable substrates, and
21
obviously, training of the staff that was
22
mentioned earlier.
226
1 At this point, we wanted to match
2
the processes that were wanting to be done
3
with the facility that we wanted to build.
4 So this is the 2991 that has been
5
mentioned before that's been used in most,
6
if not all, of islet isolations that have
7
led to transplantation.
8 It is well-known that this machine
9
doesn't come with refrigeration, and we felt
10
that it was critical to supply that
11
refrigeration as is done in the ten ITN
12
centers as well as the big three that are
13
sitting here. That was not
commercially
14
available at the time, and so we worked with
15
Gambro, the maker of the machine 2991, to be
16
able to provide that service for centers
17
that want a refrigerated COBE.
That's why
18
that company, CGMP, had to be made, so that
19
we could provide that for centers that
20
wanted to control the environment at
21
whatever temperature.
22 So this is another aspect that,
227
1
when we were analyzing the process, we
2
thought was a bit of a problem.
This is a
3
coil that changes the temperature of the
4
enzyme so that it's active at 37 degrees and
5
then becomes inactive at cooler
6
temperatures. The hosing there
is connected
7
to Dr. Ricordi's chamber.
8 The temperature of the enzyme is
9
controlled by the coil going up and down in
10
the cool water bath. With regard
to the
11 facility that we
wanted to make and the
12
process that we wanted to have, we felt that
13
it was going to be impossible to validate
14
that there wasn't some biologic waste left
15
in any of these coils as product is going
16
through this, and in addition, we didn't
17
want any water baths in the facility because
18
of the possibility of contamination in
19
water, et cetera.
20 So instead,
we started looking at
21
a blood warming bag that the product could
22
be put through instead of the coil to heat
228
1
up the enzyme, and used the manufacturer's
2
heating unit for that.
3 In order to cool it, we had to
4
design a cooling chamber since there wasn't
5
a blood cooling chamber. We had
to design
6 one that once
again that worked off of the
7
chiller, the same chiller that the COBE 2991
8
uses in order to cool it down, and were able
9
to show that the process, now using
10
disposable, essentially coil that we can
11
throw out after each isolation, provides the
12
same temperature regulation as the coil that
13
has been used many, many times.
There's
14
Dr. Ricordi's chamber that's hooked up to
15
it.
16 So in summary, we are just
17
beginning this; we certainly don't have the
18
experience that the big three or ITN or IND
19
just got approved a couple of weeks ago.
20
But the evidence that we have from the
21
isolations that we have performed is that
22
the islets performed well.
229
1 I think
moving the process from
2
the benchtop to Phase 1, Phase 2, Phase 3 as
3
was noted before with regards to GCMP and
4
BLA, moving that really requires the
5
knowledge of the pharmaceutical standards
6
and laboratory process development that our
7
consultants were able to provide to us in
8
order to get it moved towards a BLA-type of
9
atmosphere.
10 So that's all, I guess. I'll take
11
any questions.
12 DR. RAO:
Thank you. Our next
13
speaker will be Dr. Goldstein.
14 DR. GOLDSTEIN:
Mr. Chairman, FDA
15
people and everybody else, the Juvenile
16
Diabetes Research Foundation was
17
founded 30-plus years ago by parents of
18
children. In June, we had 200 of
those
19
children march on Washington in something we
20
call Childrens Congress. We resisted the
21
temptation to bring them here today.
22 But the point actually reflects on
230
1
Dr. Ricordi's introduction of where the
2
field started. JDRF was an early
supporter
3
of research for islet work; in fact
4
supported Dr. Lacy. It stuck to
it through
5
those dim years when the results were not so
6
terrific.
7 So what we are here in today is a
8
supportive role, to applaud the effort to
9
move the field from the basic research into
10
the clinic, because this has, we think,
11
demonstrated efficacy in this select group
12
of individuals with Type 1 diabetes who have
13
simply horrible lives that can be reversed,
14
and these people can be helped
15
significantly.
16 So we feel that while imperfect
17
though it may be, it's certainly currently,
18
for these people, an extremely important
19
therapy. Even if we could only
take
20
advantage of the few
pancreases available,
21
we would help hundreds of people who
22
otherwise have no particular help.
231
1 As we go forward, we recognize
2
that this will be the platform for cell
3
therapy of the future as we see alternative
4
versions of islets or alternative sources.
5 So I only brought three slides.
6
The point to this slide was a little
7
tracking on our part and crawling around,
8
simply to put up on a slide and say, people
9
all over the world are doing this.
10 Now, I don't have all the
11
spectacular results, and obviously they are
12
going to be very varied. But I
do want to
13
make a point: We support this
work in the
14
international community. And for
those of
15
you who may not know this, the FDA has been
16
intimately involved in regulating -- maybe
17
regulating is the wrong word -- when they go
18
to Italy or Switzerland or Germany, maybe
19
giving solid advice about quality is the
20
word.
21 The point I would make is that
22
your activities here will actually define
232
1 the regulatory
environment in the rest of
2
the world for this topic, because we have
3
made every effort through our funding to
4
impose uniformity. We don't want
to support
5
people doing wild and crazy things.
6 There's no other regulatory agency
7
that we are aware of that actually is as
8
interested as the FDA in making this happen
9
and coming up with guidelines and quality
10
controls. So we need that in
addition to
11
the efforts in the United States.
12 This was just to emphasize the
13
areas that we have provided research support
14
in concert with the NIH. In the
several
15
years, three to five years, that we have had
16
significant resources to do this in concert
17
with the NIH and actually private
18
organizations as well, there are many, many
19
activities throughout the United States as
20
well as throughout the world, and people are
21
trying to add on to this weekly.
22 Just to emphasize it, we have
233
1
actually seen this many transplants, and if
2
you follow the field of organ
3
transplantation in general terms and think
4
about those few people who benefited from
5
heart transplants early, or those few people
6
who have benefited from lung transplants
7
early, it may require a concerted effort on
8
all our parts to make sure that a therapy so
9
precious as this lives and flourishes in a
10
positive, constructive regulatory
11
environment.
12 We appreciate your efforts. Thank
13
you.
14 DR. RAO:
Thank you,
15
Dr. Goldstein. There is time for
anybody
16
else in the public area who would like to
17
make a comment. I'll ask again
that you
18
identify yourself and any financial
19
conflicts you may have.
20 MR. HARLAN:
I'm David Harlan.
21
I'm at the NIH, and I'm normally a member of
22
this committee. For Dickens
fans, you know
234
1
that he starts "A Christmas Carol" with the
2
words, "Jacob Marley was dead." That fact
3
has to be established before the rest of the
4
story can be told.
5 So I have a
couple of facts to
6
establish before I make my comment.
The
7
first one is that I promised Gail I would
8
emphasize that I'm here speaking from the
9
public, not as a BRMAC member. In fact, I
10
emphasize that by not wearing my uniform
11
today. I'll wear it tomorrow.
12 The second fact is:
Just like
13
what Dr. Goldstein said, I think that we
14
have to recognize that there's a baby in the
15
bath water, that there's really some magic
16
to islet transplantation. I also
wish to
17
echo what Dr. Millis said, that the progress
18
that's been made is due in no small measure
19
to the three people sitting to the right
20
side of the room.
21 The best evidence that I can cite
22
for the miracle of this therapy is two
235
1
patients that are here with me today that
2
had diabetes between the two of them for 60
3
years, and now have enjoyed, between the two
4
of them, four years off insulin.
5 That said, our agenda here is to
6
decide if this therapy is ready for BLA, and
7
regardless of the answer to that question,
8
if the answer is no, what needs to be
9 established before it
is ready for BLA.
10 My contention, the point I wish to
11
make is that based upon the law that we
12
heard about this morning that requires that
13
a process have control over the starting
14
material, that the process itself be
15
standardized, and that you have established
16
criteria for potency, that we are not there
17
yet with this field.
18 Another point
that was made this
19
morning is that what enters into this
20
discussion is that this, unfortunately, this
21
very promising therapy is very, very
22
expensive. I think that we need
to
236
1
recognize that while this is not yet ready
2
for BLA application, it is an expensive,
3
promising therapy. And it may
not be in the
4 purview of this
organization, but somehow,
5
some way needs to be found to see this
6
investigational therapy supported so that it
7
can continue to develop.
8 I'll have further points as this
9
meeting progresses, I'm sure.
I'll
10
continually emphasize that I'm speaking not
11
as a member of BRMAC.
12 DR. RAO:
Thank you, Dr. Harlan.
13 MS. MEYERS:
I think that was a
14
very important point about the cost, because
15
from my perspective, I always think in terms
16
of what will happen when whatever gets on
17
the market and you have to face an insurance
18
company.
19 What I don't understand here is
20
that it seems that the manufacturing of this
21
product is going on in academic
22
institutions. From what I
understood, tell
237
1
me if I'm wrong, this company that was
2
started for the manufacturing at University
3
of Chicago is at the University of Chicago,
4
right?
5 MR. MILLIS:
No, it's not. This
6
is not a company for the manufacture.
This
7
was just a company that was made up so those
8
COBEs could become refrigerated.
That's the
9
only aspect of it.
10 MS. MEYERS:
But you are really
11
not making these cells?
12 MR. MILLIS:
No, no. That's done
13
at the university. Like I say,
it was the
14
university essentially that made us do that
15
to say this is the only way that this aspect
16
can become commercially available to people,
17
is this retrofitted COBE, because Gambro had
18 no desire to
retrofit their COBEs.
19 MS. MEYERS:
But the point I want
20
to make is that in order for people to get
21
reimbursed for this when it does get that
22
far, it has to be made by a commercial
238
1
company, not by an academic institution.
2
Just getting a bill from the academic
3
institution is going to raise questions at
4
the insurance company; was this research or
5
not? So you need to get the
commercial
6
sector involved here, because otherwise, it
7
will never be made available to the average
8
person.
9 DR. RAO: I
don't know if this is
10
the purview of this committee right now in
11
terms of the questions that we should
12
discuss. I mean, it's an
important thing to
13 keep in context, but
I don't think we should
14
be discussing that right now.
15 DR. NOGUCHI:
Just one comment.
16
What we have in this country at the present
17
time, yes, that is the model.
However, if
18
we consider that we do regulate blood, red
19
cells, for transfusion, as an example, we do
20
have both licensed and non-licensed blood
21
banks. It's not without its own
problems,
22
but many of those are at universities as
239
1
well and are distributed by universities and
2
hospitals, which are not necessarily
3 considered to be
pharmaceutical
4
manufacturers.
5 So I think the future discussion
6
and something to keep in mind is that there
7
is no singular way that we see at FDA that
8 it must be
done. We see these as all
9
possibilities. Part of the
discussion is to
10
really to try to grasp what is actually
11
feasible. What can we actually
do to
12
deliver things that work?
13 DR. RAO: I
want to emphasize that
14
if the public wants to make a comment, it
15
should be directed in general rather than as
16
a response to any of the specific
17
discussions that come to the chair.
18 DR. SALOMON:
I just wanted to
19
make a comment to Abbey. I think
that the
20
concern you have that there has to be a
21
commercial buy-in here is not correct. What
22
in general that we're talking about when we
240
1
are talking about it is really reimbursement
2
for the efforts required by these centers to
3
develop the islet product for
4
transplantation. So it's to
purchase the
5
organ, to support the cost of the various
6
reagents, plastic, collagenase enzymes, the
7
purification material, whether you use the
8
Opti Prep or whatever, and manipulating it
9
and doing the assays for sterility and
10
endotoxin content.
11 Everything costs money, as you
12
know. That's what we are talking
about.
13
What we really want to do in the first phase
14
is to be able to go to third-party
15
insurance, Medicare, CMS, and say this is
16
what it costs to do an islet transplant.
17 Right now, there's plenty of
18
precedence for that in multiple different
19
areas, from bone marrow transplantation and
20
preparation, to actually charging a fee for
21
procurement and handling and
shipping of an
22
organ for an organ transplant.
So it's
241
1
okay.
2 DR. RAO: Dr.
Kurtzberg?
3 DR.
KURTZBERG: I want to even go
4
further and say that I think turning it into
5
a commercial operation may be the wrong
6
approach. If the real goal is
providing
7
organ donors and organs for the people who
8
need it, a federally sponsored program
9
probably has more likelihood of being
10
successful at that.
11 DR. RAO: So
on that note,
12
somewhat surprisingly, I find that we are
13
still all on-time. I thank
everyone for
14
making sure that they stayed on-time.
15 We will break for lunch and we
16
will reconvene at 1:20.
17
(Whereupon, at
12:30 p.m., a
18 luncheon recess was taken.)
19
20
21
22
242
1 A F T E R N O O N
S E S S I O N
2 (1:35 p.m.)
3 DR. RAO: I
guess we should get
4
started on the other part of this whole
5
session, which is trying to answer the
6
questions proposed by the FDA.
7 Like the morning, I'm going to ask
8
that people raise their hand to be
9
recognized, and that we just take turns in
10
doing this and try to make sure that
11
everybody participates.
12 People from the audience are
13
welcome to make comments, and they need to
14
be recognized just like other people.
15 If you look
at the list of
16
questions, we are trying to focus mostly on
17
the manufacturing and product quality
18
questions which are posed there.
You should
19
have them in your folders.
20 DR.
SHERWIN: Could I make a
21
comment? I've been on a lot of
review
22
committees over the years. It
just seems to
243
1
me that before addressing these questions, I
2
don't have a good sense of the clinical
3
outcomes.
4 Normally when you evaluate a drug,
5
you evaluate it from top to bottom, from the
6 very beginning up to
the clinical outcomes.
7
Then you get a sense of what the key issues
8
are. I don't have that without
knowing the
9
whole picture, because my assumption is what
10
I'm hearing from people here that everything
11
is great and the clinical outcomes are
12
perfect. I don't know that.
13 I don't have a good feel, even
14
though I'm a diabetologist, that that is
15 exactly what's
going on. So it seems to me
16
that it's very hard to address these issues
17
without looking at what's been achieved so
18
far.
19 I know that's not what you want to
20
hear, but that's my sense.
21 DR. RAO:
Well, part of the issue
22
we are going to try and discuss tomorrow in
244
1
terms of what kind of clinical outcomes will
2
be necessary and important in terms of
3
looking at the cells and what will be
4
measures which will help predict clinical
5
outcome, so some of those issues will be
6
there.
7 On the other hand, it's also
8
important to remember, I think, that what
9
the FDA is trying to look at today, at least
10
from what I understood, is to ask is, if
11
people wanted to use pancreatic islets in
12
terms of transplants, what would be the
13
minimum set of criteria if it were to be
14
regulated as a product, even in terms of
15
just safety that you would have to look at?
16 How would you define a lot in
17
terms of manufacturing so that one could
18
generate standardized clinical data that one
19
might believe?
20 I know it's hard, but it's hard to
21
see how one could do it by first having a
22
lot of clinical data with a non-validated
245
1
set of samples and then try
and extrapolate
2
from that as to what samples may go to it.
3 I think what you said is true, you
4
have to keep that as a caution in terms of
5
we can't make a statement or it can't be
6
definite.
7 This is just going to be in a
8
sense an advice to the FDA that they have to
9
take in account including such issues.
10 DR. BLAZAR:
Cynthia, did you want
11
to make any comments here?
12 DR. RASK:
Yes. This Advisory
13
Committee is a bit different than the usual.
14
It's not for a product approval.
We don't
15 have any data
in-house. It's been submitted
16
and reviewed by FDA.
17 So the purpose of this meeting is
18
not to evaluate any of the data that have
19
been presented, but really to talk in
20
theoretical terms as to what would be
21
required for a product submission to support
22
licensure of this type of product, if that
246
1
helps.
2 MS. LAWTON:
Yes, I just wanted to
3
make a comment on the question, because it
4
asks us to talk about acceptance criteria of
5
these organs. Although we heard
a little
6
bit about that in the presentations, I don't
7
know that we have much information really on
8
those different factors and how they
9
influence the ultimate cells that are
10
transplanted and then what the clinical
11
outcomes are.
12 I do think it's important,
13
however, for us to think about these
14
different things so that in all of these
15
aspects, understanding what measurements are
16
being done, and if we can have some kind of
17
standardized measurements done across, or
18
not necessarily exactly the same methods,
19
but at least similar measurements, that we
20
can then understand how those factors
21
ultimately impact the clinical outcomes
22
would be important.
247
1 DR. RAO: I think that's a real
2
good start. I mean, that's the
first
3
question. In the pancreatic
islet field, it
4
seems that that's one place where there's a
5
clear distinction from other tissues, that
6
you don't have a lot of control on source.
7 It would be nice to hear the
8
committee's comments on that, that since
9
there's not a whole lot of control on
10
source, in general, it's always going to be,
11
at least currently, the criteria that you
12
use discarded pancreata which would not be
13
viable for use as organ transplants.
Is
14
there any specific guidance the committee
15
thinks it can make?
16 DR. LEVITSKY:
We heard presented
17
by Dr. Ricordi some statements made that
18
islets that came from people who had a
19
higher BMI, which I guess means they were
20
better at releasing insulin because they
21
were used to insulin resistance, worked
22
better; that younger islets worked better.
248
1 He had a couple of other things on
2
the list, too, that they shouldn't have been
3
considered brain dead for too long before
4
the islets were taken.
5
But we really were not
presented
6
with the data that went along with that so
7
that we could evaluate that in any sort of
8
statistical manner. The data may
well be
9
there. I suspect Dr. Ricordi has
it. But
10
it's hard for us to judge without that.
11 DR. SALOMON:
Yes, I can respond
12
to that because I've actually seen the data.
13
I think the big question here is, this is
14
really important, this is the beginning of
15
the bisection of this from whole organ
16
pancreas transplantation to islet
17
transplantation.
18 I think that Dr. Ricordi and I
19 think Dr. Shapiro tomorrow
will make the
20
same point, Dr. Hering, you have to start
21
with the quality and characteristics of the
22
donor organ.
249
1 With that said, we could go off on
2
a big discussion here. I'm not
sure that
3
this is the right venue for that.
I mean,
4
there are statistics there, and I think they
5
didn't show you the data just because I
6
don't think they thought that this committee
7
has the expertise as we are right now to
8
start talking about the fact that there's 20
9
percent that go here and 3 percent go there.
10 I mean, all the data is available.
11
If everyone feels otherwise, I can go back
12
and I can tell you the exact numbers, within
13
reason, because it's very important.
14 DR. HARLAN:
I want to first make
15
a statement about my earlier disclaimer.
16
When I say I'm not here in a official
17
capacity, that is not as a member of a
18 BRMAC. I am a full-time government
19
employee, and what I say here represents
20
that.
21 I want to follow-up on what both
22
Dr. Salomon and Dr. Sherwin said.
It's my
250
1
perception that in fact the state of the
2
field right now is that we can, in a
3
post-hoc ergo-hoc say this patient did well,
4
therefore, the isolation islets were good
5
and the isolation went well and the donor
6
was good.
7 We can find correlates that are
8
associated with that. But as far
as
9
validating anything in a prospective way,
10
we're not very good at that yet.
Anybody
11
that does this can tell you that they got an
12
organ that anybody would have said would be
13
horrible, and they got good islets and vice
14
versa.
15 So I think it's premature. I
16
think we need more data before we can really
17
come up with a criteria for question one.
18 DR. SHERWIN:
I think the big
19
issue here, to put this into perspective,
20
and the one that we're concerned with, for
21
example, in my position as chair of the
22
NIH/NCRR islet cell resources program, if
251
1
the obvious idea here is to move the field
2
forward, you move the field forward with
3
good islet preparations that lead to good
4
validation of this as a procedure.
5 At the moment, there are a number
6
of impediments to doing that optimally that
7
start with the selection and distribution of
8
organs.
9 We're working
right now with the
10
UNOS Kidney Pancreas Committee and with HRSA
11
as a group to give very specific
12
recommendations to UNOS about allocation
13
that we feel will significantly improve
14
distribution of these organs, better organs
15
to these guys, to make good islet preps and
16
move the field forward.
17 DR. RAO: Dr.
Mulligan.
18 DR. MULLIGAN:
My impression is
19
that the criteria that Dr. Ricordi presented
20
were very reasonable for a start.
I guess I
21
thought it would be most helpful to talk
22
about the principles of the criteria.
252
1 There's obvious infectious disease
2
issues, and I think that we don't need to
3
harp on those. I think they've
covered that
4 whole
business. Probably most important to
5
us here is the function; where there's organ
6
characteristics that affect the function.
7
We heard something about how the temperature
8
affects things.
9 I don't know that we really need
10
to get into detail, but there's clearly some
11
principles. I guess what I'm
still a little
12
fuzzy on that I think is important here is
13 there's yield
issues, how many cells do you
14
get, and there's kind of how good are the
15
cells that you get.
16 I think in the question, there's a
17
little focus on maximum potential for
18
yielding absolute numbers. I
think like in
19
biotech in the past, where different people
20
have different ways to grow up a biotech
21
product, I don't think the FDA is
22
necessarily asking them what's the yield
253
1
they get of this or that recombinant
2
product. But more when you
purify, is it
3
pure?
4
So I guess my guiding
principles
5
with this is to take what is out there, if
6
we think it is a reasonable set of criteria,
7
and leave it flexible based on some guiding
8
principles.
9 So we
might say that we think it's
10
important, not so much that we have so much
11
concern about the maximum yield, but are the
12
cells that you are going to get tested in
13
such a way that you are comfortable?
14 So we heard about a couple of
15
different tests. If those tests
could be
16
then used to validate in a sense the
17
procurement criteria; to say when you take
18
an organ that's put in at this temperature
19
for this amount of time and you do this to
20
it, do you typically get this kind of yield
21
in those cells, or this sort of function of
22
those cells?
254
1 So I actually don't think it's all
2
that controversial. I assume
from where we
3
are at this point that we can do no more
4
than give guiding principles and just say
5
this makes sense, this is what you should be
6
looking for and that type of thing.
7 DR. RAO: Dr.
Childress.
8 DR.
CHILDRESS: I was just going
9
to follow up on Dan's comment and say that
10
obviously, UNOS has already put into place
11
draft proposals for revised allegation
12
criteria. I think Jim Burdick is
probably
13
going to comment on those tomorrow.
I'm
14
going to raise some of the ethical
15
considerations that I think they involve.
16 DR. RAO: Go
ahead, Dr. Sherwin.
17 DR.
SHERWIN: I didn't hear among
18
the various criteria glucose levels of the
19
donor. You know, you are a
little stuck.
20
You've got all these obese people that you
21
are sort of forced to use. Most
of them
22
probably have metabolic syndrome.
All of
255
1
them are under stress, and the islets are
2
probably not functioning as well.
3 But at least to get a hemoglobin
4
A1C is some assessment; since we know
5
about 30 percent of people with diabetes
6
have it and don't know it, particularly in
7
that population, I would like to be sure
8
that these people, even though they say they
9
are not diabetes, you know, are not.
10 Then some people in distress who
11
are very sick, I mean, we have been
12
monitoring people in the ICU recently
13
continuously. Most of them have
glucose
14
levels over 200 a good part of the time.
15 What that impact on beta cell
16
functions might be and viability is another
17
issue. So I think that that
would be the
18
thing that I would recommend, perhaps a
19
little better assessment of the donor from
20
the diabetes perspective as perhaps an
21
exclusion criteria.
22 DR. RAO: So
here it seems to me
256
1
is a really important issue, is that there
2
are generalized criteria which are already
3
in place in terms of looking at donors, and
4
those criteria seem reasonable.
5 But in addition to those criteria,
6
perhaps there should be some specific
7
criteria which might have some predictive
8
influence on islets. Possibly
that would be
9
the thing that ÄÄÄÄ by having that
10
information.
11 DR. RICORDI:
If I may just
12
comment on that, that is exactly a very
13
important point as we move towards higher
14
BMA donor, you have to exclude.
We are
15
doing it routinely; hemoglobin A1C from the
16
donor is very important as a way to exclude
17
transplanting islets from a diabetic
18
subject.
19 DR. RAO: Dr.
Mulligan.
20 DR. MULLIGAN:
Well, as I
21
understand your point, Bob, I think you are
22
being consistent with what you said before,
257
1
that if you don't really know the details of
2
the clinical outcome, you don't really know
3
what's important.
4 I think that's the key. I mean,
5
what you are suggesting is that we should be
6
more intelligent about the source of the
7
cells so that they have a greater chance of
8
being more successful clinically.
9 That's, I think, not where we are
10
asked to be looking at this point.
I think
11
we can't do that. I guess the
way I was
12
looking at it was, given the data that
13
exists here, and I think you are saying
14
maybe you don't know exactly what the data
15
is, but given what we have heard here, can
16
we make some judgment call about criteria
17
for a procurement?
18 I would say that would
19
specifically exclude the kinds of things
20
that you are talking about. They
really
21
incorporate the more mundane characteristics
22
of the viability of the cells, the makeup of
258
1
the cells and those functional
2
characteristics that they do have data on at
3
this point.
4 DR. SHERWIN:
I thought one of the
5
issues was who the donor selection was.
Is
6
that not part of this or not?
Did I miss
7
it?
8 DR.
RAO: You are absolutely
9
right. That question, the issue,
we are a
10
little handicapped in what one can do.
11 DR. SHERWIN:
I mean, obviously it
12
should be looked at, because there must be
13
data out there. Even though
things seem
14
logical doesn't always mean they're logical.
15
But my bet would be that somebody with
16
diabetes probably is not the best source.
17
So it makes logical sense.
18 DR. RAO:
Let's hold that thought
19
for a minute.
20 DR. BLAZAR:
I was hoping you
21
could clarify a little bit.
We're trying to
22 provide some advice. Most of the discussion
259
1
is going to be based on statistical
2
probabilities of whatever variable we're
3
looking at with potential
outcomes.
4 What I'm struggling with here is
5
are there going to be research questions?
6
Is there going to be prospective information
7
that needs to be collected so we understand
8
what predicts outcome? Are there
going to
9
be another set of recommendations that are
10
going to be placed as absolutes or relative
11
indication or contraindications for going
12
forward?
13 DR. NOGUCHI:
Let me just take a
14
minute here to try to address a couple of
15
the issues. I think the comments
by
16
Dr. Sherwin are extremely pertinent to our
17
discussion.
18 Even going back to what Abbey
19
said, we have had bone marrow and organ
20
transplantation take place in really the
21
absence of a public, and especially from the
22
patient's point of view of what do we
260
1
actually need to think about, given the fact
2
that this is a very complicated thing, even
3 though it's a
simple outcome.
4 Why don't we have the best minds
5
at work? So these comments are
right
6
on-point. It may or may not be
something
7
that we decide is a necessary component for
8
exclusion. It needs to be
examined. We
9
think there are some basic points that we
10
have in place of donor screening for
11
infectious diseases that are fairly
12
non-controversial.
13 But for us, we have 28 INDs, or
14
probably more even coming in today.
They
15
are not all uniform. We deal
with each
16
individual sponsor and IND at a time.
We
17
are continuing to see questions come up
18
about how can we move it forward in a more
19
comprehensive fashion.
20 That's part of the challenge that
21
we've kind of thrown at you, which is to say
22
we are at the beginning of a newer
261
1
technology where traditionally it's been
2
done mostly in academic centers, with
3
relatively intense oversight at the center.
4 But now we are looking at it in a
5
more global way, with the experience of a
6
lot of bone marrow and organ
7
transplantation. Above and beyond
simple
8
things like make sure that you screen the
9
donor and that there's a certain level of
10
viability, what in your best judgment are
11
those other things that this community,
12 within the context of the 28 and increasing
13
number of INDs that are being done, how can
14
we help guide that?
15 It is certainly one case. If we
16
tell a sponsor something, that's one thing.
17
But if we get advice that this needs to be
18
looked at because it may have a bearing on
19
the ultimate outcome, then it becomes easier
20
for everyone, including the sponsors, to try
21
to pool their resources and really work
22
together to get a common set of data.
262
1 So we agree with you, this is not
2
by any means a normal situation where we say
3
look, we have a certain clinical outcome and
4
here's the product and let's review that.
5
We are saying that, well, it has been said
6
before, we know there are some people who
7
are benefiting from islet transplants just
8
as some are benefiting from pancreas
9
transplants, and that we want to maximize
10
that benefit.
11 We are at a point where we have
12
basic information. What
next? What other
13
information is going to be, in your expert
14
opinion, useful to help us determine on a
15
dynamic basis how to move forward?
16 DR.
LEVITSKY: Continuing to beat
17
this dead horse, we understand that there
18
have been about 300 transplants done on
19
these protocols. That means
there's a lot
20
of data out there.
21 It's hard to suggest other things,
22
aside from the things like not choosing
263
1
someone with diabetes and not choosing
2 someone who's
HIV-positive or has got
3
Typhoid Fever to add to this collection
4
without knowing a little bit about what the
5
data look like; rather than just kind of
6
having been given a vague overview, which is
7
all we've gotten so far. A very
nice vague
8
overview. It sounds very
exciting, but --
9 DR. ALLAN:
I'm sort of
10
simple-minded. I just wanted to
weigh in.
11
I mean, I didn't really understand why we
12
were even doing this. I was sort
of
13
weighing in with these guys, or with her.
14 Then you made it very clear. In
15
other words, you are getting inundated with
16
all these, but you said you've got 28 INDs
17
coming in from all these different places.
18
So you may be getting concerned that you
19
need to have some sort of criteria in which
20
you can do something. I
understand that.
21 So that makes it more simple to
22
answer these questions, because we don't
264
1
have to be all that specific. We
just need
2
to provide some sort of a framework of
3
criteria that you can work with as advice to
4
you guys.
5 Is that correct?
6 DR.
RAO: Dr. Salomon?
7 DR. SALOMON:
I think that we're
8
making this too hard. There's a
tremendous
9
amount of experience, empirical and
10
supported by thousands of patient data sets.
11
There were 18,000 donors last year.
So that
12
there's incredible amounts of data.
13 What it tells us is that if the
14
donor is between 15 and X-number of years of
15 age, we'll take
donors well into their 60s,
16
and those are decisions that are made by
17
individual centers, exactly where they set
18
this upper limit.
19 If they are not diabetics, a
20
hemoglobin A1C, or an obvious history of it,
21
and certain extremes of body weight, et
22
cetera, and then after that, there's all
265
1 kinds of
criteria that are appropriately
2
based on, obviously, if they had slashing
3
abdominal trauma, it's not going to be a
4
donor. You know, that kind of
thing.
5 So I think that essentially, the
6
field is okay there. Where the
real focus
7
ought to be is in potency assays that
8
predict further down the road in the
9
manufacturing part that you have a good
10
quality islet prep.
11 Other than that, unless my
12
colleagues, with all respect to them, both
13
here and in the audience, would say no, no,
14
no, there's got to be some real -- I think
15
we are okay. I think the donor
thing is
16
okay.
17 DR. RAO: Go
ahead.
18 DR. WEBER:
Doctor, I'll offer a
19
little more clarification. I
certainly
20
agree with all the comments
that have been
21
made. But I'll encourage us to
think beyond
22
just the donor organ characteristics.
266
1
Obviously, in HRSA's organ procurement
2
system, organs are offered to various
3
locations. Time does elapse
during that
4
time period.
5 So some of these issues relate to
6
when that organ finally does become
7
available for islet transplantation, how
8
much time has elapsed? Has that
organ been
9
stored in conditions that are acceptable?
10
Are there predefined criteria?
11 We
recognize there is being data
12
collected here. In many ways, we
are trying
13
to get at the sense that, first of all, we
14
should all have some general agreement that
15
yes, there should be predefined criteria;
16
that certain situations are going to be
17
organs that just aren't going to be useful
18
because they weren't handled properly, while
19
they were waiting to determine whether they
20
can actually use them for an islet
21
transplant, to prepare islets.
22 So there are loads of nuances
267
1
here. I think in general, the
principles,
2
if I wanted to decide on it and say yes, you
3
do need to decide on criteria, and those
4
criteria should be driven by data.
5 I think Dr. Hering wants to say
6
something.
7 DR. RAO: Go
ahead, Dr. Hering.
8 DR. HERING:
I think that
9
Dr. Weber raised a very important point. If
10
you really ask what data is out that you can
11
analyze, I think a study has been done in
12
Edmonton a few years ago analyzing donor
13
data.
14 The single most important factor
15
determining outcome of islet isolation was
16
whether the islet team was given an
17
opportunity to procure the pancreases.
This
18
was associated with success, with an odds
19
ratio of 10.9.
20 The next highest odds ratio
21
was 1.3 or 1.4, and this was hyperglycemia
22
in the donor, and so forth. This
really
268
1
shows where the effort should go.
I think
2
timing of allocation is absolutely critical.
3 This has another implication. If
4
the islet team is given an opportunity to
5
recover the pancreas, the islet team can
6 also choose
adequate pancreas preservation
7
techniques like the ones mentioned by
8
Dr. Ricordi, the two-layer technique, which
9
again really impacts on the islet yield and
10
islet potency based on the limited data that
11
is available. This is the most
compelling
12
information that is available.
13 I think timing of allocation is
14
the single most important point.
15 DR. RAO:
Let's continue on that
16
and maybe consider this as that there are
17
issues with how you test donors in terms of
18
determining whether you will allocate that
19
pancreas. But once you have the
tissue,
20
there are certain criteria saying whether
21
that's going to be acceptable in terms of
22
isolating pancreatic islets.
269
1 You know, there have been various
2
tests that have been proposed and have been
3
discussed as to what's being done currently.
4 Did the committee have specific
5 additions, questions,
suggestions?
6 MS. LAWTON:
If I could just make
7
one comment, just to follow-up what the
8
folks from FDA say. I think it's
absolutely
9
critical that we need to continue to collect
10
this type of information and sort out what
11
information we need to collect.
12 Because today, we have been
13
talking about how do we move this product
14
forward for the sake of
patients. How do we
15
get to the BLA process and how do we approve
16
those BLAs?
17 To me, we've talked about the
18
process. We've talked about how
we need to
19 control the
process and validate the
20
process. You cannot do that
unless you
21
understand the starting material.
So these
22
are key factors that we have to understand,
270
1
and how it impacts how we process the
2
material and what our ultimate product is.
3
And then what the efficacy and safety of
4
that product is.
5 DR.
HARLAN: I want to support
6
basically what Dr. Salomon said and what
7
Dr. Hering just said. I think
that we don't
8
know enough actually to, other than the
9
obvious infectious safety concerns -- I
10
would move that we collect lots of data on
11
the donor, but not automatically exclude any
12
donor, and let the centers do that, and then
13
just collect all the data and see what pans
14
out in the end.
15 But I don't think we know enough
16
to exclude anybody automatically right now.
17
For instance, all brain dead donors have
18
hyperglycemia. If a center wants
to try to
19
isolate islets from people with hemoglobin
20
A1Cs of 6.5 and they have an in vitro assay
21
that predicts success, let them.
22 But I just don't think we know
271
1
enough to mandate criteria on the donor,
2
which is the question under review right
3
now, to stipulate those criteria.
4 DR. RAO: Dr.
Millis?
5 DR. MILLIS:
Once again, just to
6
change perspective a little bit, we have
7
been doing kidney transplants for decades,
8
liver transplants for decades, thousands of
9
experiences among all of us, and still at
10
all of the UNOS committee meetings,
11
acceptance criteria are still debated, hotly
12
debated. What should be accepted
as an
13
organ for which patient is still hugely
14
controversial.
15 I don't think the data certainly
16
is here in regards to islet transplantation
17
today. Certainly, we ought to
try to get
18
that. Certainly, we make
incremental steps
19
ever year in other organs as to this is a
20
good risk, this is a poor risk.
We ought to
21
collect that for islet transplantation as
22
well. But the data is just not
robust
272
1
enough.
2 DR. RAO: So
does that sort of
3
seem like a sense, if I can try and
4
summarize a little bit, that clearly there
5
are requirements and certain predictions
6
that will possibly be made based on donors.
7 But we don't have enough data to
8
be able to make any definite statement or
9
make any definite recommendation on that
10
basis. However, there's a strong
sense that
11
we should be collecting information which
12
might have possible predictive value.
13 The
information should definitely
14
include things which might affect islets,
15
such as glycose-isolated hemoglobin, glucose
16
levels and other standard criteria.
17 If we can move on to the next sort
18
of step of which part of this question was,
19
you have a pancreas and you decide that you
20
are now going to disassociate it, and there
21
are going to be protocols for disassociation
22 which give you a certain
number of islets of
273
1
a certain size.
2 You heard today from some of the
3
centers on how they decide what's quality
4
and what they test. Is there
anything
5
specific the committee feels is not covered?
6 For example, early in the
7
questioning, it became clear from
8
Dr. Mulligan's questions, for example, that
9
perhaps there should be additional testing
10
on other cells growing. Can we
predict that
11
they will continue to grow after you've
12
transplanted them?
13 Are OCR/DNA ratios the best way to
14
go in terms of assessing what you have got
15
in terms of the quality of islets?
That may
16
be an important thing to look at.
17 DR.
SALOMON: So do you want us to
18
start with just these product release
19
assays?
20 DR. RAO: I
just want you to have
21
them up there to remind people that these
22
are possible assays that we have already
274
1
discussed, and whether there should be
2
something in addition or whether this is
3
adequate.
4 DR.
SALOMON: Well, first of all,
5
islets don't grow, which is a big statement.
6
And in all of biology, nothing is as black
7
and white as that. Richard knows
as well as
8
I do.
9 But essentially, I don't think we
10
ought to be worrying about the islets
11
growing or not; just surviving,
12
revascularizing and functioning in vitro is
13
a lot. If we could get that to
work, that
14
would be great.
15 You know, whether there's stem
16
cells in these islet preparations et cetera
17
is just not clear at all. So
unlike stem
18
cells, like hematopoietic stem cells, which
19
is where a lot of people on this committee
20
are coming from, that engraft and then fill
21
the niche up, that's not the way islet
22
transplantation works to this point, at
275
1
least as far as I'm aware.
2 With respect to product release
3
assays, the viability assays with dye
4
exclusion are pretty close to
useless. They
5
are real easy to do, and I think they will
6
continue to be done, but they are pretty
7
close to useless. I think
Bernhard has
8
shown some very good data that
9
scientifically demonstrates how variable
10
those results are when you use a better
11
assay.
12 I think that the oxygen
13
consumption data that Bernhard showed, and
14 I've seen other people's
work on that, is
15
really compelling and a proof of principle,
16
sure. You want to validate it in
much
17
larger groups, and that's the way it ought
18
to move forward.
19 But at this early point, it's
20
something that can be done very rapidly and
21
something I think should definitely be a
22
part of it.
276
1 Insulin content is something that
2
is also very easy to quantify.
It's a
3
number. In a way, it backs up
the other way
4
Bernhard showed you of doing percentage of
5
islet --
6 DR. RAO: Do
you mean insulin
7
release or insulin content?
8 DR. SALOMON:
I'm talking about
9
actual insulin contents. You
essentially
10
just take a given mass of tissue, or do it
11
by DNA. You extract all the
insulin and
12
measure as a function of grams per islet
13
tissue in the infusion to the patient.
That
14
would be one way of doing it. So
that's
15
another very good measurement.
16 Obviously, gram stains and
17
endotoxins is good, and I think the idea of
18
having several other assays in development,
19
like looking at apoptosis assays, those are
20
now something you can do by flow cytometry,
21
so it's very rapidly done.
22 There are ways to do
277
1
phosphorylation, activity assays like you
2
could do a P-38 or a junKinase (?) assay,
3
for example, now on the flow cytometer, so
4
there's a bunch of ways that, if there was
5 enough
incentive as this field moved
6
forward, that you could do these kinds of
7
assays.
8 So I think that's kind of the
9
direction that we ought to go.
There's one
10 caveat, though, that
I'll finish with. All
11
these assays are not the kind of assays that
12
are normally found in the clinical lab.
13 Therefore, when you go forward in
14
a clinical program in multiple centers
15
around the United States and start using
16
assays that by their nature are not
17
standardized -- no criticism of the assays;
18
it's just facing the reality -- we're going
19
to have to be very careful that part of the
20
process of going forward for a new program
21
into this is to demonstrate that they can
22
reproducibly get data with their assays, and
278
1
that those assays are reproducibly
2
equivalent to what's been demonstrated by
3
the people who led forth with the first
4
successful programs.
5 That's going to be an interesting
6
process. There's going to have
to be a
7
process developed here of validating a new
8
program's set of assays in order for this to
9
be sensible.
10 DR. RAO: One
more sort of
11
addition maybe, and you can look at this,
12
what do you think about the fact that you
13
should also be looking at the other cells,
14
other than the beta cells
present in the
15
islet?
16 DR. SHERWIN:
Do you mean the
17
asinar (?) cells or the ductile cells?
18 DR. RAO: All
of them. You will
19
notice that when they presented the data,
20
for example, they looked at the percentage
21
of beta cells relative to the other islet
22
cells.
279
1 DR.
SHERWIN: Well, I think the
2
relative proportion of beta cells is the
3
most crucial piece. The other
thing I would
4
be interested in is perhaps ductile cells,
5
just because of the potential that there
6
might be primordial cells there, so that
7
could be of some interest down the road.
8 But clearly, the beta cells would
9
be the most important. And the
rest, I'm
10
not sure would be as crucial.
11 DR. SALOMON:
Yes. I think what
12
you are referring to, Mahendra, is that an
13
islet consists of not just insulin-producing
14
beta cells, but also glucagon, pancreatic.
15
I'm sure you know that, of course, these
16
other cell types.
17 From a biological point of view,
18
it's likely that true glucose homeostasis is
19
contributed to in a more complex way than
20
simply the beta cells sense glucose and put
21
out insulin. We know that. There's a lot
22
of data on the roles of glucagon, for
280
1
example, and pancreatic polypeptide and
2
sematastatin (?) in controlling glucose
3
homeostatishomeostasis. I think
they all
4
ought to be measured.
5 But I
think at this early point, a
6
strict product release assay would probably
7
be fine if it focused on insulin content and
8
beta cell function, beta cell content.
9 DR. BLAZAR:
So Dan, you
10
summarized that very well. I
guess the
11
question is what should be mandated for
12
assays to be collected; what are the
13
research questions that may become mandated
14 assays as the FDA
reviews these protocols;
15
what might they recommend versus what they
16
may say you absolutely need to collect.
17 Then, as you said, it has to be
18
assays where all the sites that are
19
submitting the INDs are able to have this as
20
validated assays.
21 I don't know if you or Bernhard or
22
someone, but it seems to me we should get
281
1
our hands around that question first, and
2
I'm still confused.
3 DR. RAO: Dr.
Kurtzberg?
4 DR. KURTZBERG:
I'm troubled by
5
the fact that the FDA at this point would be
6
making those decisions. I don't
feel like I
7
could make those decisions. I
think there
8
are experts in the room and around the
9
country that have a lot more ability to do
10
this.
11 I would say the FDA ought to bring
12
together those experts and let them put
13
together a program that will in a systematic
14
way define what questions need
to be asked,
15
and in a systematic way determine how to
16
best accomplish that.
17 I just feel uneasy sitting around
18
this table making those decisions or even
19
making recommendations, because I think
20
there are people who have expertise in this,
21
and it's at the level where it still should
22
be in their hands.
282
1 I don't think enough has been done
2
to take it to a panel and say you guys
3
decide. I also don't think that
we should
4
make this so expensive that nobody can do
5
it, because it's great to sit here and
6
theorize about testing 20 assays that we
7
have no idea if they are predictive of
8
anything, but there's got to be funding for
9
that.
10 In the meantime, there are people
11
who might benefit from transplants, and you
12
don't want to impede that.
13 So I would recommend bringing
14
together a working group of people in the
15
field who have the expertise to answer some
16
of these questions, rather than ask a panel
17
that doesn't really have, at least some of
18
the people like me, don't have the expertise
19
to answer.
20 DR. RAO: You
know, three of the
21
major centers which are doing transplants
22
are here and we heard some of their
283
1
recommendations. Would that
constitute a
2
working group?
3 DR. KURTZBERG:
It's a good start,
4
but they are not answering these questions
5
now.
6 DR.
RAO: Maybe we should ask them
7
to.
8 DR. RICORDI:
I just want to make
9
a general comment. I think that
we have to
10
define what are the critical tests required
11
to identify whether the product is safe and
12
is viable and will work in patients.
13 Using this very simple system so
14
far at our three institutions, we have had
15
close to 100 percent success rate in having
16
a biologic effect with this safe product.
17 From the other set of questions,
18
it is whether the research priority to
19
define the much more elaborate and what will
20
become maybe the ultimate level of testing
21
for potency testing of the islet
22
pre-transplant, when this may become a BLA
284
1
for when you have a product that is an
2
unlimited source, that you treat millions of
3
patients.
4 But right now, what we are hoping
5
to have as a set of guidelines for the
6
objective here is to be able to do this few
7
hundred transplant patients and obtain a
8
status that is beyond the IND that would be
9
the BLA for an orphan biologic, or whatever
10
would be appropriate with the set of testing
11
that would allow us to do these patients.
12 Because the absence of this BLA
13
would be the major impediment to the
14
development of a cure for diabetes, because
15
right now, the research funding that could
16
go towards research to bring us to the next
17
level is going to pay for the
18
immunosuppression in all the patients that
19
have been transplanted successfully.
20 So that right now, until we are
21
still in the end, this status for procedures
22
that have been proven to work, I have in my
285
1
institute $1 million budget this year to
2
pay $40,000 a pop for the immunosuppression
3
of patients that have been successful.
We
4
calculated in two years, we will be
5
bankrupt.
6 In other centers, I think it's a
7
challenge that we are facing in diabetes
8
research; if we are blocking this kind of
9
initial, limited development, recognition
10
that we can hope for some kind of
11
reimbursement, not for the whole processing,
12
but just for the immunosuppression, maybe,
13
would severely impair our ability to move
14
forward with the real cures.
15 DR. RAO: To
sort of address what
16
Dr. Kurtzberg said, do the experts in the
17
field feel comfortable with this minimal set
18
that you presented up here as being
19
adequate?
20 DR. RICORDI:
Bernhard says as
21
well. But I feel comfortable
that given
22
that it is criteria that we have now in the
286
1
last two years, the three leading centers
2
here have obtained very good clinical
3
results.
4 Am I satisfied completely? Not
5
for sure. That's why we have
five years of
6
planned research and continue to improve.
7
As soon as the continuous multiparametric
8
monitor is confirmed and the OCR is
9
confirmed, we continue to add additional
10 tests that will be more representative.
11 But I feel we have a good starting
12
point to start.
13 DR. HARLAN:
I think that
14
Dr. Ricordi raises excellent points about
15
funding. But I think we should
not consider
16
that right now.
17 The problem is, as scientists
18
doing investigation, you like to start from
19
what you know and then build on it.
The
20
trouble with this field is that at each step
21
of the process, there's significant
22
unknowns.
287
1 One other point is that these
2
criteria do work in some centers, but
3
applying these same criteria at other
4
centers, they don't work. So it
tells you
5
that these criteria are not sufficient to
6
predict success.
7 Other centers have used these
8
exact criteria and have had zero success.
9
So these are not the answer.
10 DR. SHERWIN:
What about an in
11
vivo test?
12 DR. HARLAN:
I was going to go
13
into that.
14 DR. SHERWIN:
Well, it seems to me
15
that that's pretty simple.
You're already
16
culturing islets. They are sitting
around
17
in an enriched media, presumably.
If you
18
waited three days, you could probably tell
19
whether they work or not.
20 You know, if you set up strict
21
criteria of the number of islets to do a job
22
within three days and so on, that would give
288
1
you probably, I would think, the best way of
2
doing it.
3 DR. HARLAN:
Well, I think I am
4
not allowed to respond to that directly.
5
But I wanted to continue the point and I'm
6
almost done. I'm following the
rules very
7
carefully, I think.
8 But Dr. Hering presented some very
9
nice, I think, data today to say that our
10
best test right now is the nude mouse assay.
11
The trouble with it is it's not prospective.
12 But that gives
us a target.
13 What I would suggest is that the
14
FDA consider asking any center with an IND
15
that's doing islet preps, regardless of what
16
islets you get, whether they are
17
transplantable or not, you apply the various
18
in-development assays that prospectively
19
might predict success, and then test them in
20
the in vivo mouse model.
21 If they are
acceptable for
22
transplant, then obviously, you give them to
289
1
the patient. That's the gold
standard, how
2
the patient does. But that we
begin to try
3
to validate these various tests that we
4
heard about, and we heard about a lot of
5
them today with great promise:
OCR/ATP,
6
beta cell mass, P-38 kinase.
7 But let's begin to try to validate
8
those prospective studies in in vivo
9
clinical testing and in this nude mouse
10
model so that we know prospectively whether
11
a prep is good or not.
12 Right now, we don't have that
13
prospective capability.
14 MS. LAWTON:
Can I ask a question
15
on that bioassay, because I agree it makes a
16
lot of sense to develop that type of
17
bioassay, but I also know from experience
18
that these types of bioassays have very
19
large CVs.
20 You have to do very large numbers
21
of animals. I don't know whether
that's a
22
limitation for us here with the number of
290
1
cells that we have, if you have to use such
2
a large sample to do the bioassay.
3 So I just ask the question on the
4
assay itself.
5 DR. RAO: Dr.
Ricordi?
6 DR. RICORDI:
Just a comment in
7
general without responding, but the bioassay
8
will not work, because we would have to
9
prolong the culture to one week before you
10
transplant, because the first two days that
11
you put it in that mouse, the nude mouse
12
bioassay is very good to
validate which one
13
of the other surrogate tests that then you
14
can use instead of the nude mouse model,
15
because the first two days of insulin
16
independence, after you put islets in the
17
mouse could be because the islets are dying
18
because they are damping and you always have
19
three or four days to see sustained normal
20
glycemia.
21 So then you would have to prolong
22
culture to one week, and then what do you
291
1
know that maybe in the last three days of
2
culture, you lost function, so you should do
3
another nude mouse at one week.
4 So it is again a very important
5
point. We have to study new
tests. But we
6
have the prospective tests right now that
7 allow us to have
success in a predictive way
8
at several institutions.
9 DR. RAO: Dr.
Salomon.
10 DR. SALOMON:
Yes, I think that
11
Camillo is absolutely right.
This gets to
12
Bruce's point as well. What can
you do to
13
establish the equivalency of assays as new
14
programs step up for their BLA?
At this
15
point, I'm presuming it's going forward.
16
The answer is: These are the
assays.
17 What I think Bernhard and Camillo
18
are saying, and I think, Dave, that's what
19
you were saying as well, is these assays are
20
not hard to do. We do them. We have 100
21
animals right now in our animal facility
22
under various aspects of this as we speak.
292
1 So what you do here is you do
2
enough of these tests to validate your set
3
of potency assays, and then, obviously, when
4
you are ready to transplant a patient, you
5
use the potency assays as a surrogate
6
marker.
7 I feel very strongly that we don't
8
want to come away with a recommendation
9
saying that every time you do a transplant,
10
you have to do a nude or a skid mouse
11
transplant before you can give it to the
12
patient. That's missing a point
there.
13 DR. HARLAN:
If I made that point,
14
I didn't wish to make that point.
The gold
15
standard here is going to be how the patient
16
does, not how a mouse does.
17 I'm just saying every single islet
18
prep, whether it's transplanted or not,
19
should go into that nude mouse, not in a
20
prospective way, but in a retrospective way,
21
to see: A, is the nude mouse
really
22
predictive of human experience.
And those
293
1
preps that we don't transplant, can we use
2
these in vitro assays to predict how they
3
would do in the nude mouse?
4 I'm not saying wait for a week and
5
transplant patients with islets.
Go ahead
6
and transplant them at Day 3, but transplant
7
a nude mouse, too, and gather that data.
8 DR. RAO: So
if I understand that
9
right, it's a recommendation for additional
10
information. Is that right?
11 DR. LEVITSKY:
I was at the first
12
of these meetings when we were first
13
discussing this several years ago.
I
14
remember that at the time, the Edmonton
15 protocol data was
under embargo by "The New
16
England Journal." So we
felt like we were
17
having a discussion inside of a paper bag.
18 I have this same feeling again in
19
that I hear hints that these same wonderful
20
tests work in the hands of the wonderful
21
gentlemen behind me, but they don't work in
22
some other people's hands.
Without knowing
294
1
if there are some hints as to why these
2
wonderful tests work in one place but not
3
another, it is very hard to be able to know
4
how to respond to most of these questions.
5 DR. REEMS:
My name is Jo Anna
6
Reems and I'm with the Puget Sound Blood
7
Center in Seattle. We are one of
the
8
novices with regard to islet
9
transplantation. We participated
in the ITN
10
trial as well as we performed a couple of
11
SKI transplants as well.
12 I'm responding to the comment with
13
regards to these release criterion and
14
whether they're
sufficient. I can say that
15
we also have to keep in mind that our
16
products may be meeting these release
17
criteria, but there are other components
18
that need to be dealt with, one of them
19
being with regards to the transplant site
20
and whether or not the patients are being
21
maintained with their immunosuppression.
22 So with regards to those sites
295
1
that may not have had the success that other
2
sites have, it may not have been necessarily
3
due to the product meeting the release
4
criteria and then not being able to result
5
in insulin independence, but rather because
6
of difficulties with the immunosuppression
7
regimen.
8 DR. RAO: It
is a good point. I
9 want to extend this just a little bit in
10
sort of to what Dr. Salomon said already.
11
So there are two issues here.
One is can
12
you have an absolute set of criteria which
13
are used by every center, or can you have a
14
defined set of potency criteria which are
15
applicable to an individual center and you
16
hope that you can extend them to other
17
centers by having some sort of comparability
18
sort of issues so that there will be a way
19
to compare the data and compare standards.
20 So it seems to me that what data
21
has been presented suggests that at least
22
there are some kind of potency criteria that
296
1
work in one set of center's hands.
We don't
2
know whether they can extend.
But what we
3
need then is some way of being able to
4
establish additional things so you can
5
mandate these tests right now.
6 You can ask for there to be
7
comparable potency criteria that would be
8 used for any
application which looked at
9
using islets for transplant.
Does that
10
seem --
11 DR. O'NEIL:
This is the composite
12
list of the three gentlemen who are really
13 pioneers in the
field. So I think that no
14
one would disagree with the items that are
15
on the list as far as release criteria.
16 I would just suggest that in a
17
cell identity assay, your pre-product by
18
divazone (?) standing, which is a very
19
subjective analysis, and I would suggest to
20
have cell identity as a post-release
21
criteria as well, with more sophisticated
22 techniques than
that, with some of the data
297
1
that Bernhard presented on the different
2
cellular components that are present in the
3 graft.
4 I also want to question the
5
potency assay that is up there with the
6
insulin release glucose stimulation static
7
assay. We've heard that there's
really no
8
correlation between that and
clinical
9
efficacy.
10 In light of it being a
11
post-release assay, I would suggest that the
12
nude mouse transplant, where you are going
13
to gather data to look at potentially other
14
predictive assays that are real-time, that
15
that would be a more appropriate
16
post-release assay than the
17
glucose-stimulated insulin release.
18 DR. HERING:
If I could comment, I
19
think nobody can really predict what type of
20
predictive assay will eventually be applied.
21
But I think we don't need to define this
22
very well at this point in time.
I think my
298
1
understanding is whoever wants to submit a
2
BLA must be prepared to provide compelling
3
evidence that this issue has been addressed;
4
must present a validation plan; must provide
5
good evidence that what he or she proposes
6
is solid and has been validated and can
7
withstand scientific scrutiny and so forth.
8 I think whatever that assay is, I
9
think could be sufficient and could define,
10
really, the process. People can
be creative
11
and innovative, and we should probably not
12 limit this.
13 DR. RAO: One
issue that seemed to
14
have come up several times with different
15
people has been the idea of using a nude
16
mouse assay to validate your tissue, and do
17
it enough times so that you can have some
18
independent assessment of predictive power
19
of your test.
20 Does there seem to be a consensus
21
in the committee that that would be
22
something useful to do or information to
299
1
collect, or still an open question?
2 DR. KURTZBERG:
I think it's still
3 an open
question. I don't think it's
4
appropriate to require everyone to do a nude
5
mouse assay.
6 But I was going to say something
7
different, which is I think philosophically,
8
in cellular therapies beyond pancreatic
9
islet cell, I'm not sure it makes sense to
10
demand that you have an assay that proves
11
efficacy.
12 Even in other organ
13 transplantation,
solid, liquid, where you
14
have more liberal standards about the organ,
15
you can never predict from the organ that
16
the patient will or will not do well.
17 I think the FDA's obligation is to
18
make it safe, not toxic. You
know, the
19
obvious things; that it is what it is and
20
it's labeled right, that it was handled
21
properly. But to require an
assay that
22 proves it's going to work,
I think is an
300
1
unreasonable expectation.
2 DR. NOGUCHI:
Just to say, we
3
would agree with that comment.
The potency
4
assay is to help the physician say that I
5
have a reasonable expectation this will
6
perform in the way that clinical trials have
7
shown or collected clinical outcome data has
8
shown.
9 This does have a different nuance
10
than trying to approve a thrombolytic drug.
11
This is transplantation. In many
cases, not
12
in the case of cord (?) blood, but in this
13
case if it doesn't work, you may try another
14
one. So it's not a question of,
given a
15
particular regimen, how many are
16
successfully reversed in some condition; it
17
is really a question of can you make this as
18
consistent as possible, knowing that
19
transplantation is a very variable type of
20
end-game that you play in terms of curing or
21
treating a condition. There are
many things
22
that go into that. But each time
you do it,
301
1
what is the assurance that what you give is
2
about as good as can be used at that point
3
in time?
4 But you are right; we are not
5
looking for a test that will say if you
6
don't pass this test you can't use it.
That
7
is not at all in any of the questions that
8
we are asking.
9 We are saying that ideally, that
10
would be a goal worth achieving should it be
11
feasible. But the rules and
regulations we
12
have cover everything from drugs, biologics
13
to now these newer kinds of things of
14
cellular transplantation.
15 So it's an ideal, but we don't
16
expect that we'll ever have that predictive
17
power.
18 DR. RAO: Dr.
Shapiro.
19 DR. SHAPIRO:
If I could, I would
20
emphasize that point. There's
138
21
transplants, clinical transplants that we
22
performed in the last four
years at the
302
1
University of Alberta; 136 of
those 138 had
2
immediate function with detectible C-peptide
3
in the recipients.
4 We haven't routinely done nude
5
mouse transplants, not that I have any
6
objection to the procedure of assays.
I
7
think that's a good idea. I
think really
8
where we need those assays in fact is not
9
for the good preparations where we have a
10
high yield of very nice looking islets with
11
superb counts; where we need those assays,
12
in fact, is where we have a marginal islet
13
mass, where we have some question in our own
14
minds whether that preparation is going to
15
do well in the patient.
16 In fact, one thing that we could
17
add to that list, those preparations where
18
we have some question about them is how do
19
those yields hold up in culture?
Often, we
20
find if we have a marginal islet mass with
21
questionable viability at time zero, if we
22
keep it in culture for 24 or 48 hours, we
303
1
might find that the counts diminish quite
2
markedly for a marginal prep, and we won't
3
transplant that.
4 DR. RICORDI:
I would like to
5
emphasize also what Jo Anna Reems said from
6
Seattle, that at the time you have a
7
preparation that has been validated with
8
your test, there are still 100 different
9
ways you can fail to have a biologic effect.
10 This is not only the failure of
11
achieving therapeutic levels of
12 immunosuppression
in the first two weeks.
13
You can also damage the cells on the way to
14
the operating room; the virologist may fail
15
to cannulate portal veins through multiple
16
attempts, and you may be sitting with a bag
17
with a pH droppings or whatever.
18 You know, this is why it takes
19
experience and multidisciplinary and
20
multi-experienced team. There is
no
21 short-cut. Everyone who tries to do this
22
cutting corners is going, has going and will
304
1
go into surprises as far as efficacy.
2
That's why in the hands
of these
3
centers, there have been certain results. I
4
believe it is the competence on the site.
5
It is not just islet processing.
6 DR. BLAZAR:
Can I ask in that
7
regard, Camillo, if the odds ratio is 10.9
8
for the pancreatic team processing the
9
tissue, what assays are most affected by
10
that, since that seems to be a clinically
11
validated outcome measurement?
12 As you have looked at all of these
13
assays, are there certain ones that are very
14
sensitive to that?
15 DR. RICORDI:
Well, I completely
16 agree that the most
important thing for all
17
of us is starting with a good organ.
A lot
18
of times, teams that accept marginal donors,
19
anything that has been trashed by a drug and
20
transplant program and receive instead a
21
poor pancreas, they may have 50 percent
22
success rate or less, or zero in some cases.
305
1 So having your team in control of
2
the procurement is the same guarantee like
3
the cardiac transplant team.
They go to
4
preserve the health. They
wouldn't let me
5
take the heart and trust me to keep the
6
right temperature, because, you know, the
7
life of the recipient is on the line.
8 So that in a way, what this shows
9
is that if you are careful and you treat the
10
pancreas as valuable as the cells should be,
11
from the very beginning, your likelihood of
12
success is very high at the end.
13 If you lose control of something
14
during this procedure, you may compromise
15
your ability.
16 DR. BLAZAR:
But which assays?
17
Since the focus is on assays, which assays
18
correlate with whether or not the pancreatic
19
islet team procures the organ, since that
20
seems to be the most compelling validated
21
clinical outcome measurement.
22 DR. HERING:
If I could answer
306
1
this one: At that time, I think
the islet
2
yield was the only outcome measure.
3 DR. RICORDI:
Yield and viability.
4 DR. HERING:
Viability, yeah.
5
Well, I think more studies clearly should be
6
done. This needs to be
addressed, but this
7
was the main outcome at the time.
8 DR. RICORDI:
But you know, yield
9
and viability may not be sufficient, if you
10
have problems with procurement.
For
11
example, if the viability of the pancreas
12
was compromised or the temperature during
13
procurement was too high, if you have
14
accepted an hypoxic donor or a pancreas that
15
was already compromised at the time of
16
procurement that the temperature was too
17
high, you may have those apoptotic changes
18
at the time you have your islet product that
19
you may not be able to detect with only a
20
viability assay.
21 So you may say I count 600,000
22
islets and they are 95 percent viable, but I
307
1
have apoptosis already triggered the next
2
day. I will lose 60 percent of
them.
3 That's why if you culture two or
4
three days, you eliminate 90 percent of that
5
risk, because if there was something
6
dramatically wrong with the donor pancreas,
7
you will see it after one night culture. If
8
the products are falling apart, the islets
9 don't round up,
so that in a way, the
10
culture is a level of safety and is an
11
integral part of the protocol for this
12
criteria.
13 DR. HARLAN:
With regard to the
14
point that Dr. Kurtzberg made
and what
15
Dr. Reems said, I agree that we are not
16
looking for any guarantee of success.
17
Rather, in the islet isolation business, a
18
truism is that in most centers' hands, you
19
get islets deemed suitable for transplant
20
about 50 percent of the time, and it's only
21
about 50 percent of the time.
22 What I'm suggesting is that for
308
1
that other 50 percent, if you do the nude
2
mouse, which is our best surrogate assay,
3
maybe we can begin to more quickly validate
4
other prospectively performed quick in vitro
5
assays that might predict success.
6 The other point I'd make, based
7
upon Jo Anna Reems' comment is, that when it
8
fails presently, you can name ten different
9
reasons why it may have failed, but you
10
really don't know. If you have
any assay
11
that allows you to say, well, it can't be
12
the islets, then we are further down the
13
road.
14 Right now, we don't have those
15
assays that say it's probably not the islets
16
because they worked in the nude mouse.
17
That's all I'm suggesting.
18 DR. SHERWIN:
I would tend to
19
agree with David's perception.
It just
20
seems to me that the goal is to get a good
21
surrogate marker for viability.
I agree,
22
from what Camillo was saying, it can't be at
309
1
this point a nude or a skid mouse.
But I do
2
think that when people come in to get a
3
license, they need to establish that they
4 have validated a
surrogate marker for
5
viability, and that should be the criteria.
6 The criteria could then be the
7
animal model to use to develop the surrogate
8
marker. Of course, as people get
experience
9
with the clinical transplantation, they can
10
also validate it from that perspective.
11
That would be the way I would propose to go
12
ahead with this.
13 DR. RAO: Go ahead.
14 DR. SALOMON:
Yes, I think again,
15
just for clarity, we need to remind
16
ourselves it's not just viability, it's
17
biological functionality. That's
what we
18
need a surrogate marker for.
Otherwise, I
19
agree with everything you said.
20 DR. SHERWIN:
Well, that's what I
21
meant. You are absolutely
correct. That's
22
what I was trying to say and didn't say it.
310
1 DR. SALOMON:
If we have a measure
2
for biological functionality, then
3
everything downstream of that reflects the
4
typical complexity of any sort of
5
transplantation procedure. I
think that we
6
can all deal with that.
7 Getting back to something, you
8
know, Rich Mulligan said, basically to
9
articulate a principle, the principle will
10
be useful to the FDA, so the principle to me
11
in this particular area is that you need a
12
set validated by you, by your center before
13
you step up and ask for a BLA for your
14
program.
15 Because I think in the end, the
16
committee needs to remind themselves that
17
the comments you made about the first time
18
we dealt with this in 2000, there really is
19
a big center effect. But at the
same time,
20
you need to be aware of the fact that when
21
islet transplantation suddenly sounded
22
in 2000 like it could be done
at a handful
311
1
of centers, centers all over the country
2
stepped up and said, well, me, too.
3
Hundreds of them.
4 That's something that if the FDA
5
wants to be responsible, needs to be very
6
aware of the fact that there are going to
7
have to be some criteria here.
We can't be
8
laissez faire about it.
9 It's a big responsibility to
10
hundreds and thousands of patients who could
11
potentially get transplanted with poor
12
quality product.
Immunosuppressed
13 induction,
immunosuppressed mice, I'm not
14
interested in giving that to anyone and then
15
giving them a worthless bunch of cells.
16 So this is really serious that
17
they have some sort of biological assay.
18 DR. RAO:
Again, let me try and
19
restate what I had said a little bit earlier
20
and see whether that captures the sense of
21
the committee. It is that it is
hard to
22
establish any absolute criteria and
312
1
generalize and make it absolute for any IND.
2 But each individual center or
3 application should
really have a clear-cut
4
potency measure, and that potency measure is
5
essentially a surrogate measure.
And that
6
should be validated by some clear-cut test
7
which will convince the FDA that that's a
8
reasonable surrogate measure for that
9
particular institute.
10 Given the level of information
11
here, it's important to start collecting
12
additional information to make sure that
13
perhaps in the future, you will be able to
14
have additional criteria to change the
15
validity of a test.
16 DR. ALLAN:
Just to flip it, the
17 Washington
transplant information -- I don't
18
remember her name -- just gave us, in some
19
cases, it's more important data to know
20
about all these centers where it doesn't
21
work. We've got these three
great centers,
22
but to get the data, if they've got
313
1
viability and they've got all of this other
2
stuff at these other centers and they are
3
not getting any success or much success, you
4
might be able to learn a lot more from that,
5
I think.
6 Because then you can put it
7
together with viability and all these other
8
assays and then come to some conclusion as
9
to whether some of these assays are actually
10
useful or not.
11 DR. RAO:
That's a really nice
12
lead-in into the last part of the question
13
session, which really suggested exactly
14
that: How do you develop some
sort of
15
comparability? How do you look
at which
16
tests across centers? You know,
look in
17
these preclinical studies as well.
18 Maybe you want to lead off, Alan,
19
and say something about that.
20 DR. WEBER: I
want to make a point
21
that we missed the second question totally.
22
So at some point, please come back to that
314
1
question before we conclude.
2 DR. RAO:
Let's take the last one
3
first since it seems to lead to that
4
directly. We've answered part of
this
5
already, right, in saying that you can't
6
have one uniform set of tests, but you
7
really have to develop some kind of potency
8
measure, and that surrogate marker is what
9
can be used as a comparable thing as long as
10
you collect additional information.
11 Is there anything else in terms of
12
pre-clinical studies or clinical studies
13
which are specific which should be followed
14
in assay, just like we suggested, maybe the
15
skid mouse assay, as a possibility that one
16
should consider?
17 MS. LAWTON:
I'll make a comment,
18
just to start the discussion.
I'm assuming
19
the comparability. You're not
going to be
20
doing it as routinely as you are for lot
21
release.
22 So I think to look at things like
315
1
the mouse model would be appropriate to look
2
at for comparability studies, where you can
3
get a little bit more information to
4
understand, but you are not having to do on
5
a routine basis.
6 DR. RAO:
Let's go back to the
7
question that we skipped. Let me
read it
8
out because I thought we'd answered part of
9
it. We can go with that in a
little bit of
10
detail.
11 DR. WEBER: I
apologize for
12
interrupting. Can we go back,
before we go
13
on to the next issue? Can I
parse this out
14
just a little bit more to make sure why we
15
were asking this question, if you don't
16
mind?
17 I think what
Ms. Lawton said was
18
appropriate in terms of a mouse assay, and
19
it wouldn't be done every time.
But there's
20
a couple of nuances, and maybe this is
21
something that wasn't clear.
22 Comparability, clearly there's
316
1
right now, there's a variety of centers.
2
There's 28 INDs, for example, and
3
manufacturing in some ways is doing fairly
4
similar, to be all different centers, but in
5
some ways it's different.
6 The challenge, of course, for the
7
FDA is when we do get data that comes in the
8
form of a license application, to interpret
9
that data in the sense that Site A, Site B,
10
and Site C, one might culture for seven
11
days; one might use fresh; one might do
12
something different that we
haven't thought
13
about.
14 We're going to be asking how to
15
interpret that clinical data based on a
16
product that may be, in our perspectives,
17
different.
18 So comparability can be thought of
19
in that way as well. There's
really an
20
important issue for us to be able to get
21
some feedback on: What you would
think
22
would be important. Maybe clinical data is
317
1
sufficient if the islets in fact cure
2
diabetes, regardless of whether they are
3
maybe enough. But maybe it's
not. There
4
are other assays, other ways of addressing
5
it.
6 MS. LAWTON:
Can I just ask a
7
question then, because you're talking about
8
a BLA situation again. Are you
foreseeing
9
that you may get a BLA application with
10
numerous sites all having a different
11
process for that one clinical trial result?
12 DR. WEBER: I
think that's a
13
possibility. It's not
precluded. There's
14
nothing that precludes that.
It's way down;
15
this is a hypothetical. But it
is an
16
important issue. It could be
within center
17
processing or across center potentially.
18 MS. LAWTON:
The reason I asked
19
the question is because generally in the
20
past for BLAs, you have to take that into
21
account in the design of your clinical
22
studies. So you understand that
if any of
318
1
those differences actually end up in a
2
different result in the clinical setting.
3 That's why I asked the
question.
4 DR. NOGUCHI:
I'll expand upon
5
that a little bit. I think what
we are also
6
broaching is that in the abstract, as a
7
concept, there may be situations in which a
8
clinical trial may not necessarily be
9
restricted to a single IND or a single
10
license folder; that some of the data may be
11
garnered from a more global assessment of
12 clinical
outcomes from many different
13
centers.
14 If we are to move to a model where
15
a traditional clinical trial as what might
16
be expected for a pharmaceutical
17
manufacturer, we are positing that there are
18
alternative methods of getting clinical data
19
that are sufficient to be convincing that
20
this preparative procedure to make this
21
product leads to a result, then I think
22
these other issues of comparability now
319
1
begin to play into much greater importance.
2 If we can only rely on three
3
centers for clinical outcome data, well then
4
that really forces us into more of a
5
position of well, maybe those are the only
6
centers that are going to be able to deliver
7 this particular
therapy.
8 We personally, I think, are open
9
to the question, and want to know whether or
10
not other centers who cannot match it can
11
eventually match that outcome.
12 Then it comes back to what John
13
was identifying as, yes, it's very important
14
to know what are those factors that
15
contribute to patient outcome?
If it's a
16
technical factor such as they are just not
17
getting the same viability and so forth,
18
that's one thing that can be addressed.
19 If it's a different question such
20
as the immunosuppressive regimen is similar,
21
but a little bit different in some other
22
areas, that becomes a different question to
320
1
be asked.
2 Part of your
question, Alison, is
3
yeah, we're trying to stretch the boundaries
4
of how we evaluate clinical data vis-...-vis a
5
product.
6 DR. ALLAN: I
may be off-base
7
here, but also the way I'm looking at it is
8
if you've got 28 other INDs, another way to
9
look at it is maybe one of those INDs or two
10
of those INDs will get 100 percent long-term
11
efficacy because they've done it a little
12
differently. You know, what
they've done
13
may not on the surface look like it might be
14
very good, but it may end up all of a
15
sudden, they get really good results.
16 So then how will that affect what
17
you may be looking to impose on these types
18
of assays or studies that are ongoing.
I
19
want to make sure you don't limit people
20
either too much in terms of what they are
21
able to do, because if it is an IND, then
22
really, it's research to some degree to get
321
1
there.
2 DR.
NOGUCHI: Well, we are
3
examining a lot of different things.
From
4
one aspect, if we could standardize all the
5
requirements you'd need in order to get a
6
reasonable outcome, that makes it
7
administratively simpler to license the
8
thing.
9 We just say these are the
10
standards and you make sure you meet all
11
these standards, and we will inspect you and
12
that's a uniform set of requirements.
13 For cellular products, we also
14
know that these will evolve. We
think if we
15
impose standards per se like that, that will
16
inhibit the field, because most people will
17
say, well, I just want to make sure I can
18
deliver this.
19 Those who are then wanting to
20
advance the field may not be able to do that
21
in a facile way. So we're trying
to see
22
what are the different ways we can do it.
322
1 Our current licensing procedure is
2 very adequate to
address the issue of two
3
different groups do it slightly different.
4
They get similar results. We can
license
5
each separately for the particular method of
6
preparation. So our current
licensing
7
supports that.
8 But the other thing is do we want
9
to license 28 centers separately, or 57, or
10
in the case of blood banks, well over 1,000?
11
The answer is we don't
think
12
that's the most efficient, but we don't want
13
to just have one center that does everything
14
either.
15 DR. RAO:
I'll let Keith make a
16 comment.
17 DR. WONNACUTT:
Yes, just to kind
18
of add to what Phil has been saying, at this
19
point, the me-too idea isn't going to apply
20
to a license application. So
we've been
21 hearing from some
of the people here that
22
efficacy data at some centers who all use
323
1
the same release criteria currently is
2
different than at other centers.
3 So what our first conclusion was
4
before Darin stopped us and brought us back
5
was, well, if they have a good validated
6
potency assay, then comparability is not an
7
issue.
8 But what you're saying by saying
9
that is that if they do the potency assay,
10
we assume that the clinical outcomes are
11
going to be the same at those two centers.
12 What we have been hearing is that
13
the clinical outcome isn't the same at those
14
two centers. So can we make
assumptions
15
from the assay to predict what's going to
16 happen in the
clinic, or do we need clinical
17
data to compare the two?
18 DR. SHERWIN:
One piece of the
19
outcome is the islet itself. But
there are
20
many other pieces of the outcome when you're
21
dealing with patient care.
That's a very
22
much more complicated story. So
324
1
consequently, this is one piece of the
2
puzzle, not the whole puzzle.
3 So you know, you can standardize
4
the islets all you want, but the ultimate
5
result is the clinical outcome, and that's
6
not simply based on that. So
that's the
7
complicating feature.
8 DR. WONNACUTT:
It's just a matter
9
of are you comfortable when you talk about
10
the clinical outcome in saying that we have
11
X potency assay at both centers,
and we're
12
comfortable saying that because X potency
13
assays were conducted at both centers, we
14
are absolutely sure that the two outcomes
15
are not related to product; they are
16
definitely related to one of those other
17
factors that you are talking about.
18 If that's true, then our
19
comparability is good. But if
that's not
20
true, then we have to wonder, is just saying
21
one assay is sufficient for comparability?
22 DR. SHERWIN:
Well, it's clearly
325
1
not totally true. In relative
terms it's
2
reasonable, the best I would suspect that we
3
can do at this point.
4 But clearly, the outcome is going
5
to depend on the immunosuppressive agents,
6 the individual
you put it in, how
7
insulin-resistant those people are that you
8
are dealing with. There will be
many
9
factors that are going to ultimately
10
determine it.
11
But I think given the
state of the
12
art, not being the world's expert in this
13
area, but I would think it's the best we can
14
do right now. And it's a
reasonable thing
15
to say that they are relatively comparable,
16
even though there will be things that we'll
17
find out to prove that it wasn't quite true.
18 DR. RAO: Go
ahead, Dr. Harlan.
19 DR. HARLAN:
I just think that we
20
don't know and we probably never will know
21
for sure, but having better assays would
22
give us better information to address
326
1
Keith's question.
2 I just want to also support what
3
Dr. Allan said. I think that
there's a
4
danger in saying that Center A, B, C or D
5
did well. And I will say that in
the
6
patients that we transplanted, we saw islet
7
function in every single one of them.
So
8
I'm not speaking from a position of profound
9
weakness here. I think our
islets were
10
good. But I think it's also dangerous to
11
think that because a center has a certain
12
level of success that that's the way to
13
isolate islets.
14 I think anybody that does this
15
recognizes that we don't know the best way
16
to isolate islets yet. By saying
this is
17
how it should be done, we limit progress.
18 So I really like saying, agreeing
19
upon an in vivo assay that may not be
20
perfect, but it's the best we got.
And the
21
best we got, based upon what Dr. Hering told
22
us this morning, other than a patient, and
327
1
in patients, you've got variables and
2
immunosuppressive drugs to worry about and
3
in the non-skid mouse, you don't.
4 Dr. Salomon made this point, too;
5
it's not a perfect assay, but it's the best
6
we've got of assessing the islet quality,
7
and we should let centers come up with their
8
own in vitro assays and their own way of
9
isolating islets, but apply that test and
10
then transplant the islets.
11 You know, regardless of that test,
12
I'm just looking for a test where we can
13
validate islet quality.
14 DR. RAO: Dr.
O'Neill.
15 DR. O'NEIL:
As long as we are
16
talking about a few centers and a few
17
numbers of assays and a few numbers of
18
patients, I can guarantee that everything
19
will be comparable.
20 I was truly dumbfounded to read
21
the word in an FDA document, because, of
22
course, you have been in the midst of
328
1
controversy about how to establish the
2
comparability of therapies. It's
a very
3
challenging thing to be satisfied with
4
evidences of comparability.
There isn't
5
going to be any comparability here, if you
6
try to do statistical tests.
7 Otherwise, it's strictly every man
8
for himself; every person's idea of
9
comparability is all in the eye of the
10 beholder.
11 DR. RICORDI:
On that
12
comparability, I think that the major
13
challenge we have is the lack of
14
comparability of the initial product.
So
15
each pancreas is different, and there are
16
differences in the processing that are done
17
that for us are comparable tests because it
18
is a procedure. It's like in a
surgical
19
procedure. I'll say I'll put
this stitch
20
here now, or within this timeframe, or I
21
don't put it. If you have
bleeding, you
22
stitch it at that point.
329
1 At the same time, it's not that
2
you say I diluted 12 minutes, whether I
3
saved three islets or not. There
are
4
hundreds of features that are part of the
5
experience factor in this kind of procedure.
6 But I would be very worried if we
7
hold progress at some institution or find
8
some way to come together, put together the
9
results from different clinical experience
10
in all these trials where you have a high
11
level of success and say let's move to the
12
next step.
13 I think comparability in this
14
sense is the fact, for example, we meet
15
every six months for three or four solid
16
days to see is there anything new at another
17
institution that should be included and be
18
part of a comparable or better procedure.
19 That is how
James tomorrow will
20
present, like we don't do any more the
21
Edmonton protocol as it was in "The New
22
England Journal" presented with fresh
330
1
islets, because we think the first reason
2
for potential failure is transplanting a
3
product immediately instead of observing
4
over two days.
5 So that's why now Edmonton,
6
Minnesota and ourselves and others are
7
transferring cultured product.
Then if we
8
have data on the PFC, a preservation that
9
improves ÄÄÄÄ now that of course is not
10 comparable.
11 It wouldn't be comparable in a
12
strict comparability for drug manufacturing,
13
because you are slightly changing the
14
procedure as you go forward. But
if your
15 outcome keeps
improving, I would hope that
16
we can still pool the data from the three
17
centers, for example, and say some were with
18
fluorocarbon and somewhat was EOW (?) but
19
they were all meeting the end product
20
criteria and they were all successful in
21
reversing diabetes. That would
be pretty
22
satisfactory to me.
331
1 It would be very dangerous, a very
2
slippery slope if we start saying we don't
3
move forward until we train the last of the
4
centers to have the same degree of success,
5
given what we know based on the last two
6
decades of experience.
7 Because then we can keep doing, as
8
one of the general criticisms on research in
9
general, when you stick too much on basic
10
science, you can spend your life asking new
11
questions and being very, very sure that you
12
have your final 100 percent sure question
13
and never move the field forward.
14 So that the
major concern is like
15
safety, make sure you have something viable
16
and can it be done in a selected institution
17
first, including as many people as possible
18
as we are capable, but going in one
19
direction.
20 MS. LAWTON:
Can I just make a
21
comment on that? I completely
agree with
22
you. I think it's really
important that we
332
1
can continue to learn, continue to improve.
2 But unfortunately, we also have a
3
current standard that we work to under BLAs,
4
and that is the expectation that we've done,
5
somewhat at least, optimized our process and
6
that we then have a controlled and
7
consistent process.
8 Any changes we want to make to
9
that process, we have to understand and do
10
the comparability studies and present that
11
and ask the FDA whether they agree in order
12
to make that change.
13 We can't keep tweaking our
14
process. I just raise that
because that's
15
the current standard that we have for other
16
approved cell therapies. I do
think it's
17
important that we have a level playing field
18
as far as what we do in the future.
19 At the same time, I'm very
20
sympathetic to how we continue to improve
21
our products and the outcomes.
22 DR. WEBER:
We certainly 100
333
1
percent agree with what Ms. Lawton just
2
said. The point I think Dr.
Wonnacutt made,
3
you know, experimental procedures lead to
4
experimental products, not licensed
5
products.
6 INDs can continue to exist in this
7
field that will be developing these
8
innovative therapies. When the
submission
9
data, both safety and efficacy, is
10
developed, that can then be supplemented or
11
no BLA can be submitted.
12 So we are not killing the field in
13
any way. But when we do talk
about a
14
licensed product, we are talking about a
15
fairly well, rigorously established
16
procedure. Again, within some
limits, we
17
are willing to be somewhat flexible, but
18
it's going to have to be well-defined.
19 DR.
HIGH: My comment is really a
20
question about the process, and it's a
21
follow-up on what Alison Lawton said around
22
what Dr. Ricordi said.
334
1 It sounds to me from what I've
2
heard today that clearly everyone agrees
3
that this needs further research and
4
refinement. Yet the people at
these three
5 centers could probably, if
they wanted to,
6
sit down and agree on standard operating
7
procedures for organ procurement, for organ
8
processing and for choice of potency assays
9
and that sort of thing.
10 So they could submit something and
11
it would have similar outcomes at the three
12
institutions. Yet it's very
important for
13
what they want to do that once that BLA is
14
submitted, they still have the opportunity
15
to meet every six months or once a year or
16
something like that to pool results on some
17
of their work, which would apparently
18
continue to be done under an IND rather than
19
under the form of the BLA.
20 So is that the paradigm that we
21
are working with here?
22 DR. NOGUCHI:
I think that that's
335
1
one possibility. I think that in
the very
2
broadest sense the purpose of a
3
comparability protocol is for, a priori,
4
assuming you have first a licensed product,
5
is then to establish a process that you
6
review with FDA and you will say we believe
7
that these are the kinds of test results
8
that will suggest to us that we can actually
9
do things a little different but still have
10
a product that meets the same or better in
11
terms of patient outcome.
12 So I think we unfortunately in FDA
13
often say things that we are not always
14
articulate about what we mean.
15
Comparability in the sense of a
16
comparability protocol is to allow a company
17
or a sponsor or a BLA holder to be able to
18
continually refine their process, improve
19
quality product, and still have a reasonable
20
expectation that it will perform as it has
21
in the past in terms of clinical outcomes.
22 We were also talking about product
336
1
comparability here in the sense of among
2
different centers making a similar product.
3
What are the assays and what are the
4
outcomes that can help us say that we can
5
actually pool across centers who have
6
independently submitted BLAs in terms of
7
outcome data? Can we do
that? Certainly
8
within the course of a BLA
approval we
9
expect that you will be able to continue to
10
refine, and that's what the protocol is all
11
about, and there may even be major changes
12
to the protocol, not necessarily under IND.
13
I think that becomes an individual case by
14
case evaluation.
15 But the intent is really to try to
16
in a way get the best of both worlds, use
17
what works, teach those, us especially, what
18
is it that prevents others from not
19
achieving the same success, extend that
20
knowledge to them, be able to bring them up
21
and not necessarily at the same pace but
22
also to not shut them out as well and to not
337
1
shut out future refinement of the process.
2 DR. RAO: Dr.
Harlan?
3 DR.
HARLAN: I agree very much
4
with what Dr. Noguchi just said but I want
5
to extend it some and that is that I have
6
zero interest in limiting centers that have
7
success with this now, none. I
think the
8
motivation is that even though they have
9
success, everybody, even the three most
10
successful centers would agree it's not
11
fully mature yet, either, and I don't want
12
to see any policies that are implemented
13
inhibit a new center from being able to jump
14
into the field, including with something
15
completely different than the way that the
16
three centers are doing but maybe better.
17 And Jonathan Allan made that
18
point. So let centers with
success continue
19
what they're doing and do refinements on it
20
but let's try to figure out rules that will
21
allow somebody else to jump in with a
22
reasonable chance of success, with a
338
1
reasonable assurance that their islets are
2
of good quality, and let them try something
3
completely different. Science
works best
4
when you do rapidly reiterative experiments
5
to try things as long as safety and
6
scientific integrity are ensured.
7 DR. RAO: So
let me ask me ask you
8
a question here. Did you feel
from what in
9
the conversation was that there was any
10
suggestion which you specifically felt would
11
inhibit anybody from performing --
12 DR. HARLAN:
Dr. Ricordi, I'll ask
13
him exactly what he meant, but he said he
14
didn't want to see their progress hindered
15
by rules that we stipulate, and I certainly
16
would hate to see that happen, too.
I don't
17
want to see any progress.
18 DR. RICORDI:
I would hate to have
19
suggested at any point that islet
20
transplantation should be continued only at
21
centers that have 90 to 100 percent success
22
rate. That would be criminal and
farther
339
1
away from what I say. All these
new trials
2
and these wonderful new techniques can
3
continue with the regular structure.
Some
4
methods are funded by their institution
5
submits an IND, and you show the
6
compatibility. What I'm worried
is like
7
let's not block the ability to deliver this
8
very limited potential small group kind of
9
therapy to patients and be reimbursing fast
10
so that we can continue the next level of
11
research. When we wait for the
magic new
12
method to come out I don't know from where
13
or to have all center in the world fully
14
trained and fully comparable so that they
15
are equal and just as good because in the
16
meantime we're doing a disservice to the
17
patients we try to serve.
18 DR. RAO: I
don't think that
19 should have been the
sense from the
20
discussions. Abbey.
21 MS. MEYERS:
I'm still struggling
22
with the concept of academic groups who
340
1
don't have regulatory affairs personnel
2
applying for a BLA. I mean,
Merck and
3
Pfizer have problems applying for a BLA.
4
How do you expect that to happen?
5 DR.
NOGUCHI: Well, we will see.
6
I will say that the particular concern that
7
you voice about regulatory affairs expertise
8
is one that's well recognized certainly by
9
FDA as well as by NIH in general as well.
10
One of the centers at the NIH, the National
11
Center for Research Resources, in fact
12
Dr. Nasik is here representing that group,
13
just had their five-year strategic planning
14 several weeks
ago. The regulatory affairs
15
component part of it was one of the largest
16
needs that were identified along with
17
expertise in product development for cell
18
therapies, biotechnology, and so forth.
So
19
there is a clear recognition that if this is
20
to happen at a university setting that there
21
needs to be a fairly rapid buildup of the
22
necessary regulatory structures for that.
341
1 It is our experience that many
2
universities are striving to rise to that
3
challenge and that they're willing to commit
4
some amount of funding to developing that
5
infrastructure. It is also true
that in
6
many respects if we could convince the
7
pharmaceutical industry that in fact these
8
particular therapies are in fact reasonable
9
for them to pursue that would be another way
10
to get this moving.
11 I think Dr. Lawton can address the
12
fact that industry is perhaps a little bit
13
in the sense of trying to see if the things
14
at the end actually work because to create
15
it into a fully pharmaceutical industry does
16
take commitment and devotion of efforts and
17
resources from the industry as
well to do
18
that.
19 We think at FDA, however, that the
20
promise of this field is sufficient, that we
21
are willing to spend our own personal
22
investment in terms of
education. We, of
342
1
course, are a little bit on the lean side in
2
terms of funding but it's certainly our
3
intellectual property that we've developed
4
over a number of years. We're
willing to
5
share that.
6 We believe that academic
7
institutions can even at this time get part
8
of the way toward learning how to produce
9
facilities that can actually meet not
10
necessarily millions of patients but perhaps
11
on the order of hundreds of patients, which
12
is what we're talking about here, and that
13
it's a reasonable opportunity to try to see
14
if it really can be done.
15 We believe that it can be done in
16
academic institutions. We would
agree that
17
an academic institution can't really deliver
18
it to the hundreds of thousands of people at
19
any one time yet. Perhaps that's
in the
20
future, perhaps not, but for right now for
21
this field we think it's a viable
22
possibility.
343
1 MS. LAWTON:
If I just make a
2
comment, coming back to the issue of
3
comparability, I mean, ultimately when we're
4
talking about these three centers we're
5
talking about the fact that they have a very
6
good success record clinically so that's
7
obviously very important. So you
could
8
argue that actually clinical is the key
9
thing for comparability as a result.
10 However, it also is a question of
11
what are you approving. Are you
approving
12
the product or are you approving the regimen
13
that gives you the clinical outcomes?
And
14
that's the dilemma we're facing here in
15
trying to give you advice on what the
16
comparability should be.
17 DR.
SALOMON: I think that's
18
right. That has been the circle
that we've
19
been dealing with but I think there's an
20
answer to that. I mean, in the
end to me
21
the clinical results are critical but not
22
what we're approving. They are
the way of
344
1
saying that a center applying for a BLA is a
2
value by the fact that they've cured a
3
significant number of patients and maybe
4
what we ought to be rather as having is a
5
conversation about how many patients, for
6
how long, and with what measures should
7 constitute a
clinical cure of diabetes with
8
islet transplantation if indeed that's going
9
to be one of the ways of validating a
10
program for BLA.
11 But the key thing here is that
12
they have a surrogate potency assay and
13
that's what should be the lot release
14
criterion. That's what makes the
data to
15
day functioning of these programs going
16
forward with a BLA possible.
Otherwise it's
17
not going to be practical.
18 DR. KURTZBERG:
If you have a
19
package that is the center, gets the organ,
20
produces the cell, does the transplant,
21
takes care of the patient post-transplant,
22
and has an outcome that's one thing.
But if
345
1
you have a production facility that supplies
2
cells, which is your product, to multiple
3
centers, then they're not going to be in
4
control of the medical care that surrounds
5
that patient which may very well influence
6
the outcome. So you're really
talking about
7
two very different scenarios.
8 DR. SALOMON:
My whole point is
9
that we're really talking about two
10
different things and that's why you don't
11
link the two in the end. Maybe I have to
12
put it into something practical because I
13
can't follow it any more.
14 So I'm a center and I want to do a
15
BLA for islet transplantation.
What do I
16
need to do to do that, right?
That's the
17
question that everyone's sitting here trying
18
to figure out first. So what I'm
thinking
19
we've all been saying all day is that what I
20
need to do is develop a series of standard
21
operating procedures done in a CGMP-like
22
facility, show that I'm getting consistent
346
1
product.
2 I need to develop a potency assay,
3
or a set of potency assays I think is more
4
appropriate at this early point, that I then
5
show that with this product and with these
6 potency assays that
correlate with it, given
7
a clinical regiment that I have control over
8
whether that's through a collaborator or at
9
the same institution or a collaborator a
10
hundred thousand miles away, it doesn't
11
matter, and I get X-number of cures, which
12
we haven't talked about, but let's say I do
13
with good measures of normal glycemia.
Now
14
I get my BLA.
15
Now, I can collaborate
with my
16
product with 20 other centers the next day
17
as long as my potency assay is good and
18
those other centers might not have the same
19
clinical success as I demonstrated in the
20
process of getting my BLA. But
that doesn't
21
affect me any more. I have a
BLA. I know
22
that if I do everything I said I was going
347
1
to do I have a good product.
2 DR. BLAZAR:
Maybe we're already
3
there.
4 DR. SALOMON:
I think so.
5 DR. KURTZBERG:
But how do you
6 know it wasn't
the package and not just the
7
product especially if your product in other
8
people's hands, which is still the same good
9
product, doesn't produce the same results?
10
Then it's partly related to your package and
11
your care and the center expertise and --
12 DR. SALOMON:
I'm sure it is.
13 DR. KURTZBERG:
So maybe some of
14
these more specialized therapies need to be
15
approached that way rather than there's just
16
a product in a box and it can go anywhere.
17 DR. RICORDI:
We know that it is
18
that product in a way. I mean, I
should not
19
say we're sure but based on our limited
20
experience we believe it is the product
21
because we obtain equal results comparable
22
or better when we ship this product to
348
1
Baylor College of Medicine in Houston where
2
they have now 13 patients all with
3
functioning islet transplants and is a team
4
that never saw an islet before but is
5
definitely a multiorgan transplant, very
6
competent team in managing
7
immunosuppression, so my hope is that we can
8
define "criteria" so that the centers
9
created also by NIH, the regional center,
10
can distribute islets to other centers that
11
don't have the capability of processing
12
islets because it is a completely different
13
expertise provided that they can guarantee
14
some elements of competence locally, that
15
they have a multiorgan transplant ÄÄÄÄ or
16
transplant surgery team that can deal with
17
eventual problems that may arise, so like
18
now we've been asked besides Houston or so
19
Dallas wants to start, Oklahoma City, in
20
centers everywhere in the United States.
21 DR. SHERWIN:
Can I follow up on
22
that? The problem we're getting
into is the
349
1
issue that I brought up originally.
It
2
depends on what you're looking at.
If you
3
want to look at the whole package you've got
4
to see the whole package to evaluate it. We
5
can't evaluate it until we see the whole
6
package.
7 If you want to focus on the islet
8
and its specific potency I think we've
9
addressed that at least as best we could
10
now. So if we're going to get
into the
11
clinical pieces around that I think are very
12
important I think you have to do that with
13
the data. That's all.
14 DR. RAO:
Well taken.
15 DR. AGBANYO:
I was just going to
16
follow up on a comment that was made on this
17
end. This situation is quite analogous
to
18
what happens in the transfusion medicine
19
community in the sense that you have blood
20
banks that manufacture platelets, blood
21
components, and then the products that
22
transfuse at the hospital level.
Now, you
350
1
set specifications for your platelets and
2
your red cells but you have absolutely no
3
control over utilization at the hospital
4
level. So there are some
similarities here.
5
That does not mean that you ignore the
6
product specification side of it just
7
because you have no control of what happens
8
on the hospital end.
9 DR. BLAZAR:
I'm wondering from
10
both the three centers and maybe Dr. Noguchi
11
what are we missing in terms of going ahead
12
with the BLA, assuming that there's some
13
clinical cure rate to be defined tomorrow,
14
because I don't understand where the
15
limitations are based on what Dan said and
16
based on the discussion this morning.
What
17
are the missing elements here?
There'll be
18
some cure rate which we'll see in whatever
19
those numbers are but what is the impediment
20
right now that's stopping the centers from
21
going forward with the BLA and the FDA from
22
evaluating that application?
351
1 DR. RAO:
Allison, do you want --
2 MS. LAWTON:
I was just going to
3
make a comment and follow up to
4
Dr. Salomon's comment around the product and
5
the comparability versus the clinical
6
outcomes.
7 I agree with you about once you
8
have a finished product then it's up to
9
other people if they want to take that and
10
look at how they use that clinically;
11
however, what I'm hearing is how do you pool
12
clinical data from three different sites and
13
understand therefore how you approve a
14
product so it almost is you're approving the
15
product but you need some kind of algorithm
16
in your labeling for people to understand
17
what makes that a successful product.
18 If then people want to take that
19
product and use a different algorithm for
20
the regimen of immunosuppressants, how they
21
deliver the cells, all of those other
22
things, that's investigational again and
352
1
they need to understand how that's going to
2 impact the
outcome.
3 DR. SALOMON:
I just want to say
4
for the record I totally agree with that.
5
That was my point.
6 DR. RAO: Go
ahead, Doctor.
7 DR.
NOGUCHI: I think that the
8
question in terms of what's holding us up I
9
would not want to answer directly; however,
10
tomorrow we will be getting into the
11
thornier clinical questions that are of some
12
major bearing.
13 We've heard, certainly, and seen
14
in the public's literature some results that
15
are preliminary but are very enticing.
We
16
certainly are nowhere convinced that we have
17
the community sense of whether a year of
18
insulin dependence is adequate, whether it's
19
seven years, whether you need normalization
20
of glycemia. What are all the nuances
above
21
and beyond a simple end point of you don't
22
need to take insulin any more?
Is that all
353
1
you need? And I think for some purposes
you
2
may say yes but then you have to start to go
3
to the intended use.
4 For those that are brittle
5
diabetics I think that any relief from the
6
wildly out of control situation that they
7
have is a step forward but if you're a Type
8
I but very stable diabetic who takes insulin
9
at a minimal use, the risk and the benefits,
10
what are those there?
11 So I think we do have a lot of
12
discussion yet to be done about what things
13
are going to be important, not just in a
14
very narrow, small population, but in a
15
larger population. That doesn't
mean that
16
we can't begin at some point to evaluate
17
licenses for a very specific population.
18 In the meantime we know. I mean,
19
you can just look at the figures.
We know
20
at this time based on what we know a minimum
21
of one and probably two pancreases are
22
needed to generate islets for one successful
354
1 reversal. If you look at whatever limited
2
statistics are available, if you do a
3
pancreas transplant, a single organ
4
transplant, that's one pancreas for a cure
5
versus two pancreases for a cure.
6 What are the limits of what we
7
need to do in terms of how would you make
8
that judgement as to when you want to spend
9
two organs versus one organ and with two
10
different modalities of treatment?
So I
11
think those are all the other issues that we
12
are by necessity having to deal with.
13 We are starting with a source
14
material that is a very precious resource
15
that this government has decided is a public
16
resource and we would derelict in our duty
17
if we didn't consider that as well as
18
whether specific groups are being treated
19 because we have
to judge that in the case of
20
the entire population as well.
So the
21
answer is we're getting close for a specific
22
end point, but for the more global end point
355
1
of islet transplantation and other cellular
2
therapies this is a beginning.
3 DR. BLAZAR:
I wasn't suggesting
4
that it was necessarily an approved process
5
but these centers could presumably come
6
forward with a narrow definition of
7
"patients," procedures, with SOPs and the
8
validated bioassays. Presumably
the centers
9 feel strongly that
their data is compelling,
10
and that could be then evaluated, and this
11
committee and other input could help set the
12
standard for the field.
13 But I'm still confused as to what
14
is limiting these centers from coming
15
forward with an application with a narrow
16
definition for being evaluated.
Maybe
17
that's --
18 DR. RAO:
Maybe I can ask --
19 DR. SALOMON:
Can I comment? I
20
mean, I have an answer for that.
Phil may
21
not like my answer but it's easier, maybe,
22
coming from me. I mean, these
guys have
356
1
been working very hard for the last three
2
years since 2000, the first time we had a
3
meeting, to come up with an integrated set
4
of data that would allow them to do a BLA,
5
and I think part of the reason they're
6
sitting quietly and listening here is to say
7
what exactly do you want us to show you for
8
a BLA.
9 So I think
what you're doing here
10
is helping move the field forward.
I think
11
the reason there's no BLA right now is that
12
these guys aren't sure what to provide you
13
and I think you're giving them a pretty darn
14
good idea.
15 DR. RAO: Let
me ask the FDA a
16
question here. So I think some
of this we
17
are repeating ourselves in terms of
18
emphasizing what we can and cannot compare.
19
Do you feel that the issue has been
20
adequately considered in terms of
21
comparability?
22 DR. WEBER: I
think yes. I think
357
1
all those questions have been excellent
2
because they're all important issues and I
3
think it will be behooving on who is going
4
to be able to provide that data in a way
5
that's interpretable. It should
demonstrate
6
comparability to address our regulatory need
7
to have a well-defined manufacturing process
8
and I think you've brought up the issues
9 that are
important to consider.
10 DR. RAO: Dr.
Harlan, you had a
11
comment?
12 DR. HARLAN:
I want to support
13
what I heard, I thought, Dr. Sherwin and
14
Dr. Noguchi say. I think it's premature yet
15
to say anything about the clinical outcome
16
because we haven't discussed that yet.
But
17
on the product I absolutely agree with what
18
Dan Solomon said but my diagram of what he
19
said is that all things led through a
20
potency assay and I think that we don't know
21
what that is yet.
22 If you separate the clinical
358
1
outcome what's the potency assay?
We need
2
that.
3 DR. RAO:
Absolutely, and I think
4
that was the sense that came through which
5 we tried to
summarize before that we don't
6
have good clinical predictors, and we only
7
have a potency assay which is specific to
8
the centers which work with it.
9 DR. HARLAN:
If you remove patient
10
outcome let's not talk about that yet.
Just
11
remove that from the equation.
What are the
12
potency assays that have been validated to
13
predict outcome? It's a circular
argument.
14
That's the trouble.
15 DR. SALOMON:
It is circular.
16
That's what I was going to tell you.
I
17
mean, basically here what we're saying is
18
there are measures for the biological
19
functionality of islets. They
function in
20
vivo to maintain glucose homeostasis.
21
There's only two in vivo models.
One is in
22
animal models and there are like a dozen
359
1
under that heading. And there
are human
2
patients with, of course, a dozen
3
modifications of that because the
4
immunosuppressive drugs and the selection of
5
the patients have all kinds of variables
6
measuring function objectively.
But either
7
way it's still biological functionality, so
8
any potency assay has to be validated in
9
animal models and in human clinical trials.
10 DR. RAO: So
do you need me to
11
summarize or is this okay? We
have covered
12
the questions.
13 DR. NOGUCHI:
The only question I
14
don't think we really covered adequately was
15
number two, and I'm not sure if you lumped
16
that into number one but if you want me to
17
I think we might want more discussion on
18 that. Is that okay?
19 DR. RAO: I
suggest we take a
20
short break and then come back and discuss
21
that question.
22 (Recess)
360
1 DR. RAO: So
we're going to try
2
and see if we can address this difficult
3
second question and I'm going to try and
4
resummarize it and see if we can go from
5
there. So I'm going to rephrase
this and
6
see if it makes sense.
7 Here we have pancreata and, as you
8
heard from Dr. Noguchi and from others, in
9
general each pancrease constitutes one lot
10
and you're going to use either one or two
11
pancreases for one to get enough islets for
12
one patient. And each time you
harvest
13
islets it may be slightly different because
14
the source material that you get is slightly
15
different in terms of the age of the
16
patient, in terms of the amount of fibrosis
17
it may have, the history of the patient, the
18 donor, that you got them from.
So there
19
might be differences.
20 On the other hand if you set up a
21
GMP facility and you set up with standard
22
SOP protocols how do you define a readout
361
1
which will say that you've standardized a
2
protocol that you can use and if you make a
3
change whether that's an acceptable change
4
or whether that requires that you now call
5
it a completely different product.
It's an
6
issue that we should consider in some
7
fashion.
8 Go ahead, Richard.
9
DR. MULLIGAN: I was just going to
10
say I don't think they mention that to us.
11
I mean, wouldn't that be a necessary part of
12
this to, like, have heard hours before
13
whether there was any variation and what it
14
was precisely?
15 DR. RAO:
It's a very good point
16
and I actually asked Dr. ÄÄÄÄ I said I'd
17
probably have to ask him a couple of
18
questions on this when we did that.
19 DR. WEBER:
I'm sorry. I was
20
looking through some slides that we did talk
21
about this issue a little bit earlier, the
22
issue being that, as Dr. Rao mentioned,
362
1
clearly the donor organs are going to be
2
variable based on all the criteria we've
3
talked about, the concern being for a
4
manufacturing facility for a GMP
product, a
5
licensed product, you need to have a fairly
6
well-controlled process and there is some
7
concern that because of, again, the source
8
material being variable that you have to
9
have a little bit of flexibility in how you
10
process the islets. Obviously
you
11
enzymatically and mechanically disassociate
12
them.
13 Some centers are using things like
14
DNase, of course, Pefablock, antioxidants,
15
other vitamins, and concern being that we at
16
the FDA when we receive a license
17
application would expect that if we're going
18 to use those
various additives there be a
19
good reason for doing so. I
mean, there
20
would be justification. So we
would expect
21
that based on the manufacturing data that's
22
being collected currently under IND that if
363
1
you wanted to add maybe an antioxidant or
2
something that you would have collected data
3
showing that, first of all, that antioxidant
4
actually has a benefit for the final product
5
or optimizes the yield or does something
6
beneficial.
7 And so what this question's trying
8
to get to, I think, is the
sense, again,
9
almost a guiding principle for GMPs, that
10
you would have to have limits established
11
for whether he would or would not use that
12
reagent, under what considerations you would
13
use that reagent. We wouldn't
want you to
14
willy-nilly or randomly add a reagent and
15
then not know whether it has any beneficial
16
effects. We don't want to be
adding things
17
unless we need to be adding them.
So that's
18
an issue we're trying to get to.
19 DR. MULLIGAN:
I would think the
20
question is what is the variation and also
21
in many of these organ digestions if there's
22
more fibrosis you may need to add more
364
1
enzyme or something and the question is is
2
that simply going to affect the yield.
And
3
so you know have to know of the variation
4
that you see is, I mean, if you have to use
5
exactly the same approach rather than using
6
more of something does that in an
7
unacceptable way reduce the yield you have
8
to a point where it's not a good
9
manufacturing process. I just
don't think
10
we have the info on it.
11 DR. WEBER:
Let me just add a
12
little bit. We obviously don't
want to
13
restrict that but we want to have some
14
limits, so for an ideal pancreas you would
15
add X-amount of ÄÄÄÄ for example, now giving
16
the variable ÄÄÄÄ X-amount of fibrosis that
17
could be varying from organ to organ.
18 There should be some predefined
19
limits. You don't want to add 1
kilogram of
20
ÄÄÄÄ for a situation that would have been
21
handled by 1 gram, for example.
And this
22
data is being collected so we're asking the
365
1
committee whether you agree that data should
2
be analyzed and evaluated and some criteria
3
established for the licensed product.
4 DR. RAO: Go
ahead.
5 DR. HARLAN:
Well, just speaking
6
from a fairly inexperienced islet transplant
7
center but nonetheless did a lot of
8
isolations, there's a lot of variabilities
9
that the donor brings to the process and a
10
lot of variabilities that occur during the
11
process. A lot of the digestion
dynamics
12
are based on visual observation on how the
13
digestion is going and you can react to
14
those observations by adding something like
15
DNase or Pefablock during the procedure to
16
minimize those effects.
17 I think what the FDA is asking is
18
can we establish criteria to determine when
19
those manipulations can be adjusted and how
20
broad should those criteria be to allow
21
enough flexibility in the procedure for an
22
effective isolation but to yield a quality
366
1
product in the end.
2 DR. MULLIGAN:
Well, I think
3
another way to look at the question is
4
essentially did you develop a good method;
5
that is, in developing a method one
6
criterion would be that it actually works
7
under as many real-life conditions as
8
possible. So that's what I don't
have any
9
sense of.
10 I mean, it may be that the way the
11
work has gone up to this point in academic
12
centers is you fiddle with this, you look at
13
it, and you get it, okay. But if
you're in
14
a company you might say we need one process
15 that may be
overkill but in the normal
16
pancreases does okay, doesn't overdo them,
17
but does cover those that are hard to
18
digest, that type of thing.
19 And so it would be good to have
20
experts tell us how much do you have to beat
21
on these things in different ways and how
22
people actually given the kind of boring
367
1
attention necessary to actually work out a
2
real life process. It's like I
remember
3
with tumor vaccines and irradiating cells.
4
The cells have different susceptibilities to
5
radiation so you choose a dose of radiation
6
that's good enough that covers all the guys
7
that don't want to die.
8 DR. RAO: Go
ahead.
9 MS. LAWTON:
If I can just add to
10 that because I'm
coming from the industry
11
perspective where the company that I work
12
for have a number of cell therapies approved
13
and so I know what we've had to go through
14
to get those products approved and I think
15
the paradigm from the FDA is that before you
16
start your pivotal studies you understand
17
what the process is and therefore what your
18
product is that you're studying to
19
understand the efficacy and the safety, the
20
outcome, of that product.
21 So this whole question, I mean,
22
obviously there are certain ranges of things
368
1
that you can do within the process that you
2
want to understand and that can be part of
3
understanding it in your clinical outcomes.
4
But my question is because we don't have the
5
clinical data how much of this is a
6
retrospective look at things that were done
7
and how much clinical data do we have on
8
each one of the different parameters that
9
might be introduced in the process and
10
therefore do we really have the right data
11
to be able to make those assessments?
12 DR. RAO: So
let's ask people from
13
the center and maybe the question is when
14
you harvest pancreatic tissue how variable
15
is the process when you do it or is it
16
pretty standardized as it is and what kind
17
of information do you collect.
18 DR.
HERING: I think as far as I
19
can tell we follow our SOP to the letter and
20
if the SOP allows us to operate within this
21
framework then we do so but we don't change
22
the process. In the past three
years we
369
1
have not changed our protocol once and if we
2
have to change for whatever reason then we
3
complete a deviation protocol and file it
4
and that is how we do it, and I am not aware
5
of any established procedure that tells me
6
if I deviate from that protocol this will
7
result in a better, higher yield.
8 For example, if the donor is now
9
45 and is female and is from this state I
10
have heard many people argue this could
11
help. I have never seen the
data. That is
12
why we follow one SOP every single time and
13
that is what we do.
14 DR. RAO: Dr.
Harlan, maybe I can
15
ask you this or anybody from any of the
16
other centers. Is there huge
variability in
17
your SOP when you harvest islet sets?
18 DR. HARLAN:
I agree with what
19
Bernard said and I'm sure what Camillo will
20
say and what James would say, that the SOP
21
has within it appropriate leeway to let you
22
exercise judgement but that's not to say
370
1
that there's considerable room for judgement
2
during the islet isolation process.
You
3
couldn't put a trained monkey in to do these
4
isolations. That's for sure.
5 DR. RAO: But
there are ranges
6
built into the SOP.
7 DR. HARLAN:
Well, I don't know
8
the Minnesota SOP but the process involves
9
instilling a collagenase enzyme and of the
10
people in this room I've done far, far less
11
islet isolations than any of those three or
12
than Dr. O'Neill or others but you instill a
13
collagenase. Then you wait until
you start
14
to see islets coming out of the digestion
15
chamber that are just right, and that's when
16 you make the
conversion. But how do you
17
quantitate that? There's
considerable
18
judgement in some of those calls.
19 MS. LAWTON:
So can I just ask
20
does that mean it's operator-specific, in
21
which case is it going to be site-specific
22
and therefore what do you approve?
371
1 DR. HARLAN:
I really think the
2
person to ask is either Bernard Hering or
3
Camillo Ricordi because both of them have
4
done probably 10 times more isolations than
5
I've ever done.
6 DR. RAO: Is
there anybody from
7
any of the other 10 centers who would care
8
to comment on this?
9 MS. REEMS:
Actually, we're more
10
like a centralized center. We're
at the
11
blood center and we are collaborating with
12
three other clinical sites and we don't have
13
control over the actual procurement nor do
14
we have as much control over the actual
15
infusion of the product. So I
know that
16
when one individual is doing the procurement
17
that they're procuring it more along the
18
lines of what's been recommended by Edmonton
19
but if that individual isn't procuring it's
20
difficult for me to provide
oversight with
21
that and that's something that we're
22
struggling with at this point in time and
372
1
how to do that.
2 DR. SALOMON:
Jo Anna, the comment
3
that we want to hear is when you get the
4
pancreas in-house and you begin your
5
standard operating procedure for islet
6
isolation how much variation is there from
7
preparation to preparation in the individual
8
steps in your hands in the center?
9 MS. REEMS:
Oh, we adhere strictly
10
to our SOPs and if there's any deviation
11
from the SOP an incident report is generated
12
for that so no.
13 DR. MULLIGAN:
I guess the detail
14
in the SOP is the definition of what an SOP
15
is. It's nice to say we follow
exactly the
16
SOP but the SOP says you can change
17
everything basically. So, I
mean, what I've
18
heard from the Minnesota group, the
19
implication was no, they add a certain
20 amount, they go for
a certain amount of
21
time, they shake for a certain number of
22
minutes, and that's it and so it seems like
373
1
there's a difference between what John said
2
and what the Minnesota people said.
3 DR. RAO: I'm
still waiting for a
4
comment from you but we'll get to any other
5
people on the --
6 MR.
GARFINKLE: Mark Garfinkle
7
from the University of Chicago.
As you
8
heard earlier, we're a very young program
9
yet to do our first islet transplantation
10
clinically, just had our IND approved and
11
have far less experience than most of the
12
other isolation centers in the room.
But
13
just in response to the direct question that
14
was asked, so we also do not control every
15
single procurement and so the shipping
16
conditions and the actual nature of the
17
procurement, whether the duodenum, whether
18
the spleen is attached, whether there's a
19
severance in the tail, all that is quite
20
variable.
21 In terms of the SOP I'd just echo
22
what else was said. We have a
wide range of
374
1
allowances and ranges for, for example, the
2
amount of time to shake before dilution. We
3
allow for some judgement and how quickly to
4
bring down the temperature during the
5
dilution phase. So the SOPs are
followed
6
but there's quite a range in those SOPS and
7
that's always a judgement call.
It's an
8
art.
9 DR. RAO: Dr.
Ricordi.
10 DR. RICORDI:
Yes, maybe I'm the
11
biggest of the three big ugliest body
12
weights and I have an experience of a little
13
over a 1,500 human islet processing but we
14
certainly follow the SOP to the letter, as
15
Bernhard said, but it is true that there is
16
a range in an SOP but this range is coming
17
down more and more defined.
18 Like the first IND that was
19
submitted, it was approved. To
abide this
20
restriction we said collagenase solution is
21
injected, islets are dispersed in a chamber
22
and then purified with density gradient,
375
1
period. That was the degree of
detail.
2 And it worked very well because
3
the patients were equally incident-free.
4
But then we were told that we could not do
5 that any longer and
that now you have to set
6
a very fixed range and specify the step.
7
But there is still a range and there is
8
still a component of experience but our SOP
9
is incredible detailed nowadays and as soon
10
as we have a pancreas for research that you
11
know for sure doesn't go to a patient then I
12
change everything I can possibly do to try
13
to catch up with all the opportunity of
14
improvement that we missed the week before.
15 But if something is done for
16
clinical we just stay exactly with the same
17
reagents, same steps, same procedures, same
18
temperature and yes, you have some steps
19
like the decision when to switch is still a
20
operator-based. It means you
have to wait
21
till you have enough free islets and that
22
you don't over-digest the content.
376
1 But it's all spelled out in the
2
SOP when do you decide when to switch and,
3
of course, if you have an experience of 50
4
pancreases better than when you take your
5
first decision.
6 DR. SALOMON:
I think there's
7
something else that has to be said here, and
8
that is at the moment these guys are still
9
using a biological preparation of
10
collagenase. This is not
recombinant
11
collagenase so there are even lot to lot
12
variations in the enzyme they're using here
13
which means in the case of the conversation
14
we're having right now that one of the
15
flexibilities that have to be built into
16
their SOPs is some adjustment based on the
17
lot of collagenase they're using at that
18
time because it can affect temperature and
19
the amount of time and the amount of
20
shaking, et cetera, to get the islets out of
21
the fibrous matrix of the pancreas.
22 DR. RAO:
Let's look at this a
377
1
little bit further and say often sometimes
2
in an SOP when you're building flexibility
3
like this you have readouts at the end which
4
tell you that you are within certain
5
parameters so if you change collagenase you
6
say is the size of the islets appropriate in
7
terms of readout when you do this, and that
8
becomes a standardized readout because you
9
can't control for time. So I
presume that
10
for most people when you're building ranges
11
in your SOPs there are readouts for those
12
ranges or some appropriate control?
When
13
you say that there is inside operator
14
variability is that built in?
15 DR. RICORDI:
Well, at the
16
beginning, like, if you say you can switch
17
to a dilution between 14 and 21 minutes, of
18
course, if you don't see any islet at 12
19
minutes you wait. It
depends. That will
20
vary with the characteristic of the
21 pancreas.
22 Like, you have incredible
378
1
difference. Like a pancreas from
a 15-year
2
old has an organ composition and structure
3
that is completely different from a pancreas
4
of a 65-year old. So that's the
first major
5
variation.
6 Then, as Dan mentioned, a lot of
7
collagenase materials may have some
8
difference even though we are trying to use
9
a ratio of activities of recombinant similar
10
lots of activities even from different
11
enzyme. And ÄÄÄÄ same, like, if
at the end
12 of the day you have 400,000 islets, 95
13
percent variable, and then survive in
14
culture and are unable to reverse diabetes
15
in your mouse and give a stimulation index
16
are very good prognostic factors.
17 What they repeat is very dangerous
18
to do, especially for an experience in new
19
patients. I would almost ban
fresh islet
20
transplantation where you put the product
21
immediately after cell processing because
22
that is when you cannot screen.
Then is
379
1
when the experience factor counts much more
2
because at that point you don't have the
3
luxury to be able to wait two days so did
4
you do something dramatically wrong.
So I
5
will strongly encourage on your center not
6
to try a fresh islet transplant so that at
7
least they have this built-in safety of
8
time.
9 DR. RAO: Dr.
Mulligan, did that
10
help give you some insight into --
11 DR. MULLIGAN:
I mean, it sounds
12
like the FDA would ask them to be doing a
13
state of the art isolation and there's this
14
pancreas, and so there's going to be some
15
natural variation. It sounds
like they have
16
the sense that they ought to have a strict
17
process and they have a clear effort to get
18
to a point where you add only this amount of
19
enzyme, only shake it this amount of time,
20
and so forth. So it sounds like
the spirit
21
of things is being adhered to.
22 DR. RAO: And
I guess it's also
380
1
important in this case that if it's three
2
different centers that the SOPs are
3
virtually identical in the three different
4
centers. Otherwise it would be
hard to
5
assess them as one.
6 MS. LAWTON:
It's another one of
7
those circular arguments because the
8
argument is do you have sufficient assays
9
that are really characterizing your product
10
well enough to be able to understand how
11 those different
changes and tweaks in
12
process ultimately impact your end product,
13
and I don't know whether we really have the
14
answer to that. Generally it
looks like
15
yes, we've got some good parameters that
16
seemed to be linked with outcome but is that
17
sufficient for telling you how you tweak the
18
process?
19 And then I have a question
20
specifically for the FDA and that relates to
21
if you're looking at the clinical data to
22
support approval of a BLA how much of these
381
1
tweaks and how do you understand what impact
2
that has had in the outcomes of those
3
patients and how do you look at that?
I
4
mean, that's really the question I have.
5 DR. RAO:
That's a separate
6 question and I
think we'll wait to answer
7
that --
8 MS. LAWTON:
But it's part of the
9
BLA discussion. It's a very
important
10
question around how do you understand that
11 in the chemical
data and what the product is
12
that you're approving.
13 DR. RAO:
Just hold it for a
14
minute while we get Dr. Harlan to make a
15
statement. Then we'd ask the FDA
to
16
respond.
17 DR. HARLAN:
Dr. Garfinkle pointed
18
out and I think as a relatively new islet
19
isolator I can say that people that come
20
into this knew our surprise by the art of
21
it. There's lots of science to
it but there
22
is as much art. And I've seen
this now from
382
1
several relatively new centers.
If you look
2
at islet yield per number of islet
3
isolations that they've done you see a trend
4
that's up but it looks like the Dow Jones
5
industrial average where there'll be periods
6
of time when it's up and then it drops
7
suddenly for three months and no one knows
8
why and then it comes back up again, and
9
that happens even at centers that have been
10
doing this for 10 years.
11 So it points out that yes, we know
12
a lot but then factors come in and impact
13
yield and islet quality that stump everybody
14
for a while. James might
comment. I think
15
this even happened with you guys for a while
16
at one point where you had an inexplicable
17
decline in islet yield on function that
18
eventually got attributed to collagenase, as
19
I recall.
20 DR. SHAPIRO:
I think it's always
21
explicable by a bad loss of collagenase
22
enzyme and those islets don't end up in
383
1
patients. The yield is too low
or the
2
viability is too low and we discard the
3
prep.
4 DR. RAO:
Maybe now we can go have
5
the FDA try and answer this .
6 DR. NOGUCHI:
And actually I think
7
this discussion is very useful and try and
8
the answer by several before, well, what do
9
we need for BLA. Well, now we're
getting to
10
the details where I think we will help a
11
little bit in articulating what we are
12
talking about and this goes to Allison's
13
question and Abbey's question as well.
Can
14
academic centers do this? And
while I have
15
said that we believe that academic centers
16
can actually rise to the challenge we are
17
now talking actually about something a
18
little bit different.
19 If the worldwide experience for
20 islet
transplants is on the order of a
21
couple of hundred let's say that any given
22
center might be talking about eventually
384
1
perhaps a hundred a year or a
reasonable
2
amount. That's a completely
different issue
3
and what that means is if you have processes
4
that are operator-dependent that is not
5
adequate for a BLA because what we're
6
talking about is being able to deliver a
7
product to needy patients in which the
8
practicing physician can have a reasonable
9
expectation that it is what it is, it does
10 what it does, and
that it's made in a way
11
that can be replicated if in fact you need
12
to do it again.
13 And for that if you have something
14
where you say will we look at this and if we
15
don't have any islets coming out we add a
16
little bit more or we let it go until we see
17
islets our question is going to be, quite
18
frankly, how do you know what an islet is.
19 How do you assess the quality of it? As you
20
begin to develop these assays such as oxygen
21
consumption we might say make sure you
22
monitor the oxygen consumption of the islet
385
1
coming off.
2 You have done a tremendous amount
3
of work from the year 2000 to what you have
4
today. But what we are talking
about now is
5 a completely
different thing than research.
6
It's a completely different thing than
7
development. This is all about
delivery.
8 When you deliver things this is
9
actually something that we know is a
10
different way of thinking because it means
11
that however you get this piece of candy
12
what you really don't want to make sure is
13
first that you get the candy and that it's
14
not a counterfeit or that it doesn't have
15
something that you could have prevented that
16
led to a medical error and a tragedy like a
17
death, and all those things really mean that
18
to get to where you understand
how to make a
19
preparation requires an artist, requires
20
somebody like Camillo, who has spent his
21
lifetime learning how to do this.
22 But when you implement it now
386
1
you're talking not even artisan.
You're
2
talking about McDonald's. You're
talking
3
about every time you make it you can go
4
anywhere in the world. You may
hate the
5
hamburger but you know what to expect.
6 And while it seems like that's a
7
little bit of a silly thing I can tell you
8
McDonald's never has never had a problem
9
with contamination. Jack in the
Box about
10
every five years does. And
there's a big
11
difference because people die from Jack in
12
the Box hamburgers; they don't die from
13
McDonald's.
14 You want the same assurance for
15
this and to do that now I could say that
16
this question, number two, while it's been
17
one that we've all averted our eyes this
18
comes to the heart of what is needed to be
19
done. Now you need to
deliver. That is a
20
different mind set than discovery.
It's a
21
different mind set than academic
22
collaboration. It's a much
different mind
387
1
set than even the fine details we're getting
2
into. This is really all about
the delivery
3
and making sure it is what it is, it does
4
what it does. So the answer is
well, we've
5
got a fair amount of work to do in this
6
area, I think.
7 DR. RAO: It
does seem that having
8 detailed SOPs with
clear-cut limits is a
9
good start in terms of being able to do
10
that.
11 DR. NOGUCHI:
And where you can
12
specify how to make a decision.
We
13
understand. There is a lot of
art in this
14
as a lot of art in medicine but to the
15
extent that you can set a decision tree and
16
algorithm so that it's operator-independent
17
you do not want to have only Camillo to be
18
able to make a preparation.
19 Snowstorms happen, disasters
20
happen, but patients still need the
21
necessary outcome. How can you
make it so
22
that anyone, almost, can operate and produce
388
1
the product that you need if they've been
2
trained under the right conditions?
That's
3
the SOPs and how you make a decision where
4
there's flexibility. That's very
important.
5
It's also very difficult.
6 DR. RAO:
Maybe we can try and
7
capture that and see if that's a reasonable
8 summary for
this. Go ahead, Doctor.
9 DR. RICORDI:
Thank you, Phil. I
10
think this is a very important point but I
11
was actually hoping that it would not go in
12
this direction because I think that we are
13
trying to set up what is a range of rules
14
for therapeutics, cellular therapeutics that
15
are something between a surgery and a
16
standardizable cellular product like the one
17
that you are using to develop in which it
18
will be very difficult to standardize the
19
SOP in the fashion that you have indicated
20
because the volume will be never such like a
21
mass production of a vaccine or
something
22
that will expose the millions of patients
389
1
with diabetes. We are not
talking about the
2
therapy that will be offered to the 16
3
million patients in United States with
4
diabetes that have to be done in Mobile,
5
Alabama, and Los Angeles with the same
6
quality, same reproducibilities, something
7
that is a challenge to treat a few hundreds
8
of patients with very specific risk factors
9
in specialized centers and with this product
10
that is safe and that can deliver a
11
functional outcome that is predictable.
12 I'm not sure that the requirement
13
of having something for something that start
14
with an organ surgery and end up with a
15
cellular infusion to have something that can
16
be reproduced, as in McDonald's, I think, we
17
can close right now all the centers and
18
change field.
19 DR. NOGUCHI:
Well, let's put it
20
this way. We consider this a
challenge for
21
all of us. Your point is well
taken,
22
Camillo, and we want this to move forward
390
1
but we also are challenging everyone,
2
especially the field, to do the best you can
3
and then do even better, and it may well be
4
that you cannot define precisely how you
5
decide whether you've added the collagenase
6
long enough or whether you've stirred long
7
enough.
8 But certainly there must be
9
general principles you could do.
You see 10
10
islets in 40 seconds. We're just
saying try
11
a little bit harder. We're not
closing the
12
door. We're not saying you must
have
13
exactly a locked-in-stone SOP before you
14
come forward but you need to strive toward
15
that.
16 And even after
you receive a
17
license, as Dr. Lawton well knows, you're
18
always improving. You're always
improving
19
your SOPs. You're always trying
to do it
20
better. You're trying to do it
with less
21
possibility of human error.
22 Where we're trying to transition
391
1
is from discovery to preventing human error
2
because it's human error that by and large
3
leads to failures of products.
Sometimes
4
it's failure of the delivery.
Sometimes
5
it's a failure inherent to the product.
But
6 too often it's simply
human failure. That's
7
where when you hear is this operator-
8
dependent. That's where we go
well, maybe
9
we can do a little bit better on that.
10 DR. RAO: Okay,
go ahead.
11 MS. LAWTON:
And just to add to
12
that, I mean, without talking about GMPs and
13
everything else, but it really comes down to
14
how do you write down in enough detail in
15 your SOPs the knowledge
that you clearly
16
have. How do you get that
knowledge in
17
there so that, as Phil says, any other
18
operator can do that by taking that
19
knowledge and being able to come up with the
20
same process or the same product at the end
21
of the process?
22 DR. RAO:
I'll allow you to
392
1 respond.
2 DR. SHERWIN:
Well, just to take
3
it the opposite way, it just seems to me
4
that as long as you have a product outcome
5
that is well standardized if there is some
6 variation because of
the methods you're
7
using to generate a product just by
8
necessity that's to show you there's always
9
going to be some variation because we
10
haven't standardized collagenase, for
11
example, as was pointed out, and I know that
12
every time that happens you have to
13
restandardize the lot and figure out how to
14
do it, so there's no way it can be a hundred
15
percent.
16 But if you have a very strict
17
criterion for potency isn't that the bottom
18
line? You're telling me that if
you can
19
establish the islets are functioning
20
biologically that that's not enough for
21
anybody.
22 In other words if they have
393
1
developed a procedure that is consistent and
2
within a relatively narrow range and they
3
can prove to you that they have potency for
4
the product that they're developing isn't
5
that what the goal is or am I wrong?
6 DR. RAO: I
think you're
7
absolutely right.
8 DR. NOGUCHI:
You are correct but,
9
as we've also discussed, there is still some
10
controversy over adequacy of potency
11
testing. But all these things
are simply
12
related.
13 DR. SHERWIN:
That has to be
14
proven to you in an adequate way that
15
they've done that. I understand
that part.
16
Obviously it's not as clear to me that we've
17
totally established that. I
accept that,
18
too. But if they could I'm
trying to get
19
the ground rules.
20 DR. NOGUCHI:
Yes, I mean, the
21
goal is to have a set of release criteria
22
that can be met that gives you a product.
394
1
When we're talking specifically about GMPs
2
and these sorts of items here these are
3
actual legal requirements about record
4
keeping.
5 DR. SHERWIN:
Sure.
6 DR. NOGUCHI:
But they have
7
nuances above and beyond just simple record
8
keeping. But you're right. Ultimately it's
9
does the thing work in the person that you
10
want it to work?
11 DR. RAO: Go
ahead.
12 DR. LEVITSKY:
It has been my
13
observation, looking at even pharmaceuticals
14
coming off the line, that companies have
15
problems sometimes with production and they
16
don't easily transfer to another plant
17 sometimes so I'm
not exactly sure. I
18
understand that you would want to see a
19
protocol which was very detailed and which
20
someone if they followed it and became
21
experienced would eventually be able to use
22
successfully. But you don't
intend that
395
1
this protocol could just be read and used by
2
someone with no experience initially.
3 I mean, there is a learning
4
process. We've heard from our
colleague at
5
NIH that there is a definite learning
6
process and if the protocol is sufficiently
7
detailed even though it does
indicate that
8
there will be some small changes based on
9
biologic variability that would meet your
10
criterion, wouldn't it?
11 DR. NOGUCHI:
Yes, specifically
12
for what we're talking about here it's in
13
the context of, everything else aside, if
14
you're the holder of a BLA this is what we
15
would be expecting because you will be
16
having new operators come in. We
expect to
17
have a training program. We
expect that
18
you'll have criteria for establishing how
19
qualified that person is. But
how you
20
actually evaluate that is then based on the
21
quality of the SOPs because when you're
22
doing quality control and you're saying
396
1
well, this product doesn't work, you have to
2
then be able to go back to the person who
3
did it. Did you do this
step? Yes. That's
4
why all these SOPs are signed in ink and
5
everything else. You changed
something.
6 It's a means
of quality control.
7
So even to be able to establish that yes, we
8
have a training program, we will ask and how
9
do you establish the quality of your
10
training program. How do you
know that a
11
new person who is processing your pancreas
12
has been adequately trained?
Well, we have
13
an SOP situation. We make sure
that they
14
can follow it. We have the
outcome data
15 from what they've
done. Here are all the
16
records.
17 DR. LEVITSKY:
So the folks making
18
the islets perhaps need to get consultation
19
from Genzyme about those forms that you need
20
to fill out but it sounds like they have the
21
capacity to do it. They just
have to know
22
what hoops they need to jump through, not
397
1
that they can't do it.
2 DR. NOGUCHI:
Oh, yes, that's
3
absolutely correct and it's just that when
4
you say what hoops to jump through sometimes
5
even just describing it is not enough.
You
6
have to try it a few times. You
need to
7
have experience.
8 We believe that you do not have to
9
be a pharmaceutical industry in order to
10
deliver these cellular products; however, we
11
do believe also that the pharmaceutical
12
industry has tremendous experience in how to
13
do things the same way every time no matter
14
who is doing that and that's what we're
15
referring to here. So we're
trying to help
16
build quality into these kinds of academic
17
centers doing this type of experimental
18
product that will lead to a reproducible,
19
licensable product with the
expectation that
20
this is not the only thing they'll be doing.
21 They'll be growing the cells or
22
expanding them perhaps under appropriate
398
1
conditions, then adding the gene therapy,
2
then going to stem cells. So if
you do all
3
this up front and you learn how to deliver a
4
relatively crude cellular product it paves
5
the way for actually being able to deliver
6
other products as well.
7 Quite frankly, the academic
8
centers right at the moment are where the
9
bulk of the activity and cellular and gene
10
therapies are. There is industry
interest
11
in this but it is relatively small at this
12
time. Part of this is we're
trying to
13
couple some of the innovation also with the
14
very rigorous and the very tedious kinds of
15
end-stage product development such as SOPs
16
that we're talking about here.
We believe
17
that for academic centers, at least small
18 scale, it's well
within their capability but
19
it is a new learning experience as well.
20 DR. RAO: Dr.
Harlan.
21 DR. HARLAN:
I have two
22
Einstein-associated comments and I'm not
399
1
comparing intelligence. I want
to use the
2
term "relative." These
are all relative.
3
So what I most respect about what has
4
happened in the islet field is that what
5
we're talking about here is the relative
6
comparability of the surgical risk because
7
surgeons need to be trained in a pancreas
8
transplant just like an islet isolator needs
9
to be trained in islet isolation, but I
10
would much rather have that learning curve
11
be on a cadaveric pancreas than on a patient
12
on the operating table, and that's what
13
Camillo's been striving for for 30 years and
14
I absolutely applaud it.
15 The other relative point I wish to
16
make is that in the two ways we've talked
17
about trying to perfect this technique one
18
is perfect the isolation process; the other
19
extreme, and they're related, is perfect the
20
product release criteria. I can
tell you my
21
perspective is we're going to get to the
22
product release criteria a lot sooner than
400
1
we're going to get to perfecting the islet
2
isolation technique because there are way
3
too many variables in the latter to fix up.
4
And I think we're probably pretty close
5
based upon the data that Bernard showed to
6
having reasonable product release criteria.
7 MS.
LAWTON: If I can just make a
8
comment, and I hate to do this because I
9
struggle with this and I think it's very
10
important that we are able to continue to
11
move this forward for the sake of the
12
patients and the technology; however, as I
13
said earlier, we also come from having
14
regulated cell therapy products and knowing
15
what we have to do, and if I've heard it
16
once I've heard it a hundred times.
The
17
product specifications don't make the
18
product; the process is the product.
19 And I think we've all heard it
20
before but it is important because there are
21
so many different factors that you could
22
think that won't make an impact but do and
401
1
won't be detected in your assays that you
2
have in your finished product and that could
3
have an impact on your patients.
4 So when I look at this question
5
when I think about it, your question about
6
algorithms and criteria, I think you should
7
be able to somehow come up with what you
8
need with that range of depending on the
9
source organ and what you have there.
What
10
are some of the algorithms of the criteria
11
that you're going to set to decide how long
12
do you, I don't know, culture them or
13
whatever based on your knowledge?
14 And if you can come up with that
15 so that you have
that defined for your
16
process I think that's really what we're
17
talking about. So it's really
coming down,
18
I think, at the end of the day to the
19
details and how you document it and how you
20
do it and then the data to support it.
21 DR. RAO: Dr.
Ricordi?
22 DR. RICORDI:
I completely agree.
402
1
There is a much more detail. I
don't know
2
if you're specifically referring to any lack
3
of detail in the existing SOPs because to me
4
are fairly detailed but I think that can or
5 should be improved at
the level of the
6
algorithm that you can define your ranges
7
depending. Right now we don't
change the
8
percent of the collagenase depending on the
9
pancreas but maybe down the line the
10
relative amount of collagenase versus
11
protease will changed, something that we
12
don't do right now. We use
always the same
13
amount.
14 But there are all these other
15
flexible errors. But the thing
that I hope
16
you appreciate is that there's not the
17
slight difference in the source tissue.
18
Like if McDonald can deliver always the same
19
hamburger it's because they don't start with
20
an osso bucco one time and with sirloin
21
steak the other. They always
have the same
22
disgusting mess that is uniform.
403
1 I didn't mean to insult McDonald's
2
but if you are right to use a limited source
3
of cells, like if you would have like a cell
4
line, stem cells, or pig islets, like any
5
animal source of cells, in animal cells we
6
can do always the same procedure and give
7
you a protocol defined to the extent that we
8
don't even have to look at a microscope at
9
the sample to know when to switch.
And that
10
is like surgeon-proof as a kind of procedure
11
they can examine if they've never been in a
12
lab, train in two weeks, and can deliver a
13
standard product.
14 But human pancreas processing is
15
extremely variable so we can go through this
16
and do SOPs that are incredibly detailed but
17
still I would hope that the experience will
18
be considered part of the equation and then
19
how you train the personnel and how you
20
document competence level and what they can
21
deliver will be important and hopefully
22
we'll soon be able to move to pig islets.
404
1 DR. SALOMON:
My comment at this
2
point is I think that we're obsessing about
3
something. I mean, we ought to
get back to
4
a principle. I'm not suggesting
that you
5
don't get it, but the principle here, to
6
articulate it, is you can't change your
7
protocol every five minutes based on how you
8
feel and we can't have a situation in which
9
on Tuesday Camillo is doing it one way and
10
Bernhard's doing it another way and James is
11
doing it a third way and then call it a BLA
12 and talk about
product even if all the
13
product comes out. That's the
point
14
Allison's making. The process is
really
15
critical even though at the beginning it's
16
all potent.
17 No, I'm
not on your case, Camillo.
18
I'm just saying that we're going around in
19
circles here. The bottom line is
that you
20
have to write a standard operating procedure
21
that's the state of the art for your
22
programs when you step up to do your BLA and
405
1
that's what everybody's going to do.
And
2
that doesn't mean it's the best that could
3
possibly happen. It doesn't mean
that lots
4
of great innovation can't be done in
5
parallel. But you've got to
stick with your
6
SOPs, and as long as you accept that as a
7
principle I think we can move on.
8 DR. RAO: I
think all that Allison
9
specifically emphasized was that if you
10
change your protocol it has to be documented
11
and you have to have evidence as to why that
12
change doesn't affect your end product
13
because the process is as important as the
14
product of that. But I think
that's a
15
principle and I'm sure you're as aware of it
16
as everybody else.
17 DR. RICORDI:
Yes, and we believe
18
actually our SOPs are so close to each other
19
that we will wrote a single one for the BLA
20
application that you will receive tomorrow.
21 DR. RAO: One
last comment before
22
we --
406
1 MR. MILLIS:
Yes, one last
2
comment. A lot of this
discussion reminds
3
me of similar discussions we've had over the
4
last couple of years when we've brought in
5
our consultants who had a lot of experience
6
with pharmaceuticals who said exactly what
7
Dr. Lawton said. The product is
the
8
process. And when we first gave
our
9
consultants what our process was they
10
laughed because it wasn't as defined as the
11
pharmaceutical, et cetera, and we've moved
12
very much towards that in these two years
13
that we've been working.
14 And to, I think, emphasize what
15
Dr. Noguchi is talking about, I think
16
academic centers can do this. It
takes a
17
huge learning curve to get over from
18
investigation, from research to a process,
19
and I think it just takes a lot of learning
20
from each of us and willing to accept the
21
fact that we're not changing every five
22
minutes and we are going to stick to a
407
1
process that has clear, defined end points
2
for each of those steps and that's what
3
we're going to live with. So I
think
4
academic institutions can do it but it's
5
something that we're totally unfamiliar with
6 and we have to
be taught.
7 DR. RAO:
That seemed like a
8
reasonable take-home message.
9 So that ends the discussion part
10
for this session. We will, of
course,
11 continue tomorrow. At this stage what we're
12
going to try and do is look at another
13
aspect of what is an important component of
14
the FDA's activities, to run an internal
15
research program, and this will be a report
16
on the internal research program, which is
17
part of the charge of the BRMAC committee.
18 Everybody is welcome to stay.
19
It's an open session. However,
we'll take a
20
couple of minutes while we rearrange the
21
table because there are certain members of
22
the committee who will be involved and
408
1
others who will not.
2 (Recess)
3 DR. PURI:
I'll actually give you
4
a little bit of an introduction of the
5
members who will give you an update of the
6
research program in the Division of Cellular
7
and Gene Therapies which I have privilege to
8
act as the acting director strongly
9
committed to bring the FDA industry, patient
10
advocates, scientists, and public together
11
in a new partnership to promote and develop
12
new therapies for the 21st century while
13
faithfully protecting human subject and
14
ensuring product safety that we do on a
15
daily basis.
16 We strive to meet the five goals
17
of the FDA strategy plan, which are
18
efficient, science-based, risk assessment,
19
patient and consumer safety, better informed
20
consumers, and counterterrorism and a strong
21
FDA.
22 As you are hearing this morning,
409
1
the type of products we deal with that
2
involve complex set of products, cellular
3
therapies-based products, gene therapies,
4
xenotransplantation, unique assisted
5
production, ooplasm transfer, and
6 combination products
containing living
7
cells, and some of this technology is
8
rapidly evolving. And to be able
to assess
9
this technology it is challenging to have
10
experts in every different field, so the
11
model we chose and we adopt is a cadre of
12
scientists who are expert in their fields
13
who not only pursue their own event signs to
14
maintain their scientific expertise but also
15
are product expert in multiple different
16
areas and you will hear some of their
17
updates of their research program today.
18 And our mission is to plan,
19
develop, implement a comprehensive risk-
20
based regulatory framework for cell and gene
21
therapies and our, of course, goal is to
22
ensure the safety, identity, purity, and
410
1
potency of novel products.
2 This division has two branches, a
3
regulatory branch and a gene therapy branch
4
and the cell therapy branch, and the four
5
labs, and you're going to hear today from
6
research programs two labs, a lab of
7
immunology and virology in the lower left
8
corner, and on the lower right corner is
9
laboratory of stem cell biology program.
10 And the Laboratory of Immunology
11
and Virology is headed by Dr. Eda Bloom, who
12
is on travel. Acting lab chief
is Carolyn
13
Wilson. She will present some of
the brief
14 overview of the
lab as well as present her
15
own program and also give you update of
16
Dr. Bloom's program. In addition
we have
17
two additional scientists, Andrew Byrnes and
18
Nancy Markovitz. They are
experts in
19
different virology fields and they'll
20
discuss their programs on abnormal viral
21
vectors and herpes virus vectors and their
22
basic underpinning of their science program
411
1
to review the product.
2 In addition there are a total of
3
14 members that include post-docs, lab
4
techs, and other research associates.
Not
5
only do these folks, as I indicated to you,
6
do a research program. They also
do a huge
7
amount of regulatory workload they have.
8
They have 200 INDs and more than 130 of them
9
are active. These INDs include
complex
10
biological products, as some of you are
11
hearing today, and involve cell therapy,
12
gene therapy, xenotransplantation, tumor
13
vaccines, and this lab also has an oversight
14
of only licensed somatic cell therapy
15
product and the licensing supplement.
16 The second lab which you will be
17
hearing from is the Laboratory of Stem Cell
18
Biology, which is headed by acting lab chief
19
Dr. Steve Bauer, and in addition to that
20
there are two other senior members in that
21
group, Dr. Marti and Dr. Aksamit, and
22
Dr. Marti's program is being reviewed here
412
1
today. There are eight members
in this
2
group that also includes post-docs, techs,
3
and other research associates, and research
4
in their lab is related to regulatory
5
questions involving cellular therapy, gene
6
therapy, and standards development and this
7
lab also does a significant amount of
8
regulatory work that involves INDs, IDEs,
9
510(k)s, and these are related to cell
10
therapy, gene therapy, xenotransplantation,
11
cell separation, and ÄÄÄÄ devices and tumor
12
vaccines.
13 So at this point I'll just end my
14
presentation after this introduction and
15
invite Dr. Steve Bauer to come and give his
16
lab summary.
17 DR. BAUER:
I'm here to just
18
review very briefly for you Dr. Marti's
19
research program within the Lab of Stem Cell
20
Biology, and this is an illustration from
21
his site visit report.
22 His career
has centered for many
413
1
years now on understanding the molecular
2
lesions that lead to chronic lymphocytic
3
leukemia, and one of the primary tools for
4
looking at CLL has been flow cytometry.
5
This is a diagram that shows staining for CD
6
5 and CD 20 by which you can identify
7
different B-cell expansions that happen
8
during the process of that disease.
9 And the central idea is that there
10
is a molecular lesion that leads to clonal
11
expansion and a precursor state called B-
12
cell monoclonal lymphocytosis.
Subsequent
13
mutations lead to a frank malignancy known
14
as CLL and through his study of CLL he's
15
been able to characterize three different
16
types of CLL shown here. They vary
in both
17
their CD 20 and CD 5 stating pattern.
18 But one of the ideas that came out
19
of his lab was to look for early disease, an
20
arising precursor state. That
led to
21
discovery of this B-cell
monoclonal
22
lymphocytosis, which is shown right here.
414
1 And so some of the highlights from
2
his research program are that he actually
3
was the discoverer of this BCML, and he
4
showed that these are monoclonal precursor
5
states in terms of their immunoglobulin gene
6
rearrangement and their restriction to
7
certain light chain expressions.
8 He also discovered that if you
9
looked at first-degree relatives of families
10
who had CLL with at least one affected
11
individual there was a huge increase of this
12
BCML in those families. And this
also led
13
to an idea that if you could see these
14
precursor states in families with CLL could
15
you find that in the general population.
16 And there was an interesting study
17
funded by CDC to look at people who live
18
near Superfund toxic waste sites and they
19
were able to observe an increase in BCML
20
relative to the general population
21
suggesting that people exposed to these
22
toxic waste environments, higher levels, had
415
1
increased susceptibility to BCML
2
development.
3 He also in collaboration with Lou
4
Stout and using their Lymphochip microarray
5
analysis was able to differentiate different
6
types of CLL into those that had unmutated,
7
although rearranged, immunoglobulin genes.
8
These were found to surprisingly express
9
ZAP-70 and they have a worse prognosis.
So
10
this is one of his contributions.
All of
11
these contributions in his research lab
12
started from a characterization using flow
13
cytometry.
14 So this research has several
15
important implications for how we regulate
16
products and how in terms of product quality
17
it's very important to do in-process product
18
monitoring and lot release and flow
19
cytometry can play a huge part in that.
So
20 his expertise in
flow cytometry has been
21
very useful with regard to that.
22 And then devices that separate
416
1
stem cells and other cell products often
2
rely on cell flow cytometry. His
long-term
3
interest in flow cytometry as a tool has
4
also led to his participation in development
5
of some reference materials, specifically of
6
Fluorescein Standard Reference Material that
7
has been recently released by the National
8
Institutes of Standards in Technology and
9
this should be very useful in helping
10 standardize flow
cytometry assays that have
11
use in the clinical setting both for sorting
12
and for diagnosis and analysis of cells in
13
the clinical setting.
14 Finally, his research has been
15
focused on looking at the mechanisms that
16
lead to CLL and it's hoped that the
17
molecular lesions in leukemogenesis
18
mechanisms that can be discovered in this
19
model might help with development of
20
therapeutic targets and also is relevant to
21
early detection of developing disease or
22
residual disease. This might be
important
417
1
for timing for different interventions.
2 And then finally in terms of
3
future plans Dr. Marti's long-term interests
4
in this disease and the approach of
5 following the
development of early
6
monoclonal lymphoproliferative states and
7
then later fully malignant CLL clones, of
8
course, he'll continue doing this research
9
and looking at neoplastic transformation.
10
This earlier slide that I showed you was
11
two-color analysis; he's looking at
12
multicolor flow cytometry and hoping that
13
expanded use of single-cell PCR will help to
14 elucidate some of
the molecular mechanisms
15
of this neoplastic transformation.
16 He's also continuing his efforts
17
in standards development and the
18
availability of this FITC standard should
19
help both in lab-based answers in CCLS
20
fluorescence-intensive documents.
This is
21
something he contributing on, working on.
22 And then also they're trying to do
418
1
microbead intensity standards that can also
2
be used so you couple this standard
3
reference material to a microbead and then
4
use those in flow cytometry to again
5
standardize the instruments and so on.
6 And then finally he's looking at a
7
continuing collaboration with NCI in
8
immunophenotyping CLLs and more finally
9
defining the different kinds of CLLs and so
10
on by using multicolor flow cytometry.
11 So that's just a very brief
12
overview of the site visit report and some
13
of the highlights that his contributions
14
have recently made both to the research
15
arena and the regulatory arena.
16 Questions?
17 DR. RAO: Any
questions? Go
18
ahead, Abbey.
19 MS. MEYERS:
I'm just wondering if
20
the people who were positive for these cells
21
who lived near the Superfund sites were told
22
what the results of these tests were?
419
1 DR. BAUER:
There was a follow-up
2
study and there was a large and ongoing
3
debate about the ethical implications of
4
coming back to people who had been involved
5
in that study. I don't remember
the
6
mechanism by which that was accomplished but
7
something like 80 percent of the people who
8
were identified in this first screen did
9
agree to have further follow-up so they were
10
informed. And then some of them
had BCML as
11
defined by Dr. Marti and others had other
12
abnormalities and that's in the site visit
13
report. I don't have the details
with me.
14 DR. RAO:
Thank you, Dr. Bauer. I
15
apologize but we're going to have to close
16
this session because of scheduling conflicts
17
and take a vote, and Dr. Carol Miller will
18
present her site visit. So I'm
going to ask
19
people to step out for a short duration of
20
time, and you're welcome back after that
21
when we open.
22
MS. DAPOLITO: And we hope do you
420
1
come back and hear some more about exciting
2
research at CBER. Thank you.
3 (Recess)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
426
1 DR. RAO: So
we are going to
2
continue and our next speaker is going to be
3
Dr. Carolyn Wilson of the Laboratory of
4
Immunology and Virology.
5 DR. WILSON:
Thank you. Actually
6
I'm going to be presenting a brief summary
7
of my research program as well as the
8
research program of Dr. Eda Bloom, who
9
apologizes for not being able
to be here
10
herself today but she had to participate in
11
a meeting this week at the World Health
12
Organization on xenotransplantation.
So the
13
first part of my talk is going to be
14
summarizing her program and hopefully I will
15
do her program justice.
16 She's been working for a number of
17
years on studying the immune mechanisms in
18
xenotransplantation in cancer with an
19
emphasis on studying cytotoxic effector
20
cells. To give an overview,
she's been
21
studying NK cells, the effect of cytokines
22
and the redox pathway on NK cell function as
427
1
well as the effect of aging and
2
transplantation on the activity of cytotoxic
3
T cells.
4 But more recently, and what I'll
5
be talking to you about today, are her
6
studies on the cytotoxic activity in models
7
of xenotransplantation with the aim towards
8
looking at whether she can modulate the
9
level of effector cell or target cell in
10
terms of the immune response in
11
xenotransplantation. Her
research is
12
directly relevant to FDA's mission as we
13
regulate xenotransplantation products and,
14
as you probably know, a major barrier to
15
successful clinical xenotransplantation is
16
the immune response and the potential for
17
rejection, which presents a major safety
18 concern for the
recipient. In addition,
19
infection also presents a concern through
20
potential transmission of infectious agents
21
and that actually is more the focus of my
22
research program, which will be discussed in
428
1
the next few minutes.
2 The interaction of the immune
3
system with xenogeneic cells is her main
4
focus and she is studying this with the
5
long-term goals to look at whether she can
6
develop means to evade the immune rejection
7
response or develop ways to reduce the need
8
for immunosuppressive therapies which, as
9
you know, have a number of their own
10
clinical sequelae. The program
is also
11
generally relevant to the regulation of cell
12
therapies.
13 So to
summarize what goes on in
14
the immune response to xenogeneic cells, the
15
first response is very quick, within minutes
16
to hours, and it's mediated by preformed
17
antibodies that humans have to a sugar
18
epitope called alpha-galactosidase.
This in
19
combination with ÄÄÄÄ causes endothelial
20
cell activation and hyperacute rejection.
21 If you can overcome hyperacute
22 rejection the next
stage is called delayed
429
1
xenograft rejection or acute vascular
2
rejection. This process takes
days to occur
3
and can also be mediated by
xeno-reactive
4
antibodies and ÄÄÄÄ potentially.
NK cells
5
are important at this stage as well as other
6
immune effector cells. Again,
endothelial
7
cell activation is the hallmark of rejection
8
at this stage.
9 A number of strategies are being
10
developed to overcome these immune blocks;
11
however, even taking these types of
12
strategies, which can lead to a process
13
called accommodation, you can still have
14
cellular rejection or chronic rejection,
15
which will take weeks to years, and is also
16
mediated by a number of immune cell
17 effectors leading to graft
destruction.
18 Areas of research in this area
19
have been looking at whether one can induce
20
tolerance or use of immunosuppressive
21
therapies and this is the focus of
22
Dr. Bloom's research, looking at whether
430
1
this aspect of the immune response can be
2
overcome.
3 One area that
she's looked at is
4
whether expression of cell surface proteins
5
may modulate these responses and in
6
particular shown here is the response when
7
cells from porcine aortic endothelial cells
8
are over-expressing porcine Fas ligand and,
9
as you can see, the DNA fragmentation of
10
human T cells is increased when exposed to
11
porcine aortic endothelial cells expressing
12 Fas ligand, and in
counterpart the cytolysis
13
of these cells is actually decreased.
And
14
this experiment was done with human T cells
15
but similar results have been observed with
16
human NK cells.
17 Another avenue that Dr. Bloom has
18
been investigating is looking at whether
19
antibodies to cell adhesion molecules can
20
modulate these activities. Here
antibodies
21
to E-Selectin, as you can see here, in a
22
dose-dependent fashion decreased the
431
1
cytolysis of porcine aortic endothelial cell
2
targets in contrast to VCAM1 where
3
increasing doses had no effect, suggesting
4
that porcine E-Selectin is recognized as a
5
target structure by human NK cells.
6 So then to summarize, what
7 Dr. Bloom's
laboratory has shown is that
8
over-expression of porcine Fas ligand on
9
porcine aortic endothelial cells augments or
10
confers the ability to induce apoptosis in
11
activated human T cells as well as NK cells,
12
reduces the susceptibility of target cells
13
to lysis by human immune effector cells. In
14
addition she's shown that porcine E-Selectin
15
is recognized by human NK cells and they can
16
be down-modulated by using chemical,
17
especially redox, manipulation.
18 And it's hoped these types of
19
studies that this demonstrate that
20
modulation or alteration of certain surface
21
markers may provide a means to protect
22
porcine cells from delayed or cellular
432
1
xenograft projection. So I'm
going to now,
2
not completely because it still relates to
3
xenotransplantation, but switch gears and
4
briefly discuss my program and then I'll
5
pause for questions after that.
6 My research is focused on the
7
study of viral and cellular determinants of
8
infectivity in a broad sense.
More
9
specifically in the context of
10
xenotransplantation we've been studying
11
porcine endogenous retrovirus and also more
12
recently looking at determinants of entry
13
both for PERV and a filovirus called Ebola
14
virus.
15 So our laboratory has become
16
interested in porcine endogenous
17
retroviruses primarily because pigs are by
18
most consensus in the scientific and medical
19
community the species of choice for human
20
xenotransplantation.
21 The presence of endogenous
22
retroviruses in pigs is notable because
433
1
unlike exogenous viruses, which can be
2 eliminated through
various husbandry and
3
screening practices, endogenous retroviruses
4
are integrated heritable omens in the
5
genome, meaning that these can never be
6
removed from pigs.
7 In addition it's important to note
8
that ex vivo culture conditions as well as
9
transplantation itself may in fact activate
10
expression of these types of genetic
11
elements and, more importantly, it's known
12
that the endogenous retroviral loci in pigs
13
are capable of producing infectious
14
retrovirus, and my laboratory was the first
15
to show that we could isolate virus from
16
human peripheral blood mononuclear cells
17
that was directly infectious for human cell
18
lines in vitro.
19 Our research program is relevant
20
to CEBR's mission. Our expertise
in the
21
data that's been generated in this program
22
has directly informed development of
434
1
science-based policy in not only the area of
2
xenotransplantation but also in the area of
3
gene therapy, in particular retroviral
4
vectors. And technical and
scientific
5
advice can be provided to the sponsors
6
because of the type of research we do.
We
7
have hands-on knowledge of assays used to
8
detect PERV as well as assays used to detect
9
replicating retroviruses and retroviral
10
vectors used for gene therapy.
11 So in the interest of time what
12
I'm going to do is just very briefly bullet
13
point, summarize, the research
14
accomplishments from my program in the last
15
five years. That was the time of
my last
16
site visit. And I apologize for
running
17
through this so quickly.
18 So we've analyzed PERV infection
19
in primary and established human cells.
20
We've been looking through collaborations as
21
well as in our own labs at the feasibility
22
of different animal models.
We've looked at
435
1
the contamination of PERV in a licensed
2
product, porcine factor 8, as well as in pig
3
plasma. We have looked at the
effect of the
4
types of genetic modifications on pigs that
5
are used to modulate the immune response to
6
see how that also affects the immune
7
response to PERV.
8 We've looked, again through
9
collaboration, to see whether pigs can be
10
identified that don't express human-tropic
11
PERV. We've worked on assay
development for
12
developing TaqMan quantitative specific
13
assays for detection of PERV genomes as well
14
as a functional analysis of the elements
15
that influence expression from long terminal
16
repeat. And more recently we've
been
17
looking at identifying a co-factor that's
18
important for PERV entry.
19 Finally, I just wanted to mention
20
a new part of the research program which is
21
to look at Ebola virus GP, which is their
22
envelope glycoprotein, to try to identify
436
1
critical amino acid residues and we're just
2
using our expertise with retroviral vectors
3
to generate pseudo-virions that carry the
4
Ebola virus GP in the envelope and using a
5
version that's deleted from use and doesn't
6
cause the dramatic cytopathicity of the wild
7
type virus as a base for introducing various
8
mutations.
9 So to summarize where we're going
10
in the future, we want to continue to focus
11
now on virus receptor interactions both for
12
PERV and for Ebola and we're also expanding
13
into a new area of looking at determinants
14
of PERV assembly. We like to
think that all
15
of the work we do helps to inform regulatory
16
decisions as well as work towards the long-
17
term goal of developing antivirals or
18
vaccine strategies for overcoming viral
19
infections. And I'd be happy to
take
20
questions on either Eda's or my work at this
21
point. Thank you.
22 DR. RAO:
Questions? Go ahead.
437
1 MS. LAWTON:
It's actually a
2
comment rather than a question but since I'm
3
here I'd like to just make it count and I
4
just wanted you to know that I actually used
5
the PERV example of research done at CEBR as
6
a really good example of how CEBR research
7
working with industry really helped to move
8
the whole field forward when we came across
9
this as an issue a couple of years ago.
So
10
since I was here I just thought I'd comment
11
on it again. Thanks.
12 DR. WILSON:
Thank you.
13 DR.
ALLAN: That's very good.
14
What do you think the block is for
15
replication on human primate cells?
16 DR. WILSON:
Well, our studies
17
have indicated that there are actually two
18
blocks. One is at entry and the
co-factor,
19
which I didn't have time to present, seems
20
to alleviate that block. We've
identified a
21
human cDNA that when we express it in
22
nonhuman primate cells allows the virus to
438
1
now enter the cells. We also
think there's
2
an assembly block, but we don't have the
3
detailed molecular analysis at this point.
4 DR. ALLAN:
When you say it's a
5
block to entry what's your assay for
6
defining it?
7 DR. WILSON:
Using the pseudotype,
8
for example, that carries lacZ and MLV
9
genome that would lacZ so we're just asking
10
virus to get in and get expressed.
We've
11
also done her prep analysis looking at
12
unintegrated DNA and shown that we see about
13
50 to 100-fold decrease in that.
14
Unintegrated DNA intermediate after
15
infection compared to permissive human cells
16
suggests it's a very early stage.
17 DR. RAO: Is
there any number to
18
how many integration sites for hosts there
19
are?
20 DR. WILSON:
You would think that
21
would be a precise method by this time but
22
you see estimates ranging from 50 to 200
439
1
depending on how people do the procedure.
2
Not all of those loci likely encode
3
replication-competent viruses but clearly at
4
least some of them do. Clive
Patience has
5
done a lot of work on NIH ÄÄÄÄ suggesting
6
that there are at least two active loci, one
7
that encodes a human-tropic and one that
8 encodes a
nonhuman-tropic, and not both loci
9
are necessarily active in all pigs that he
10
screens. But the pig-tropic one
is active
11
in all of them.
12 Thank you very much.
13 DR.
BYRNES: I'm going to tell you
14
very briefly about our work that we're doing
15
on adenovirus vectors. These are
16
nonreplicating viral vectors that are the
17
same as the adenovirus vectors used widely
18
in clinical trials right now, and we're
19
interested in some very basic questions
20
about how safe these are, particularly in
21
patients who have liver damage and, as
22 you'll see, we see some
very interesting and
440
1
disturbing pathology in the lungs when we
2
use these vectors in animals that have liver
3 damage.
4 We're interested in some very
5
simple questions that have complicated
6
answers. We'd like to know where
7
adenoviruses go after an intervascular
8
injection in both normal animals and
9
diseased animals, and the reticulo-
10
endothelial system proves to be a very
11
important determinant of the biodistribution
12
of the virus.
13 The RES is a
garbage collection
14
system that collects particles from the
15
bloodstream so if you inject bacteria or
16
other small particles, including viruses,
17
into the bloodstream they tend to be very
18
quickly removed by the RES. The
RES
19
primarily consists of macrophages known as
20
Kupffer cells in the liver. So
if you look
21
at the organ level you see most of this
22
stuff ending up in the liver, and that turns
441
1
out to be the case for adenovirus vectors.
2
So when you hear that adenoviruses are
3
hepatotropic that's one of the primary
4
reasons. They get taken up by
the RES.
5 This is very wasteful for gene
6
therapy because when macrophages tend to
7
take up the vector they tend to degrade it
8
and not get productively transduced.
It's
9
also dangerous because macrophages produce a
10
lot of inflammatory cytokines and chemokines
11
when the adenovirus gets into them, and it's
12
fairly common in both animals
and in
13
patients when giving large does of
14
adenovirus intervascularly to see some
15
degree of liver damage indicated by elevated
16
transaminasis and coagulation abnormality
17
sometimes. So we'd like to know
are there
18
diseases where giving large doses of
19
adenovirus systemically might be a
20
particular safety risk.
21 So what we've done, we've studied
22
two different models of cirrhosis in rats.
442
1
One is induced by complete ligation of the
2
bile duct and the other is induced by a
3
chronic exposure to carbon tetrachloride and
4
in both of these model systems animals
5
developed severe cirrhosis and when we look
6
at the biodistribution of the vector we find
7
that it goes to the lungs rather than the
8
liver or it's much more so to the lungs than
9
it is to the liver.
10 The reason for that is the
11
reticuloendothelial system changes during
12
liver disease. In the normal
situation,
13
most particular tracers, and this can be
14
done clinically with radioactive colloids
15
such as technetium sulfur colloid, about 80
16
percent, will go to the liver.
Most of the
17
remainder, about 15 percent, will go to the
18
spleen, and then the rest will be widely
19
distributed through other organs.
20 In the case of liver disease such
21
as cirrhosis the amount that the liver is
22
taking up is greatly reduced and to
443
1
compensate for that, perhaps, the amount
2 that gets taken up
by the liver is greatly
3
increased. And this is due to
the
4
development of intravascular macrophages in
5
the lung which are normally not present.
6
These are known as pulmonary intravascular
7
macrophages.
8 We've done studies with
9
adenoviruses now and we've seen uptake of
10
the adenovirus into these pulmonary
11
intravascular macrophages and that's all in
12
the site visit report and I'm not going to
13
go into detail with that but, as I told you
14
a slide ago, it's quite a concern,
15
adenovirus being taken up into macrophages,
16
that it might cause some undue inflammation.
17 So in stuff that's not in the site
18
visit report that we've been doing more
19
recently we've been looking at the pathology
20
and the consequences of the adenovirus being
21
taken up into the lungs of bile duct ligated
22
animals that are cirrhotic. What
we're
444
1
looking at here is the amount of water
2
that's in the lung. This is a
measure of
3
edema and we're looking six hours after
4
injection of the vector.
5 In normal rats we see very little
6
difference between the control animals and
7
the animals that have been injected with
8
adenovirus vector and here's what a pair of
9
normal rat lungs looks like. In
contrast to
10
that in cirrhotic rats when we inject vector
11
a significant proportion of the animals has
12
this very severe edema, which is fatal in
13
some cases and produces this very dramatic
14
picture where you see where we've cut the
15
lungs you can see all the foam coming out of
16
the lungs.
17 There's a large amount of
18
hemorrhages in the lungs and very severe
19
pathology which the animals don't survive.
20
And we think this is a consequence of the
21
adenovirus being taken up into macrophages
22
in the lungs and producing cytokines locally
445
1
and causing leakage in the blood vessels and
2
subsequent edema.
3 And what we're doing now is doing
4
a more in-depth study on this inflammation
5
in pathology in the lungs trying to figure
6
out what the root causes are.
We've
7
identified elevation of certain cytokines
8
such as Tumor Necrosis Factor Alpha and
9
Interleukin-6. It's particularly
highly
10
elevated in these animals both
locally in
11
the lungs and systemically in the serum.
12 And that's it and I'll take any
13
questions.
14 DR. RAO: Go
ahead, Jan.
15 DR.
KURTZBERG: I have a related
16
but unrelated question. In some
transplant
17
patients that are getting stem cell
18
transplants adenovirus will be nonpathogenic
19
and in others it will cause a fatal
20
pneumonia. Do you think that
patients with
21
liver disease related to other things that
22
may be affected by the transplant would be
446
1
at higher risk for adenoviral pneumonia?
2 DR. BYRNES:
I don't know. I
3
haven't heard anything about that.
4 DR. KURTZBERG:
I mean, it's the
5
virus, not really the vector.
6 DR. BYRNES:
It might be because
7
you do get systemic infection in that type
8
of situation where you have this a
9
immunosuppressed patient with a systemic
10
adenovirus infection. It might
cause
11
something like that. I don't
know.
12 DR. KURTZBERG:
But you can take a
13
pool of immunosuppressed patients and some
14
of them are going to get lung disease and
15
some of them are not, and what I'm asking is
16
could liver disease influence which gets
17
lung disease?
18 DR. BYRNES:
It could. As you
19
know, people with severe cirrhosis and other
20
liver diseases are very susceptible to lung
21
infections and respiratory distress and
22
things like that and the reasons for that
447
1
are not completely understood but it could
2
influence something like you're saying.
3 DR. ALLAN:
That patient with
4
cystic fibrosis that died from the
5
adenovirus vector --
6 DR. BYRNES:
He had an ornithine
7
trans-carbamoylase deficiency.
Is that the
8
one you're thinking of?
9 DR. ALLAN:
Oh, is that what it
10
was? I'm sorry. Never
mind. Never mind.
11
Because I was thinking well, if it was lung
12
disease then you would expect that it would
13
probably target the liver in terms of a
14
disease mechanism. Do you know
why the
15
patient died?
16 DR. BYRNES:
Nobody knows. We
17
think there's some similarities here in the
18
mechanism but it's hard to say for sure at
19
this distance in time. I think
this is
20
something that could happen in humans
21
because it's known that humans do have this
22
type of change in their reticuloendothelial
448
1
system during liver disease and from our
2
animal experiments we know that can
3
influence where the adenovirus goes.
So it
4
has the potential ÄÄÄÄ humans with liver
5
disease.
6 DR. RAO:
Correct if me if it's
7
just my ignorance. Shouldn't the
kidney
8
have lit up, too, in the camera with the --
9 DR. BYRNES:
We've looked at the
10 kidney and very
little adenovirus goes
11
there.
12 DR. HARLAN:
I'm going to expose
13
my ignorance, which I do readily.
Is it the
14
infectious particles or the particles per
15 se, which I
understand vary widely depending
16
on the prep, that are associated with this
17
pathology?
18 DR. BYRNES:
When we dose it's
19
just like when people do a clinical trial.
20
It's always on the basis of particle
21
concentration because the methods for
22
measuring the infectious particles, as you
449
1
know, are so widely variable
that you can't
2
really compare lab to lab. It's
known that
3
the inflammation induced by adenoviruses has
4
little to do with the expression of
5
adenoviral genes or the expression of the
6
transgene. This type of
inflammation that
7
we're looking at very early within hours is
8
due almost completely to the adenoviral
9
capsid and something about that is very
10 pro-inflammatory
and that's not really
11
understood that well yet.
12 DR. HARLAN:
Thank you.
13 DR. RAO: Dr.
Nancy Markovitz.
14 DR. MARKOVITZ:
I'd like to tell
15
you a little bit about my research and some
16
background information.
17 Over the past few decades
18
virologists have constructed a collection of
19
recombinant herpes simplex viruses, each
20
deleted in a gene encoding one of the over
21
85 virally encoded proteins of the herpes
22
simplex virus. This systematic
and
450
1
labor-intensive process has resulted in the
2
identification of several gene deletions
3
that affect neurovirulence of the virus.
4
These viruses can be used to identify the
5
cellular pathways through which the virus
6
causes disease and thus aid in the
7
development of vaccines and enteroviral
8
therapy for diseases caused by herpes
9
simplex virus.
10 These viruses are also being used
11
by several biotech companies as HSV viral
12
therapy or the use of viruses for the
13
experimental treatment of cancer.
And since
14
HSV is a pathogenic neuroinvasive virus
15
capable of causing encephalitis, death, or
16
long-term neurological sequelae it's an
17
important concern that these neuroattenuated
18
viruses are safe, those used in clinical
19
trials, and that we begin to understand the
20
mechanism through which HSV causes disease.
21 For a summary of some of these
22
companies and their products I direct you to
451
1
the summary in your package as well as an
2
article in the popular press magazine, the
3
October issue of Scientific American, that
4
discusses briefly some of the different
5
companies, Biovex, Crusade Laboratories, and
6
Metagene.
7 My research program focuses on two
8
cellular pathways through which viral gene
9
deletions are thought to affect
10
neurovirulence. The first, the
interferon
11
PKR eIF2-alpha pathway that shuts down
12
protein synthesis in virally infected cells
13
and which is blocked by the protein
14
phosphatase I interaction with
gamma 34.5
15
through dephosphorylation of eIF2-alpha.
16
And, of course, this is a very well worked
17
out pathway and it is the focus of many
18
different laboratories throughout the
19
country.
20 However, a less understood pathway
21
is actually unknown, really, at the
22
biochemical level pathway. It's
believed to
452
1
include both the herpes simplex viral
2
protein kinase UL13 and one of the UL13
3
viral substrates, the ICP22.
4 My lab is using a variety of in
5
vivo and in vitro methods to identify the
6
details of the host-virus interaction in
7
these two pathways believed to influence
8
disease progression. Compared to
the wild
9
type virus lab strains viruses deleted in
10
34.5 are attenuated at least 5 logs
11
following intracranial inoculation into
12
mice; however, while these biochemical
13
pathways that I mentioned previously have
14 identified the pathway through which such
15
mutation shuts down protein synthesis in
16
certain cell types in vitro they do not tell
17
us which normal cell types in vivo support
18
productive replication of 34.5-deleted
19
viruses and this is clearly a very important
20
clinical question.
21 In vivo studies, including my own,
22
have reported on the ability of 34.5-deleted
453
1
viruses to replicate in normal cells in the
2
mouse brain and suggest that these viruses
3
can replicate productively in normal neurons
4
in vivo. It is unclear whether
the virus is
5
able to replicate productively in other
6
normal cell types in vivo.
7 The research my laboratory has
8
conducted over the last few years at CBER
9
suggests that ependymal cells may also be a
10
reservoir in which 34.5-deleted viruses can
11
replicate productively. One
avenue that we
12
wish to pursue is the report by some labs
13
that ependymal cells are mitotically active
14
and are a source of neural progenitor cells,
15
and this is consistent with the observation
16
that 34.5-deleted viruses replicate better
17
in dividing cells and the reason that they
18
are preferred for treatment of tumors.
19 Future efforts in my lab will
20
investigate the ability of other normal cell
21
types in vivo to support productive viral
22
replication and to examine the fate of these
454
1
virally infected normal cells.
By
2
identifying a way to prevent death or
3
abnormal function of these
normal cell types
4
following viral infection we can begin to
5
find ways to make normal cells less
6
susceptible to infection by these new
7
therapeutic viruses.
8 Regarding
the ICP22-UL13 pathway
9
affecting neurovirulence we're interested in
10
identifying other cellular and viral
11
proteins that participate in this pathway.
12
Like ICB22, UL3 is also a substrate for UL13
13
and the multiple isoforms of the UL3 protein
14
are shown here. These top three
are
15
mediated by the viral kinase UL13.
In
16
addition UL3 co-localizes with ICP22,
17 another
substrate of UL13, and this together
18
suggests that perhaps UL3 may play a role in
19
this neurovirulence pathway.
20 In the course of making a series
21
of functional UL3 deletion viruses to test
22
this hypothesis by which we used single
455
1
amino acid substitutions we have made
2
several new findings. We have
identified
3
the transcript encoding UL3, which is
4
different than previously reported, as well
5
as the translation initiation site, also
6
different than previously reported, and we
7
found that UL3 is smaller, as
reported
8
earlier, and have begun to identify the
9
putative UL3 TATA box. We've
just finished
10
up creating a new UL3 deletion, which is a
11
functional deletion, and we will next
12
determine the effect of this deletion on the
13
neurovirulence of the virus in vivo in mouse
14
models and we will identify any synergistic
15
or antagonistic interactions between these
16
proteins, UL3, ICP22, UL13.
17 In parallel we are planning other
18
experiments to determine the cellular
19
partners of UL3 and UL13 and the residues of
20
UL3 that are phosphorylated by the protein
21
kinase UL13 as well as the cellular kinase
22
we suspect also phosphorylates UL3.
456
1 These studies will lead to a
2
general understanding of herpes simplex
3
virus neurovirulence and biochemical
4
pathways through which herpes simplex virus
5
affects disease. For the FDA
this will
6
provide the scientific expertise and new
7
methods to evaluate the safety of herpes-
8
based products that are already in clinical
9
trials and proactively provide information
10
on the safety of other types of
11
neuroattenuated herpes viruses that may be
12
in the pipeline.
13 And finally the elucidation of
14
these pathways will help us evaluate targets
15
for use in vitro as surrogate markers of
16
neurovirulence and thereby we hope reduce
17
the costly preclinical testing required of
18
new products.
19 Thank you for your attention. I'm
20
open to questions.
21 DR. RAO: I
have a suggestion just
22
based on stem cell work that we do.
There's
457
1
now a simple marker for ependymal cells,
2 CD24, and that
seems to work quite well in
3
distinguishing from other cell types.
It
4
might be useful in using --
5 DR. MARKOVITZ:
I don't know if
6
you've looked at the image in your package
7
but actually morphology is a really
8
excellent marker for ependymal cells.
So
9
we've shown here mice were injected
10
intracranially with a virus and the virus we
11
had was a 34.5 deleted virus
containing lacZ
12
so we can see which tissues were infected
13
and this is a very characteristic example of
14
ependymal cell and you can see that they
15
have these very characteristic cilia
16
protruding from the cellular membrane here
17
into the lateral ventricle space.
18 And if you look within the nucleus
19
you see the progeny virus, herpes simplex
20 virion, and this is
at day 3 postinfection.
21
Some of these capsids have DNA and they're
22
electron dense and others are empty.
So
458
1
this data, while it does not
demonstrate
2
that it's infectious, suggests that these
3
cells support productive infection to this
4
point. What we're next
interested in doing
5
is seeing if we can cut out small sections
6
of the ependymal cell layer and put them in
7
tissue culture and see if they will then
8
infect a monolayer, for example, of viro
9
cells.
10 DR. RAO: That's right. I meant
11
more in culture.
12 DR. MARKOVITZ:
Oh, in culture
13
where --
14 DR. RAO:
It's not so useful in
15
looking at --
16 DR. MARKOVITZ:
CD24 or 25?
17 DR. RAO:
CD24.
18 DR. MARKOVITZ:
Twenty-four.
19 Thank you.
20 DR. RAO: Dr.
Rosenberg?
21 DR. ROSENBERG:
Hi, good
22
afternoon, all. I'm going to
switch gears a
459
1
bit and update you on the consolidation of
2
the research divisions of the Office of
3
Therapeutics, formally of CBER, into CDER
4
and, as shown by this slide, which was
5
lifted from a recent issue of The Scientist,
6
the perception in the outside world and
7
actually from some elements within FDA is
8
that the research divisions of OTRR
9
following consolidation into CDER will
10
implode because of the transfer and in fact
11
this is not the case and CDER has expressed
12
strong support for our programs and we are
13
anticipating the benefits of an extended
14
period of support.
15 So I think there are two reasons
16
for this support. First, we're
bringing a
17
novel model into CDER, one which they have
18
not had before, which is the researcher
19
reviewer model, and I think I'm preaching to
20
the choir here as to what the benefits are
21 of that
model. And, secondly, basically
22
there are some very critical and timely
460
1
issues pertaining to regulation of
2
biological therapeutics for which they
3
really very much need our help and that, of
4
course, is the issue of follow-on or generic
5
biologics as well as immunogenicity of
6
biological therapeutics. And, again,
I
7
don't think I need to tell this crowd of the
8
importance of those issues.
9 So where do we fit into CDER? Our
10
group is in a new office in the Office of
11
Pharmaceutical Sciences. Helen
Winkle is
12
the director of that office. We
are in the
13
Office of Biotechnology Products and our
14
acting director is Dr. Yuen-Yuen Chiu, and
15
our two divisions are under her leadership.
16 And what exactly has been the
17
support? We have been given a
commitment
18
for a minimum of five years of research
19
support and as evidence in part for that
20
there is support for the
continued use of
21
the CBER PCE Committee for tenure and
22
promotion of a scientific staff and we
461
1
anticipate that in fact we will have
2
continuing interactions with this committee
3
for review of our scientific programs and as
4
well support of the CBER PCE Committee for
5
Promotions and Tenure.
6 So that's basically the update
7
that I wanted to give you, that we have
8
strong support for our continued research
9
efforts and we intend to be around for quite
10
some time. So thank you very
much.
11 Any questions?
Thanks.
12 DR. RAO: Go
ahead.
13 DR. CARBONE:
Hi, I'm Kathy
14
Carbone, the acting associate director for
15
research. I wanted to point out
that PCE
16
stands for Promotions Conversions Evaluation
17
Committee.
18 DR. SHACTER:
While we bring up
19
the slides I realize it's been an incredibly
20
long day so, those of you who have been in a
21
day of meetings, don't worry.
I'm going to
22
talk for about three minutes and just give
462
1
you a very brief overview just to tickle
2
your interest in nongene therapy-related
3
topic and then be finished with it so this
4
will really be very brief.
5 I'm Emily Shacter and I'm the
6 chief of the
Laboratory of Biochemistry in
7
the Division of Therapeutic Proteins, which
8
is Amy Rosenberg's division, now in CDER,
9
and basically what we look at is cancer
10
chemotherapy, apoptosis, and oxidative
11
stress and we try to look at the
12
interactions among these three very big and
13
interesting topics.
14 And just to summarize to you very
15
briefly in a schematic diagram of what we
16
work on what we're looking at here is a
17
tumor mass. So let's picture
that these
18
green cells are a tumor mass and these
19
little molecules here are chemotherapy drugs
20
and we're trying to understand how immune
21
system cells in this tumor mass might
22
interact with the drugs or the cells
463
1
themselves to influence the ability of
2
chemotherapy drugs to kill the tumor cells.
3
And in particular my lab is focusing on
4
macrophages, which we know are present in
5
many tumor tissues, and neutrophils.
6 The macrophages we figure are
7
there mostly to take up the dying tumor
8
cells and to remove them from the system,
9
like you might imagine in this little image
10
here, and the neutrophils we know have a
11
great capacity to secrete oxidants and these
12
oxidants could interact with the drugs
13
themselves, with the target cells, with the
14
whole system and influence the ability of
15
the chemo drug to kill those tumor cells.
16 What are these oxidants that we
17
look and, again, we're almost done, we're
18
just about there, but you can't really talk
19
about oxidants without at least showing a
20
diagram of a neutrophil making an oxidant
21
because this is what you'd have in a
22
biological system in your body if you have a
464
1
neutrophil. It's exposed either
to a tumor
2
cell or a bacterium, whatever.
It's
3
stimulated to undergo an oxidative burst.
4 And you get this series of
5
oxidants formed, superoxide anion, hydrogen
6
peroxide, hypochlorous acid, hydroxyl
7
radical. We know these oxidants
from our
8
home use of peroxide for treating infections
9 and the
disinfectant HOCL, which is bleach.
10
Our body makes these same oxidants in order
11
to deal with an infection and also with
12
tumor cells.
13 So we look at these oxidants that
14
are generated by these neutrophils that
15
might be in a tumor tissue to look to see
16
how they might impact tumor cell killing.
17
And we also study the potential ability of
18
antioxidants like this iron chelator
19
Desferal to inhibit the formation of certain
20
oxidants that could impact the system.
21 When we look at cell death and
22
tumor cell death generally the field looks
465
1
at two kinds of cell death. Both
are terms
2
that I'm sure you know, necrosis and
3
apoptosis, and in necrosis we picture that a
4
cell swells up and basically bursts and
5
spills out its intracellular contents into
6
the extra cellular milieu where, because the
7
body isn't used to seeing these
8
intracellular components, you could have an
9
inflammatory response that could influence
10
the entire system and possibly cause some
11
deleterious side effects.
12 In apoptosis, in contrast, you'd
13 have a cell sitting
in a tissue like this
14
that goes through a series of very
15
controlled biochemical steps and ultimately
16
results in the formation of these little
17
bodies called aptotic bodies.
They're
18
little subcellular pieces that are falling
19
of off this dying cell and these then get
20
taken up by a nearby macrophage or some
21
other phagocytic cell. Because
the
22
macrophage then gobbles this up
and
466
1
basically degrades it inside there's thought
2
to be really no subsequent sequelae from
3
this. It's thought to be a
relatively
4
silent process or, if anything,
5
anti-inflammatory because of certain
6
cytokines that are formed.
7 So what we have looked at in our
8
studies is the impact of oxidants on both
9
steps of this system, cell killing and the
10
phagocytosis, and we find that they actually
11
interfere with both steps. And I
won't show
12
you any data for that because I promised I
13
wouldn't but because of the silent type of
14
system we generally consider that we want
15
our chemotherapy drugs to kill tumor cells
16
by causing apoptosis so an interference by
17 oxidants in the system and
converting it to
18
necrosis, like we see in our studies,
19
actually could have negative sequelae to it.
20 But because we see that oxidants
21
inhibit both steps, are pretty universal in
22
what they do, we've actually come up with a
467
1
hypothesis. It's not really a
novel
2
hypothesis but we have a testable system
3
that the administration of antioxidants in
4
some cancer chemotherapy protocols might
5
improve the efficacy or safety of drug
6
treatment. What would be these
7 antioxidants? It's like
Vitamin C, Vitamin
8
E, all the multibillion dollar antioxidant
9
supplement industry pills that some of you
10
might be taking. And so we've
been testing
11
this hypothesis. We have some
data to
12
support it and we'd like to continue
13
pursuing that but it's a very interesting
14
avenue of research.
15 And the other thing that we've
16
looked at is what are the co-factors
17
required for the silent uptake of these
18
dying cells by the macrophages?
And one
19
very interesting and exciting finding that
20
we made is that a protein called Protein S,
21
which is actually an anticoagulant protein,
22
is required for this process and we have a
468
1
ÄÄÄÄ paper on this and least another whole
2
series of very interesting studies on the
3
role of Protein S and this whole apoptosis-
4
phagocytosis process in both autoimmunity
5
and in sepsis so we're pursuing that.
6 And just so that you'll know that
7
we are actually doing something that results
8
in publications I just thought I'd show some
9
of our more recent publications down on the
10
bottom.
11 Since we do
work at the FDA we
12
want to show mission relevance.
You've been
13
hearing it all the time. And
even though
14
sometimes that actually might sound as a
15
little bit of fluff I think it really is
16
true that our research contributes to our
17
regulatory mission.
18 These are, for example, some of
19
the products that we regulate in the
20
Laboratory of Biochemistry, which is the
21
group that I supervise, and, as you can see,
22
it's an incredibly wide array of products,
469
1
thrombolytics, antithrombotics, angiogenics
2
and antiangiogenics, apoptosis, imaging
3
agents, oncology agents, et cetera.
And our
4
research actually touches directly on some
5
of the molecules that we study, so we
6 actually have
hands-on experience with some
7
of the products that come in but it also
8
impacts more indirectly to give us the
9
ability to work with and understand the
10
whole manufacturing processes and all that
11
protein purification that we have to deal
12
with when we deal with these products and
13
when we're dealing with industry.
14 Last but not least I just want to
15 talk for a
moment about the people in my lab
16
who do the work. And actually
since our
17
site visit my lab has dwindled from five
18
people down to two so if I could get some
19
time away from my regulatory work to
20
reconstitute my laboratory then we'll
21
actually continue some of those more
22
exciting studies that I've talked about.
470
1 I've had a number of postdoctoral
2
fellows in the lab, including Howard
3
Anderson, who did the Protein S work and a
4
number of other great studies that we've
5
done. I've had a series of
neonatology
6
Fellows who come from across the street from
7
the Uniformed Services University to learn
8
how to do basic research in our lab so that
9
they can understand the difference between a
10
control and not when they're doing their
11
clinical studies. So we put them
on to
12
basic stuff to try to help teach them that.
13 And then I have two people who are
14
currently in my lab, Baolin Zhang-Jun and
15
husband, and they're working on signal
16
transection pathways that control cell
17
survival.
18 So that's all.
Thank you for your
19
time, and thank you for your support of our
20
research program. We really need
it and we
21
need your continued support. So
thank you.
22 DR. RAO:
Thank you Dr. Shacter.
471
1 DR. SHACTER:
And if by some
2
strange fickle finger of fate you have any
3
questions I'll be happy to take them.
And I
4
won't be upset if you don't.
5 DR. RAO: Now
we're going to go
6
into the closed portion of the meeting.
We
7
have the committee reports.
8 (Whereupon, at 5:58 p.m., the
9 PROCEEDINGS were continued.)
10 *
* * * *
11
12
13
14
15
16
17
18
19
20
21
22