FOOD AND DRUG ADMINISTRATION

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

9:00 a.m.

CHAIR URBA: Good morning. My name is Walter Urba and I'd like to welcome you to the Cell, Tissue and Gene Therapy Advisory Committee this morning to discuss cellular therapies derived from human embryonic stem cells and scientific considerations for preclinical safety testing.

Gail, would you like to read the statement?

MS. DAPOLITO: Good morning. I'm Gail Dapolito. I'm the executive secretary of the committee.

Before I start, we would like to request that you silence all cell phones and pagers, please.

The Food and Drug Administration is convening April 10 and 11, 2008, meeting of the Cellular, Tissue and Gene Therapies Advisory Committee under the authority of the Federal Advisory Committee Act of 1972. With the exception of the industry representative, all participants of the committee are special government employees or regular federal employees from other agencies and are subject to the federal conflict of interest laws and regulations.

The following information on the status of this advisory committee's compliance with federal ethics and conflict of interest laws, including but not limited to 18 U.S.C. Part 208 and Part 712 of the Food, Drug and Cosmetic Act, is being provided to participants at this meeting and to the public. FDA has determined that all members of this advisory committee are in compliance with federal ethics and conflict of interest laws.

Under 18 U.S.C. Part 208 Congress has authorized FDA to grant waivers to special government employees and regular government employees who have financial conflicts when it is determined that the Agency's need for a particular individual service outweighs his or her potential financial conflict of interest.

Under 18 U.S.C. Part 712 of the Food, Drug and Cosmetic Act, Congress has authorized FDA to grant waivers to special government employees and regular government employees with potential financial conflicts when necessary to afford the committee their essential expertise.

Related to the discussions at this meeting, members and consultants of this committee have been screened for potential financial conflicts of interest of their own, as well as those imputed to them, including those of their spouses or minor children and for the purposes of 18 U.S.C. 208, their employers. These interests may include investments, consulting, expert witness testimony, contract and grants, gratis, teaching, speaking, writing, patents and royalties, and also primary employment.

The committee will discuss scientific considerations for safety testing for cellular therapy products derived from human embryonic stem cells. This is of particular matter of general applicability. The agenda also includes several updates.

Based on the agenda and all financial interests reported by members and consultants, no conflict of interest waivers were issued under 18 U.S.C. 208(B)(3) or 712 of the Food, Drug and Cosmetic Act.

Dr. Kurt Gunter serves as the industry representative for the committee, acting on behalf of all related industry and is employed by Hospira, Incorporated. Hospira has less than a five percent equity interest in Novocell. Hospira does not hold a Novocell board seat and no Hospira employee serves in an advisory capacity to Novocell. Industry representatives are not special government employees and do not vote.

In addition, there are many regulated industry and other outside organization speakers making presentations. These speakers may have financial interests associated with their employer and with other regulated firms. The FDA asks, in the interest of fairness, that they address any current or previous financial involvement with any firm whose product they may wish to comment upon. These individuals were not screened by the FDA for conflicts of interest.

This conflict of interest statement will be available for review at the registration table.

We would like to remind members, consultants and participants that if the discussions involve any other products or firms not already on the agenda for which an FDA participant has a personal or imputed financial interest, the participants need to exclude themselves from such involvement and their exclusion will be noted for the record. FDA encourages all other participants to advise the committee of any financial relationships that you may have with any firm that could be affected by the discussions.

For topics such as these being discussed at today's meeting, there are often a variety of opinions, some of which are quite strongly held. Our goal is that today's meeting will be a fair and open forum for discussion of the relevant scientific issues and that the committee discussion proceed without interruption. Thus, as a gentle reminder, individuals will be allowed to speak into the record only if recognized by the Chair and FDA has scheduled a time for public comment during the open public hearing listed on the agenda at approximately 12:30.

Regarding press inquiries, please, address all press inquiries to Peper Long and Karen Riley, and they're standing over here, of the FDA public affairs office. And, also, the primary spokesman for the FDA is Dr. Celia Witten.

Thank you, Dr. Urba. I'll turn it over to you.

CHAIR URBA: Thank you, Gail.

I'd like to start with an introduction of the members of the committee. Dr. Gunter?

DR. GUNTER: Good morning. My name is Kurt Gunter and I'm the industry representative, non-voting.

DR. WOO: Savio Woo, I'm professor and Chairman of the Department of Gene and Cell Medicine at the Mt. Sinai School of Medicine in New York City.

DR. WEIR: Gordon Weir, Joslin Diabetes Center and Harvard Medical School.

DR. CHIEN: Ken Chien, I'm the director of the Cardiovascular Research Center at MGH; and a professor in the Department of Stem Cell Regenerative Biology at Harvard.

DR. SNYDER: Evan Snyder, professor at the Burnham Institute for Medical Research and director of the Stem Cell Program and Stem Cell Research Center.

DR. SALOMON: Dan Salomon, molecular and experimental medicine, The Scripps Research Institute; and director of The Scripps Center for Organ and Cell Transplantation.

DR. FRIEDLANDER: Martin Friedlander, professor in the Department of Cell Biology at The Scripps Research Institute; and also chief of the Retina Service in the Department of Ophthalmology at The Scripps Clinic.

DR. TOMFORD: Bill Tomford, I'm an orthopedic surgeon at Mass General Hospital; professor at Harvard Medical School.

DR. CHAPPELL: Rick Chappell, professor of the Department of Biostatistics at the University of Wisconsin Medical School.

DR. CALOS: Michele Calos, professor in the Department of Genetics at Stanford University School of Medicine.

MS. DAPOLITO: Gail Dapolito, Center for Biologics.

CHAIR URBA: Walter Urba, medical oncologist, Child's Research Institute, Portland, Oregon.

DR. GERSON: Stan Gerson, professor of medicine, Case Western Reserve University; director of the Case Comprehensive Cancer Center and the Center for Stem Cell and Regenerative Medicine.

DR. KWAK: Larry Kwak, chairman of the Department of Lymphoma and Myeloma at M.D. Anderson Cancer Center and associate director of the Center for Cancer Immunology Research there.

DR. TAYLOR: Doris Taylor, professor of medicine and of physiology; director of the Center for Cardiovascular Repair, University of Minnesota.

DR. FIRPO: I'm Meri Firpo, also from the University of Minnesota at the Stem Cell Institute.

DR. GOLDMAN: I'm --

MS. DAPOLITO: Dr. Goldman, could you turn your -- yes.

DR. GOLDMAN: I'm back on? Steve Goldman, I'm the professor of neurology/neurosurgery at the University of Rochester and the head of the Center for Translational Neuromedicine there.

DR. ALLEN: Matthew Allen, I'm associate professor of Small Animal Surgery at the College of Veterinary Medicine at the Ohio State University.

DR. CHAMBERLAIN: I'm Jeff Chamberlain, professor of neurology at the University of Washington; and director of the Muscular Dystrophy Research Center.

DR. SCHNEIDER: Bruce Schneider, medical officer of the Center for Biologics, FDA.

DR. CHEN: Theresa Chen, Center for Biologics, FDA.

DR. BAUER: I'm Steve Bauer, branch chief of the Cell and Tissue Therapy Branch, FDA.

DR. WHITTEN: Celia Whitten, office director of the Office of Cell, Tissue and Gene Therapies at FDA.

CHAIR URBA: Thank you. We'll start out with an introduction from the FDA given by Dr. Steven Bauer, Chief of the Cell and Tissue Therapy Branch of Cellular and Gene Therapies of CBER.

DR. BAUER: Well, good morning, everyone.

Members and guests of the Cell, Tissue and Gene Therapy Advisory Committee, on behalf of my colleagues at the FDA, we welcome you and thank you for your participation in this meeting.

My introduction will describe the background, goals and focus for today's meeting. This meeting is being convened to provide the Food and Drug Administration with scientific insight regarding safety issues in the development of cellular therapies derived from human embryonic stem cells. No specific products will be discussed for regulatory review purpose.

Invited speakers will present information that focuses on scientific issues concerning development of these types of products. Members of the committee will be requested to consider this information and provide a response to FDA questions.

Could I have the next slide?

The Office of Cellular, Tissue and Gene Therapies is charged with oversight of all cell-based therapies. FDA has convened a number of advisory committee meetings over the previous years to discuss scientific issues related to development of cell therapies.

This slide lists selected recent advisory committee meetings and a recent workshop related to cell therapy products. These meetings have provided significant feedback about the scientific issues involved in the development of quite a few novel cell therapy products, many of which use stem cells from a variety of sources.

Information from these meetings is posted on the CBER website where it can be accessible to researchers and the public, as well as FDA. Also of likely interest to many of you in attendance today is an upcoming NIH research symposium on clinical applications of cell therapies on May 6^th this year. We expect continued activity and discussion as these therapies advance in their development.

There is considerable interest in development of cell therapy products derived from stem cells due to their ability to self renew and proliferate in tissue culture while maintaining pluripotency. The FDA has considerable experience in the evaluation of investigational cell therapy products. Nevertheless, the use of cellular therapy products derived from embryonic stem cells prevents specific safety questions that the committee will consider today.

This slide shows several biological properties of embryonic stem cells and how those properties result in particular safety concerns. As this diagram, where my pointer is, illustrates, these cells have the ability to self renew and proliferate in tissue culture while maintaining pluripotency. The pluripotency is suggested by these downward arrows and the upward arrow suggests the self renewal. But these properties allow production of large numbers of undifferentiated stem cells.

And as illustrated here, the cells maintain their capacity to differentiate, and they can be induced to differentiate along specific cell lineages under carefully controlled laboratory conditions. As they differentiate, they proceed through immature stages and finally differentiate into mature, functional cells. Today we will be discussing cell products that are intended to be more mature than stem cells, as indicated in this rectangular area.

The immature stem cells have the ability to generate teratomas, which may contain differentiated cells originating from all three embryonic tissue types: endoderm, mesoderm, and ectoderm. Although this characteristic provides evidence of pluripotency, it also raises a potential safety concern, tumorigenicity, as shown by this arrow. When administered to animals in sufficient numbers, these cells give rise to teratomas comprised of either differentiated or undifferentiated cell types depending on the micro environment at the site of administration.

Because cells that are more mature do not pose the same risk of teratoma formation, current efforts are focused on using cells that are downstream or more mature than embryonic stem cells. Nonetheless, cell therapy products derived from human embryonic stem cells could be heterogeneous in their composition and include cells that have differentiated to variable degrees. Residual undifferentiated stem cells may become teratomas, some of which may become malignant.

A related potential safety concern is that residual undifferentiated stem cells and partially differentiated cells may have the ability to migrate from the site of administration. Since they will contain the capacity to proliferate and differentiate further, such cells may undergo inappropriate differentiation and form ectopic tissue which could potentially have tumor-like effects or disrupt function at unintended sites.

One of the FDA's major responsibilities is to assess potential risks and benefits and take steps to reduce potential risks to subjects who are enrolled in investigational clinical trials. Safety assessment is based in part on the preclinical safety studies performed by a sponsor, and the design and conduct of the studies are critical to evaluating whether it is reasonably safe to conduct the proposed clinical investigations.

This slide quotes from Title 21 of the Code of Federal Regulations, Part 312, which describes safety assessment studies. As pointed out here, adequate information about the pharmacological and toxicological studies should be provided. The kind, duration and scope of animal and other tests required varies with the duration and nature of the proposed clinical investigations.

So the preceding slides provide some of the background that leads us to the focus for today. Namely, safety issues regarding the use of cellular therapy products derived from human embryonic stem cells. The goal of the meeting is to obtain expert advice regarding product characterization, preclinical testing, and design of clinical studies both to enhance recipient safety and improve the assessment of therapeutic activity in clinical trials of such products.

We will have discussion from our advisory committee members and additional experts on the panel. These members received a briefing document assembled by my colleagues and me from the Office of Cellular, Tissue and Gene Therapies. This document is available on FDA websites. The document contains additional and more detailed background information and some specific questions for discussion.

In the next few slides I will highlight the issues we would like to focus on today.

The topic of inappropriate differentiation, including tumorigenicity, is of a particular interest to us, especially with regard to the animal models that could be used for preclinical safety assessments. We are seeking the committee's perspectives on issues related to the selection of animal species, animal models, and design of animal studies that are relevant to clinical trials.

The physiological environment and the anatomical location where cellular products are administered, as well as where the cells end up, may have a significant influence on safety. Therefore, the site of cell implantation and the fate of cells after administrations are topics for discussion today.

Another topic is related to cell dose extrapolation, in particular as it relates to the composition and purity of the cell product. As I mentioned earlier, these products may be heterogeneous in their composition and contain some undifferentiated stem cells.

Low levels of undifferentiated cells are of concern, but total cell doses for animal studies are lower than total doses proposed for human studies. Does this difference in absolute numbers of potentially tumorigenic cells alter the ability of the animal studies to predict patient safety in clinical trials?

Finally, study duration is important to discuss since engraftment and duration in animals should be sufficient to address concerns about inappropriate differentiation and tumorigenesis.

Another topic for discussion is product characterization. The sensitivity, specificity, robustness, accuracy and precision of assays used for characterization of these products must be sufficient to provide a reasonable assurance of safety. Assays used as process controls and for lot release should be capable of detecting and accurately measuring low levels of undifferentiated stem cells or other cellular impurities. These may present an unacceptable risk because of their ability to form tumors, differentiate inappropriately, or present other safety concerns.

As this slide shows, we are also asking the committee to discuss issues related to patient monitoring and other aspects of clinical trial design that will help assure safety in recipients of these products. What tools exist for monitoring the infused cells during clinical trials? Where do the cells localize and in what amounts? Can we detect inappropriate cellular differentiation and function in the recipient?

An important clinical issue relates to cell dose levels. Given the potential risks of these products, data supporting a reasonable possibility of efficacy may need to be particularly strong.

Clinical design parameters should permit evaluation of potential clinical benefits even where possible in early studies. Such expectations of and proposed evaluations of potential therapeutic action are generally based on preclinical demonstrations of proof of concept, and specific requirements for such data will vary among products in clinical indications.

I have discussed the focus of this meeting and described some of the issues on which we are seeking the committee's insight and perspectives. To add to the basis for discussion, we have asked some guest speakers to address different aspects of the issues.

This morning we will have presentations from our guest speakers. Dr. Melissa Carpenter, Dr. Jonathan Dinsmore, Dr. Jane Lebkowski, Dr. Ole Isacson, and Dr. Jeff Bulte. This will followed by an open public hearing in the afternoon and then the panel discussion of FDA questions.

We value the expertise, insight and perspectives that our panelists and guests bring to this meeting and the opportunity for open discussion. Thank you to the members of the committee, our guest speakers, and Gail Dapolito and other FDA staff who have helped organize this meeting. Thank you.

CHAIR URBA: Thank you, Dr. Bauer

(Applause.)

CHAIR URBA: Any questions from the committee?

(No audible response.)

CHAIR URBA: If not, thank you. We'll move on to our first presentation from Dr. Carpenter, Developing A Safe Human Embryonic Stem Cell Product For Diabetes.

DR. CARPENTER: Okay. I'd like to start by thanking the committee for the opportunity to speak today. And what I'd like to talk about is utilizing human embryonic stem cells for the treatment of diabetes and talk about the safety issues surrounding that.

What I want to start with is by reminding you that in the case of diabetes, cell therapies have already been pretty well tested. What we know about diabetes and about cell therapies for diabetes is, that if you implant primary islet cells into these patients, you will see a positive effect.

This has been done, perhaps most recently and most notably, in the Edmonton Protocol, and in this case what it's done is that primary islets are isolated from donor pancreases and put into the liver of diabetic patients. The result of this is that these patients have, in many cases are completely insulin independent, which is a profound effect in the diabetes population.

Now, there are hindrances to this. These grafts ultimately do fail and the patients are on complete immunosuppression. But the point of starting with a slide like this is to tell you that islet cell therapy or cell replacement therapies can have a profound effect on the patients and now it's a question of being able to do this in a manner which is going to be safe using human embryonic stem cells. So that's where we go next.

So how do we do this kind of a therapy with a human embryonic stem cell product knowing that we're going to have an effect on these patients that could be positive?

Well, there's two aspects to this, and a lot of what I'll be discussing today is characterizing the cell product in vitro. But then, also, you're going to want to evaluate the cell product in vivo and this is going to take a number of different aspects. You're going to want to be looking at the efficacy, the stability, any potential toxicity, and tumorigenicity.

So what does it look like when you're talking about characterizing a cell product? And, specifically, I'm going to be talking about the work that we're doing at Novocell for diabetes as an example for this.

Now, we've used a process in which we are differentiating the cells down a specific lineage. So we're starting with human embryonic stem cells and we're driving them through specific, discrete differentiative steps through endoderm to an insulin-producing cell population. Now, in doing that, we've got a starting material and the cell product and we need to characterize various steps here. It's not simply just characterizing the cell product in order to demonstrate that you've got a safe and effective situation, you need to start by characterizing your embryonic stem cells, your starting material, remembering that your ES cells are not your cell products, they're your starting material.

There's the intermediate populations that you're going to step through. Likely there will be some assessment of these populations. And then there is your final cellular product, and in our case that would be a population of cells that's insulin producing.

Now, what I'll do is I'll start first with the starting material. How do you characterize this and what have we done toward that? Well, the answer to that is we started by generating a human embryonic stem cell line. We did that under clinical manufacturing conditions and this allowed us to start with derivation of these cells using a GMP compliant human feeder. Most of you probably know that ES cells are derived on feeders. We used a human feeder that we generated under GMPs that passed points to consider testing.

We chose our reagents very carefully so that they were compliant with clinical manufacturing. We've done quite a bit of characterization and I'll show you a few slides about that. We put down some banks and we points to consider tested the cells. What this tells us is that our starting population is free of the pathogens that we tested for, so that gives us an initial safety profile on the cells.

Now, the next step is characterizing the cells in a number of different formats ranging from genetic identify to the phenotype of the cells looking at markers, looking at the cytogenetic stability of the cells, and differentiation of the cells. Obviously, we want the cells to differentiate appropriately. So I'm going to show you a little bit about these different steps.

Now, genetic identity is really very straightforward. It's a simple fingerprint. You can do this at an outsource lab, a multiplex PCR that gives you, in this case, different loci are tested and they give you different PCR bands and you can get a unique profile for each cell line that you have.

What this allows you to do, I mean this sounds very simple, but you need to make sure if you're handling multiple cell products in your facility that the cell product that you're delivering is the cell product that you think you're delivering. And simple mistakes can be made in labs and you need to be able to track your cell product, and one simple way to do that is just simply doing a fingerprint.

The next thing that you want to know is the phenotype of the cells. So what tools are available for this? Well, there's lots of different tools and this gives you an example of flow cytometry.

Now, for ES cells we all know that there's sort of a standard panel of markers that you can use to identify an ES cell. Some of these are tied to a pluripotent phenotype. Some of them are surface markers that we just know happen to be on the cells and we haven't identified their function yet.

Now here what I'm showing you are two flow cytometry panels. On the left is TRA-1-81 stained with Oct4, and you see SSA4 stained with Nanog. And you can see the upper right on both of these panels that the bulk of the populations are co-stained by these markers. This is what you would expect for human ES cells, and so we have characterized this celling that we've generated called the CyT49 line over multiple passages and in multiple conditions and we demonstrate that these cells have an abundance of expression. That's going to tell you that your population is largely containing human embryonic stem cells with an appropriate phenotype.

The next step is the cytogenetic analysis of the cells, and this is actually one of the more critical components. And cytogenetic analysis is usually done most typically with a G-band, and I'll talk about that a bit in the next slide. But essentially what is performed is you send yourselves to a cytogenetics facility and they'll assess generally 20 cells and you'll get a report back that says that those 20 cells had a normal or an abnormal karyotype and then you can move forward with that data.

And so what you'd like to do or what we've done is to test the cells on feeders in your normal conditions of growing them. You want to test them after you've cryopreserved them. That's a fair bit of trauma to a cell. You want to make sure that the cells that are coming out of your bank, out of a freeze are normal. And then you want to test them in the culture conditions that you're going to be using for your scale up. So here we've tested the cells that have been grown under feeder-free conditions, and in all cases we're seeing a normal karyotype.

Now that doesn't tell you that 100 percent of the cultures that we test are 100 percent normal. There are some lines that come up in some cultures that come up with what we could call the mosaic karyotype, meaning that some of the cells in the culture are abnormal. And depending on the abnormality that is seen, there are different levels of concern.

Now, remembering that cells in culture, cells that are maintained over

long-term culture, do change in culture. That's an expected situation. The cells in your body are dividing and changing, and in many cases there's aneuploidy.

So what we need to do is measure aneuploidy and measure it over time to determine if the cells are stable. So when you think about cytogenetic analysis, there's lots of different ways to assess this. One can do G-banding; one can do SKY analysis; and there's other molecular techniques that are now being used that are more sensitive.

But when you start collecting this data, one of the things you do need to consider is what to do with the data. So to date what we know is that aneuploidies do occur in these cultures. They occur at different frequencies in different labs under different culture conditions. And those aneuploidies can be different kinds of aneuploidies.

Things that are associated with embryonic carcinoma cells, one of the close cousins to ES cells, are things like trisomy-12 and trisomy-17. Those kinds of aneuploidies we worry about more than others that are not associated with a poor clinical outcome.

Now, to date, what to my knowledge, a correlation between a mosaic or an aneuploid population of ES cells in an adverse event, and an adverse event would be the cells no longer differentiate, the cells make a tumor in an animal, the cells are metastatic in an animal, these sorts of things have not yet been correlated with a mosaic or aneuploid population.

I think many groups are now collecting data to try to draw this correlation. We don't know yet what an aneuploid population will do.

Now, we all think that we want to go forward with as normal a population as possible, but, remember that as you grow cells in culture, it's unlikely that 100 percent of your cells will be absolutely euploid and absolutely unchanged from the time of derivation. So the question is to be able to chart that over time and measure the stability and make sure that that's not connected to an adverse situation.

Now, that takes me straight to stability. There's two different ways to look at stability or probably more than two. Here what I have is, are the cells stable over long-term culture? So ES cells, by definition, are these cells that proliferate for very long periods of time. Well, we want to know if, as they proliferate, do they change in some way that will be adverse.

We also want to know about stability over long-term preservation. We're all making cell banks and we intend to use these cell banks for cell therapies, so that's another type of stability that needs to be assessed.

When you think about the stability of the cells as I just described, in terms of long-term culture, what are you going to assess? Well, in our case, you've got the ES cells. You're driving them down this differentiation. And at some point you're to need to make more ES cells, whether that's before you put down your master bank or after you put down your bank and right before you actually differentiate the cells. At some point you need to do an expansion.

So how do you know if the cells have retained their fundamental characteristics? Well, there's many things that you can measure. You can measure by ability. You're going to measure whether the composition is consistent.

Is the phenotype of the population -- remember this is a cell population -- is that going to be consistent? Are the cells retaining their karyotype? Is it normal? Is it drifting? Is it stable? And are the cells able to differentiate in a consistent way and differentiate to the population that you want for your cell product?

Now, let's move to the cell product. So I told you about the starting material. Now let's talk about the cell product. Now as you drive the cells through differentiation, you're going to want to measure markers for these cells to make sure that the cells are being appropriately stepped through your process.

And here I've listed in our case a number of positive markers. But remember that it's not just the positive markers that you're going to need to measure. It's the negative markers. It's the markers for cells which you would prefer not be in your population or the non-target markers. So there's a number of things that you're going to want to be measuring as you take the cells through your process.

You are going to want to, at some point, enrich this likely. That could be a mechanical enrichment, that could be an enrichment that has to do with the optimization of your differentiation protocol, and all this is going to result in your end process testing which you're going to want to be predictive of your outcome.

So in terms of your cell product, what is it that you're going to want to assess? What's likely to be in this?

Now, Steve just told you that it's likely going to be a heterogeneous population, and I'm going to tell you the same thing, that your cell product is probably not going to be homogeneous.

So what is this assessment going to include? Well, you're going to want to assess your function cell. In our case, that's the cell population that produces insulin. That's our functional cell. But there's likely going to be accessory cells, and in our case that might be an endocrine population and endothelial population.

You're also going to want to assess for inappropriate cells, so that would be undifferentiated ES cells, or maybe cytotoxic cells or cells that are doing something adverse.

And then the last thing I've got on this list is "bystander" cells, and by that I mean that there might be some cells that don't appear to be doing anything, maybe a fibroblast population, that doesn't appear to be functioning, but is there and you need to identify it and you need to know that it's there reproducibly and quantifiably. So there's multiple different aspects of this that you're going to want to measure.

Now, how are we going to do this? Well, there's many different tools that you can choose from immunocytochemistry to flow cytometry to QPCR. And I'll give you a couple of examples of how this might work out.

Now, here I've got an example of some QPCR data in which human embryonic stem cells, undifferentiated cells, we're using OCT4 to track those, are spiked into fibroblasts to assess can we measure a low threshold of ES cells in a somatic subpopulation. And the answer is, yes.

So here what you see on the left is ES cells showing OCT4 expression. Here's 10 percent ES cells; here showing a much reduced expression. And then this aspect is blown up here and you can see that we can measure as few as 0.04 percent of the population spiked into a fibroblast population.

Now, this is an unoptimized assay. This is straight out of our research group. This is not our QC profile. It's just to give you an example of the kinds of assays that you can use and how you can test them. So this tells you about the RNA.

But what you might want to know is about the protein and you can use flow cytometry in the exact same way by doing a spiking experiment. And here what we've done is we've spiked in ES cells into human fibroblasts, and here you see in this population 97 percent of the cells were expressing OCT4. And here we've got the spiked population of 50 percent, five percent, and 0.5 percent, and you can see in the flow cytometry profile that there's just about 50 percent here, there's just a little over five percent, and here we're running anywhere from 1.5 to two percent. That's telling you that our threshold for this kind of assay is the noise level's going to be about two percent.

So that doesn't sound very sensitive, but this is, again, this is an unoptimized assay and there's many different ways to optimize this including doubling up or tripling up the markers to increase your resolution to optimizing your antibodies. But it gives you an idea of the kinds of assays which are available and how you might want to go looking for any ES cells in your final cell product.

Now, let me summarize the in vitro characterization. You're going to assess your starting material. You're going to look at the identify and stability of this cell product. That's going to require you to validate a lot of assays. These need to be sensitive and this is going to be -- the sensitivity of your assays is going to be balanced against your products, so you don't want to use all your cell product in QC. So there's a bit of a balancing act that goes on here. And then you're going to want these assays to be predictive.

Now, let's move onto safety and tumorigenicity, and I'm only going to talk about a couple of bullet points here. So, obviously, dosing is important, biodistribution is where do the cells go, and do they maintain their identity once they're there.

In a diabetes situation, the stability has a couple of different aspects to it. One of those is glycemic control. We're not just looking at whether the cells maintain their identity, but, also, do they maintain their function. And then there's obviously the tumorigenicity that you just heard about from Steve, and I'll be talking about that.

I'm going to talk about the in vivo aspects of that, but I would point out that the in vitro analysis and assessment of tumorigenicity may ultimately be one of the most sensitive ways that we can assess for the presence of ES cells or the presence of any sort of a tumor-forming population.

So let me start by asking about what is going to be a relevant animal model. So when you're talking about using human cells and you're putting human cells into an animal, you're doing a xenograft. And we know that the most permissive sites that we have available to use for this kind of a xenograft are immunocompromised animals.

So immunocompromised animals come into buckets if you will. There's rats and there's mice. And it's very permissive, you can put the cells in and you can do great modeling. You can do your tumorigenicity assays.

But what if you wanted to do a large animal assay for some of your functional analysis? Well, what would you choose? Well, you've got the large animal models require immunosuppression. So you're doing immunosuppression in a xenografted format.

So the drugs that you're giving to suppress the immune system may very well not be the drugs that are in your clinical protocol. And what that will do is it will confound your data to some extent, so you need to balance that.

So that brings up the question is, will a large animal study be meaningful in the presence of those kinds of compounds? Those compounds can often be toxic to the cells that you are delivering. So you do need to keep this into consideration. And, unfortunately, I don't think there's an easy answer to this question, but it will need to be balanced with the outcome that you have in your small animal immunocompromised animals and the patient population that you're intending to go into. So all of this will be need to be put into a balance.

Now, moving onto what we're really looking for is the presence of tumor formation. We're concerned that there's going to be a stray ES cell in our cell product. And one of the things I think is very important to clarify is in the literature the word teratoma and the word teratocarcinoma seem to be bounced around in a rather equivalent way.

And I think it's important to point out that a teratoma is generally considered to be a benign tumor, and a teratocarcinoma is generally considered to be a malignant tumor. And we can talk about the semantics, but what I think is really important to understand is that when you talk about, is there going to be any tumor formation, to qualify that in terms of whether or not it's a malignant tumor. And, in most cases, what is thought to occur for ES cells is a benign tumor.

Now, the risk of teratoma formation, again, this is going to be a balancing act just like in the last slide, it's going to be balanced with the patient population and the implant site, and there may be different risk assessments based on where the cells are placed, whether it's in the subcutaneous space or in the CNS, these are all things which need to go into the balance in terms of determining what the risk profile will be.

So in terms of tumorigenicity, what is going to be the appropriate assay? So the first question that we always ask is, how many ES cells does it take to make a teratoma? And that seems like a very straightforward question.

However, when you think about it, and I think Steve mentioned this as well, is it the absolute number of cells or is it the frequency of the cells. If it takes ten cells to make a teratoma, or a hundred cells, and those are in the context of a thousand somatic cells or differentiated cells that's your cell product, is that different than if they're in the context of a million cells? And so these are experiments that we can do to sort that out, and, of course, each cell line will need to be measured likely.

The implant site, we know from the literature, can affect teratoma formation. That some sites appear to be more permissive. There's lots of different aspects to that, but some sites appear to be more permissive and maybe that's a self survival issue.

Are other cell types tumorigenic? You're going to want to know that as well. And what about the immune status of the recipient? Now, in some cases you could envision that these products will be into immunosuppressed patients, and will that in some way affect the possible tumorigenicity of a stray ES cell.

And my last question here is, what does a negative result mean? So what we all anticipate is that we will test our cell product in animals, we will let them go for a long period of time, and we will not see any tumors. So what we need to know is, what's the sensitivity of that assay so that we can make a judgment about what that absence of tumor means. And that goes back to the first question, how many ES cells and in what format does it take to make a teratoma?

Now, one of the other things that I'd like to point out is that the different ES cells, mouse ES cells and human ES cells, are really quite different from each other. And I think it's really important to understand that when a mouse ES cell makes a tumor, it may not necessarily be predictive of when a human ES cell can make a tumor. They have very different qualities, and we could spend a long time on this, but I'll just point out a couple of things.

We know that mouse ES cells are much more tolerant of single-cell dissociation than human ES cells and that makes a large difference in their ability to be tumorigenic. We also know that their requirements for self renewal, which would be part of the tumorigenesis, are quite different. So I point out that mouse cells are not human cells and we need to keep that in mind when we start building predicted assays.

Now, let me finally summarize with telling you that you're going to need to characterize your cell product. You're going to need to demonstrate safety and efficacy. This is going to require sensitive and specific assays, and that these assays are going to need to be predictive.

And then I'll conclude by going back to the beginning and saying, in the case of diabetes, we know that cell replacement therapy can have a tremendous benefit for the patients, and now we're looking at ES cells as the source for this kind of a therapy, and that in terms of the immunosuppression issues around this. And Novocell has developed an encapsulation technology which we're not here to talk about today, but there is great promise here and we need to sort out the safety issues and be quite careful about it.

So with that I will close and take questions.

(Applause.)

CHAIR URBA: Thank you, Dr. Carpenter.

Obviously, these issues will come up for lots of discussion this afternoon. So, to try and stay on time, if the committee members have any questions, it would be good if they were points of clarification rather than beginning the long discussions that will start this afternoon.

Dr. Taylor?

DR. TAYLOR: The only question I had is when you do your marker analyses, do you do that on a non-purified population or does it always have to be a purified population of cells?

DR. CARPENTER: Well, right now we do this on what we call an enriched population of cells. Once we have established enrichment, we will do the same procedures. So we'll take the population of cells and run it through a battery of assays, but likely that -

(Technical difficulties - No recording for 25 seconds)

Do you mean in terms of the RNA expression or genotype?

DR. TAYLOR: The genotype.

DR. CARPENTER: It's a single genotype. These are cell lines that are generated from a single embryo.

Does that answer your question?

(No audible answer.)

CHAIR URBA: Other questions?

DR. SNYDER: What are the most common aneuploidies that you have seen and that you actually would allowed as not being significant?

DR. CARPENTER: We have not ruled out anything as being not significant yet. But we have seen, occasionally, a trisomy-12 in a mosaic format. We've never seen trisomy-17 yet in San Diego. And, occasionally, we see a deletion of part of chromosome 18.

DR. SNYDER: Very often we can see a loss of one of the --

(Technical difficulties - No recording for two seconds)

-- in the cell lines. Is that something you've seen?

DR. CARPENTER: We may have seen it once or twice, but it's not something that crops up. And one of the things to point out is if you see similar aneuploidies cropping up in a cell line every so often, it could be that you have a very low level of that aneuploidy in your cell bank that can crop up occasionally and you need to monitor that. But we haven't seen much in terms of the X chromosomes.

DR. GERSON: In vivo selection is something that would be of interest in the context of safety. Are there data on aneuploidy persistence in vivo?

DR. CARPENTER: I'm not aware of any. One of the questions I think that that points to is, if you had a mosaic population, let's say that 30 percent of your cells were aneuploid and you implanted them, who would survive? And is the aneuploidy going to give you a survival benefit or is it actually going to inhibit survival and be destroyed? And I don't know anybody that's has done that yet that's sharing data about it, but that's certainly an experiment that needs to be done.

CHAIR URBA: Okay. Thank you.

We'll move on to the next presentation from Dr. Dinsmore, Senior Vice President and General Manager, Advanced Cell Technology and Mytogen.

DR. DINSMORE: I'd also like to thank the organizers for the opportunity to speak today.

And my presentation is going to address the issues relevant to embryonic stem cells, but realize that when we're working with embryonic stem cells we're also practicing cell therapy, which is not an established field at this point relative to its therapeutic application. So there's a number of complicating factors here and I think it impinges on the question of, in your preclinical efficacy studies, what are you actually showing, and that one spend sufficient time in the preclinical models in order to fully identify what the therapeutic benefits are and risks and potential risks.

Just in general, not only do you have to worry about whether there are embryonic stem cells in the population, you have to worry about what's happened to those cells. So in the embryonic stem cell case this is a unique situation where your cellular product is derived entirely in vitro. Other kinds of cell therapies are derived from either fetal or adult tissues. So nature has done a certain amount of work in generating those cells here, you're starting purely from an in vitro situation.

So things happen in vitro because we don't know how to reproduce and culture all of the appropriate signals and factors that are there in a developing embryo and in the adult body. And so there are certain general characteristics of cells that one needs to be very cognizant of in terms of growing these. Rendering any of these to a commercial scale manufacturing is difficult. So usually the easier they are to grow, the better. But, in fact, the easier they are to grow, the more problems you may have.

So of keen interest in any of these is not just looking at the karyotype analysis, but looking at the behavior of the population. Are the cells growing at a constant over time? What's the stringency for growths? Do they need feeders, not need feeders? Are they growing on plastic? Do they need extracellular matrix compounds? All of these things can tell you that there may be shifts in the population and cells may not be the original embryonic stem cells but have subtle differences now that no one can pick up by looking at a karyotype.

The factors that accentuate these risks as I mentioned are things that related to scale of production when you're producing large numbers of these cells. It's much more difficult than when you're looking to try to produce smaller numbers. So scale is an issue. Either culture process, purification process, differentiation process also will affect this.

And not only can there be problems with the original embryonic stem cells, but there can be problems that arise with the derivatives of those embryonic stem cells. Again, this being a purely in vitro generated product has unique safety risks.

One of the areas that was originally thought that cells could get instructions in adult tissues on how to differentiate, this, by and large, is not the case, and I think everyone is directed towards the need to pre-differentiate yourselves in vitro prior to using them therapeutically.

And, again, here, I think driving them to a terminally differentiated state is the ideal relative to the potential safety of what you're applying. And so, again, not only eliminating the undifferentiated embryonic stem cells, but making sure you haven't got some kind of altered progenitor in the population as well.

And, also, long-term followup in the animals and, again, by and large, in a small animal when you're looking at these things, tumors pop up relatively fast, in months, so I'm not sure that it's necessary to follow an animal for its life time, but sufficient to know that within the initial starting population if you had a transformed or -- cell that could grow uncontrollably, you would identify it in, say, three or four months time. So, again, long-term followup is relative.

And so, I'm very fond of combining the proof of principal efficacy studies with the safety studies because looking at the ultimate function of these cells will tell you a lot about the adverse function of these cells. One of the keys there is you will be able to account for the cells you put in and the cells you get out or don't get out, and choosing the appropriate models, having the appropriate markers before you start the experiments so that you know that you can identify the cells within the graft recipients.

And route of cell delivery is an important feature. So systemic delivery is much different than site-specific injection. When you inject cells into an adult tissue, there's cataclysmic cell death. There are immune inflammatory reactions. There are many things which are impinging on that cell population. And so, again, there are separate rules relative to whether it's systemic or direct injection. Specifically with systemic, cells travel throughout the body, you really have to look everywhere for where those cells go.

Now there is -- relative to cell therapies, there are cases where there are therapeutic benefits but there is no apparent cell survival. So what does that mean? Do the cells migrate away from the site of administration? Or is there a nonspecific effect that they're releasing factors that are stimulating endogenous repair mechanisms?

And so, again, depending upon whether you, in the end of your studies, can show cell survival or not, steers the way in which you look at your safety studies. So if you cannot demonstrate cell survival, then one needs to answer whether it's actual cell death, immune elimination of the cells.

And if it's immune elimination, one would be wise to look at a number of different species combinations because you always want to assure that you haven't missed something and it does matter which species you combine relative to the immune response.

When you're looking at situations where there's benefit in the presence of no obvious cell survival, it's important to use control cells, not just saline control.

Also, blinding your experimental team relative to what they're putting in, this is difficult with cell therapies, because when you have a cell suspension, you know it. It's cloudy. If you have saline, you know. It does affect how you deliver what's in a syringe whether you know it's cells or not. So, appropriately blinding is critical to your overall outcomes.

And, as always, ectopic survival does need to be assessed even when there's no apparent survival at the site of administration. When you can find an identifiable graft things are simpler, you have markers, you can look at the presence of that graft, you can look at the components of that graft and it allows you to better assess the risk of overgrowth and any unwanted or undesired cells. Again, you can compare that against the overall efficacy at the end of the experiment.

Cell survival assessments, one of the things that with more time you always find out more problems. Most of the tags that are out there in one cell type or another have been shown to be toxic or give rise to an artifactual activity, their fluorescence or enzymatic activity. Some markers are transferred from cell to cell due to cell fusion events, or macrophage activity which comes in and takes up damaged and dead cells.

And then there is this issue that has come out from the cell therapy world which is cell fusion events so there is an actual fusion of the cells delivered with host cells. And so are you looking at actual survival or fusion? And these things are important, again, in analyzing your overall efficacy in your animal models.

And one of the benefits, and we'll get to it as we go along, is that if you're doing cross-species transplants, you have unique markers that you can use to identify them. Because we want to look at the function of the human cells, it means we will be putting them into either immunosuppressed animals or immunodeficient animals.

The site of administration does matter. So nude mice and rats are perfectly fine if you're transplanting into an immunoprivileged site. They're not fine if you're looking at other sites of administration because you will get eventual immune rejections. So, again, if you want to look at what the unfettered survival of the population is, you need severely immunodeficient animals to appropriately assess that, and, ideally, having eliminated BET and K-cell activities.

The cell source, again, important. We've heard a lot about the cell source, but I agree that human cells are different than those from other species, and they will need to be tested and that will require the use of an immune suppressed or immune deficient animal.

The shipping and storage conditions are important. So what you do in a laboratory needs to be modeled for how you'll deliver it to the patients. And so the shipping and storage conditions have unique effects on the cells, as well as unique risks if you're talking about cryopreserving and having your clinical team thaw the cells and delivery them.

On the cell manufacturing side, again, the process and source of cells you want to make sure are the same as you'll be using clinically. Eliminating any undifferentiated cells, the period of the population, again, we just heard a lot about that. Things that affect the period of that population is whether cells are grown as monolayer or clusters. How is the differentiation done? ES cells can hide out in clusters of cells are less likely to be sustained long term if you're culturing cells in monolayer. The selection process you're using for the differentiation is important and how that is done and whether cells are grown after the selection process.

Mode of delivery is key as well to the safety and efficacy in these models. The injection device can have effects on the cells. You need to look at that both in vitro and in vivo. Looking at the short term and long term effects.

And cells, when passed through a device, can reveal deficiencies. There's something like being caught in a crowd in New York that sweeps everything along with it. Well, that's what cells do oftentimes when they're in an injection device and they bring out particulates and contaminants that aren't seen otherwise and those cause problems in the recipients. Again, these things should be tested as best as possible in the preclinical models.

Sometimes the testing of the delivery devices will require the use of large animal models as well. They don't necessarily have to be immunosuppressed so long as you can look at the safety of the injection device.

From there, for a look in general at tumorigenesis and biodistribution studies, again, those can be designed with the efficacy studies if you're doing them in the appropriate hosts species. And if you're doing independently, again, you want to make sure that the recipients are sufficiently immunodeficient.

And then, again, that you're using the same size scale-up process to produce the cells for your safety studies as you'll be using clinically. It doesn't work to use a smaller scale-down version of what you'll use clinically because the number of cells you're generating is very important in terms of the demands on the system and what can happen during the grow up of large numbers of cells. And, again, you want to mimic your storage and shipping conditions as well.

So I just want to now to shift gears a bit and give a specific example of work that's been done at Advanced Cell Technology. This is looking at a retinal pigmented epithelial cell in here in patients with degeneration of the retina. There's loss of pigmented cells which are there to support the photoreceptors in conditions like

age-related macular degeneration. There's no real treatment, so unmet medical need.

These are cells that are hard to derive from fetal or adult sources, so, ideally, embryonic stem cells provide a unique source for these cells and one can generate these readily from embryonic stem cells, first taking them through embryoid body formation and then actually picking pigmented cells by physical appearance then culturing those cells over a period of months to grow up sufficient numbers and this is done in monolayer where cells are grown -- these are epithelial cells and have a typical epithelioid morphology; then the final maturation stage, where the cells begin to increase the amount of pigment; and at the end looking at various measures, both markers and functional measures for the characterization.

So at the start of the process the embryonic stem cells are characterized in terms of normal karyotype. After the expansion and growth of the RPE cells, again, you're looking for normal karyotype. The cells are grown in monolayer over a long period of time and in conditions that do not favor embryonic stem cell growth. So at the end you have a well characterized population which you can look at markers for both the embryonic stem cells and for the retinal pigmented epithelial cells and show an absence of the stem cell markers and a presence of the differentiated markers in the large percentage of the cells.

And the final screen says well in terms of functional phagocytosis, elastin secretion, and the PDEF secretion. These were taken into animal studies looking in an accepted animal model for retinal degeneration, the RCS rat, immunosuppressing. They're put into the eye, which is an immunoprivileged location, and one can assess the survival integration. Cells do survive long term in these animals and, as you'll see, integrate to prevent photoreceptor loss.

And, actually, these experiments in rats can be used to extrapolate the dose to a human, but we've tested in primates as well, the validity of those assumptions drawn from the rat studies.

This is just a histological examination of an animal about 120 days out after administration of the human ES-derived RPE cells. In panel B is higher magnification of the area where the photoreceptors have been rescued versus C in an area where cells were not administered and have continued degeneration. The thickness of the RPE layer is preserved, as well as the morphology. Whereas in the deteriorating zone, there is a disorder array and thinning of the RPE layer.

In moving from the animal testing to the patients, we thought a lot about the application that we were going after in terms of limiting the potential risks and maximizing the change for potential benefit to patients. So, again, going into the eye, you have an immunoprivileged site, which is an issue with embryonic stem cells because we don't have exact matches to the recipient.

Cells will have to be protected either by immunosuppression or some other means. There is a relatively small dose required. We're going into the macula. It's a confined area within the retina. So we need a relatively small calculated now a million and a half cells in order treat a human eye successfully. And it's a self-contained space that limits some of the issues relative to where the cells go. It can be easily visualized and you can deal with most any problem, be a standard ophthalmological procedures in terms of any adverse events.

And so I'll stop there. I know I presented a lot of material very quickly, but invite anyone with questions to ask away.

(Applause.)

CHAIR URBA: Thank you.

Go ahead, Dr. Allen.

DR. ALLEN: You mentioned in both your rodent and also in your large animal, your primate and your human primate model you used immunosuppression. I know that that's immunoprivileged site. But how long is that protocol effective for, is that for the lifetime of the study as you have it?

DR. DINSMORE: Yes.

DR. ALLEN: And what is what, what do you do in humans?

DR. DINSMORE: In the rats it is continuous. In the primates, there are

short-term studies. Again, we're looking at just the mode of delivery to verify that you can do it appropriately. Those are into aged primates. And based on the lack of any overt immune response in any of these, in the patients we would look to do transient

short-term immunosuppression with local steroid administration and not have long-term continuous immunosuppression.

DR. ALLEN: Are any plans in place to do, for example, a non-human primate study with transient immunosuppression and then let it go out and monitor the immune response?

DR. DINSMORE: You know, we're fairly confident, based upon the accumulated data now, that these cells and the now allogeneic situation will not be rejected.

DR. SALOMON: Two things came up that may be relevant to the conversations the committee has later. The first is, you were very careful parsing your words that your injections led to I believe something like restoration or it didn't really deal with the idea that these differentiated specifically into pigmented epithelial cells in the site, that's the first question. Which is fine if it, you know, works. I'm not arguing. It's just something that will come up, this term mechanism.

The second thing is, did you do a controlled-cell injection to control for the idea of just the injection effect?

DR. DINSMORE: Yes, we did, so I did go through that quickly. The cells are terminally differentiated as we start so we have the actual pigmented epithelial cells which are injected. We identify actual surviving differentiated human RPE cells in the rats. So I did go over that rather quickly. But we have replenished the RPE layer in the site of injection.

In a model here we have continual degeneration outside of that area. You can use that area as your control for the absence of the cells. So, in fact, these are fully differentiated in vitro and maintain that differentiated state in vivo.

I forgot the second half of your question.

DR. ALLEN: The second was, did you think that then in that model you needed to do a controlled cell injection?

DR. DINSMORE: So we did. A control cell there was to compare to ARPE-19, which is a transformed retinal pigmented epithelial cell human which is being used in clinical in an encapsulated form and these cells performed about -- log level better than the ARPE-19 cells.

DR. FRIEDLANDER: A lot of specific comments, but in the interest of what's relevant to the committee just one comment and one question.

The comment is that the eyes relatively immunoprivileged sites and absolutely immunoprivileged, and so these studies in intact eyes is very different than doing a diseased eye, but we can discuss that later.

The question I had was you said 95 percent of these cell differentiated into RPE cells. What do the other five percent of the cells do?

DR. DINSMORE: Minimally 95. In fact, the results of the characterization have 99 percent plus of cells. They would be cells of endothelial morphology, but not staining for a specific marker. So endothelial cells of some type.

DR. FRIEDLANDER: I guess my question is, are there any residual undifferentiated embryonic cells?

DR. DINSMORE: Yes. It's undetectable in terms of the stem cell markers and then 99 percent in terms of the specific markers we're interested in.

CHAIR URBA: Dr. Chappell?

DR. CHAPPELL: You mentioned the difficulties involved with funded blinded with blinded or comparators. Have you over come those or do you think they can be overcome?

DR. DINSMORE: In terms of doing our animal experiments with mimicking the cells or what specific example relative to

the --

DR. CHAPPELL: Well, I don't think the animals would know. But the ones -- you said that you would inject saline differently than you would inject --

DR. DINSMORE: I'd say that one needs to make sure that your investigators are completely blinded to what they're doing. So because cell suspensions are much different than saline, if you don't either use -- it can be complicated relative to cell therapies. What is an appropriate control cell? Because they may do something. They may make the situation worse. So it is not necessarily simple to choose your control cell, but to work towards having some way of blinding your experimenters to what they're actually injecting. Because these animal experiments take a long time. Oftentimes the injection procedures are complicated. When you get to the saline, it's like, okay, and the cells are typically long and hard to come by, so it's just natural that one take more care and to try to eliminate those biases as best as you can to completely blind it and have it unbiased.

DR. CHAPPELL: And do you think you've solved that problem yet?

DR. DINSMORE: The best is to have a control cell type to inject. There are other cases if you have a surviving graft and you have a benefit that's in relation to the graft to the control is less of an issue. And when you have a surviving graft, you can look at the consistency of the graft, the composition of the graft, the size of the graft relative to the number of cells put in, so you can do all of the accounting which makes the control so much simpler. I think some of the things become more complicated when it's hard to find the cells, or when it's a low frequency of survival, exactly what is going on, and I think it's very important to have a comparator cell if it's feasible so you can look at the relative benefit. So if you have islets from ES cells and you have islets from human islets, do you get equivalent results? Do you get much poorer results? Do you get better results? Some choices of therapeutic cell from embryonic stem cells is, as I said, the pigmented epithelial cells have been derived from fetuses. The problem is they can't be dissociated so you're dealing with sheets of RPE cells, which are much more difficult to manipulate and sustain and implant. So the embryonic stem cells source here is actually an advantage because these cells can be derived in the single-cell suspension. Will it adhere to the Brooks membrane when you test them in vitro? And you can inject and integrate in the animal lungs.

CHAIR URBA: Dr. Weir.

DR. WEIR: Yes. I just wanted to get into the karyotyping a little bit. Melissa Carpenter nicely brought up some of the complexities, and I just wondered what level of analysis you were using for your karyotyping?

DR. DINSMORE: It's sent off for routine cultured karyotyping and the problem with embryonic stem cells is lines vary relative to their stability. There's the presidential lines and then there are probably hundreds of other lines that have been generated. If you work with embryonic stem cells you know that no two embryonic stem cell lines are exactly alike. Some more readily differentiate into some cell types than others. So it's very difficult to say relative to one population or another how stable or unstable they may be.

In our hands on gross examination karyotypes are normal. The growth is constant over the period of time. The master cell bank is a relatively small master cell bank because we're generating in a batch about 100 million differentiated cells. Those cells are expanded from a differentiated precursor, so we use very few embryonic stem cells in the initial step. So that's why I said a scale of the process is important to what the potential issues may be with the originating embryonic stem cell population.

CHAIR URBA: Dr. Gerson.

DR. GERSON: Thank you for finishing up with that comment because I just wanted to further investigate a question that I have on the issue of stability and potency related specifically to culture and culture passage.

So in this particular application, just as an example, I think you'd like to achieve potency on the basis of self persistence in the space of the retina, not necessarily self proliferation. And yet you have cells coming in that have been in monolayer culture or single-cell suspension culture and you've been culture expanding them. And you've shown us a discriminator for a loss of ES capability, but not loss of proliferation capabilities.

I have two questions. One is, what's the impact of cell passage and how do you know that you've optimized conditions of your cells in a batch? And, how do you manage the transition from a cell proliferation set to a cell maintenance expectation?

DR. DINSMORE: Yes. I'm not sure I can answer it completely.

Through the process of expansion, as I mentioned, the cells of an endothelial morphology are relatively nonpigmented. That's under one set of culture conditions. And they shifted to a set of differentiation, further differentiation media which then that's that month period of following where the cells are pretty much in stasis and begin to accumulate pigment and take on their final state that we use for administration. And so we would rely on the outcomes in the animal testing to verify that we've got an inappropriately matured cell population.

CHAIR URBA: Two more questions. Dr. Chien?

DR. CHIEN: I was fascinated by your ability to get this very rare cell type out of human ES cell starting material. I think that's an amazing achievement.

It also raises a questions as, how do you know that you have the cell that you really want? What are the criteria? I mean this is a very specialized cell. If you look at the transcriptional profile of the cells that you're isolating from ES cells, how do they compare to the real McCoy?

And, obviously, you use a few markers, immunocytofluorescence, but there's a much more, perhaps, complicated functional readout that one could get and is there an intermediate step where you can identify a progenitor that's fully committed, still can replicate and then differentiate that would then be sort of like an intermediate point that you could capture and then you characterize that product, that intermediate product much more carefully than an ES cell because it might be less heterogenetic.

DR. DINSMORE: Right. That's a very good point and something that we have a specific whole program devoted to generating embryonic progenitor cells that can be propagated as in clonally and then derive the differentiated cell types of interest.

What I presented here was just a superficial overview. We look at about 15 different markers for the pigmented epithelial cells which are all present in these cells and present in the bona fide pigmented epithelial cells from the retina. The gene expression profiling is done as well and is imperative to the real McCoy. And the retinal pigmented epithelial cell in this case, it's a purely spontaneous event in these populations. And, again, with embryonic stem cells there are certain cell types that come out spontaneously from embryoid bodies. Certain neuronal cell types, certain blood cell types, cardiomyocytes, smooth muscle, and --

DR. CHIEN: But you don't have to purify it and you don't have to direct it and you can end up with 95 percent of --

DR. DINSMORE: -- so the purification comes after the embryoid body formation, the spontaneous differentiation where you have actual pigmented cells. So we're actually then driving them back into a proliferative stage to expand them, and then allowing them again to mature and pigment. So we're actually picking, visually, and then fully growing those pigmented cells.

DR. CHIEN: So it's eyeball a metric, basically?

DR. DINSMORE: Right now it's low tech.

DR. CHIEN: I'm just trying to understand the technology. The other question, though, you didn't answer. The transcriptional profile of the real McCoy versus the ES derived. How close is it?

DR. DINSMORE: I would have to defer you to the exact data. That's not my area of special expertise. But, suffice it to say, they are overlapping and I leave it to someone like yourself to say how closely overlapping. To a neophyte relative to that particular analysis, they look very close.

DR. GOLDMAN: I think that addresses my question with regards to the 95 to 99 percent enrichment of phenotype. The methodologic issue of they were selected. These are embryoid bodies that are pigmented. So I just want to clarify to make sure I understand.

So the incidence of the RPE phenotype in the embryoid body mix may still be a low event?

DR. DINSMORE: It is a low event.

DR. GOLDMAN: Okay. And so then it's just those embryoid bodies that appear to be biased towards generating RPEs who have been manually selected out. And then the progeny of those embryoid bodies are 95 percent plus?

DR. DINSMORE: Again, you're dissociating and actually picking the pigmented cells. That was as looking at the array of cells you can get from embryonic stem cells without having unique surface markers for precursors, one of the sorts of things you can pick out -- well, one thing they could see readily in these cultures were pigmented cells in the population.

DR. GOLDMAN: So there are lots of pigmented phenotypes. So are these being further selected?

DR. DINSMORE: Right. So the pigmented cells were pulled out, then grown, and then, through a series of experiments, identified as retinal pigmented epithelial cells. Again, that's a series of markers and I only gave a few here, but there's 10 to 15 unique markers that are all expressed.

DR. GOLDMAN: Are there any specific sorting steps that are involved or this is post hoc --

DR. DINSMORE: -- it's just progressive, early on progressive pulling of the unique cells. So you can render this to a FAC sort. We just haven't gotten to that point again with the relatively small numbers cells that we're dealing with at this point. That's the next step we'll move to. And I think it is one of the aspects of this that make it a somewhat easier process to move forward with in taking the path of least resistance.

CHAIR URBA: Thank you, Dr. Dinsmore.

We'll move onto the next presentation. Dr. Lebkowski, Senior Vice President of Regenerative Medicine from Geron Corporation.

DR. LEBKOWSKI: Okay. Thank you very much, again, to the organizers for allowing me to present to the committee. And a lot of my comments are going to echo of Melissa's and Jonathan's looking at very important in how to culture, characterize and the preclinical studies that are involved in establishing and developing an embryonic stem cell based therapy.

For my presentation, I am going to be looking and addressing some of the considerations that we made and took into consideration for development of some of our candidate therapies, and illustrate with some examples from one of our products, which is an oligodendroglial progenitor based therapy for the treatment of spinal cord injury.

So the embryonic stem cells essentially represent a new cell type for potential broad therapeutic application. However, before this can be realized, there are many necessary technological developments that have to take place so that we can make this for widespread use and makes these populations safe for use.

We need to simplify, standardize and scale the growth of the undifferentiated stem cells. We need to develop reproducible methods to selectively differentiate those cells into our target populations where we can know and characterize what the untargeted or the extraneous phenotypes are. We need to develop parameters to characterize the differentiated cell populations and define their efficacy, potency and safety of those cell populations.

We need to develop methods to deliver those cells to the target tissues so that they can act effectively and, preferentially, not go to untargeted sites. We need to define the need for immunosuppression and we also need to demonstrate the safety and efficacy in clinical trials. And then eventually develop scaled, low cost production methods so that these can be available widespread.

I am going to be referring in a couple of examples to experience with our product, which is referred to as GRNOPC1. GRNOPC1 is a population of cells that are differentiated from embryonic stem cells. They contain oligodendroglial progenitor cells. We know that these cells produce various neurotrophic factors. We know that these cells can induce myelination when implanted into, for instance, a Shiverer, mouse, or when implanted into an animal with a spinal cord injury.

So we think that these cells are important in many functions in spinal cord injury, namely tissue sparing, namely the survival and potential expansion or extension of axons, and also inducing myelination of axons.

One of the very most important things in looking at developing an embryonic stem cell based therapy, and I think Melissa addressed this very nicely, is identifying and establishing your differentiation process. And one of the most important features in developing that differentiation process is characterizing the materials that you're using for that differentiation process.

Most importantly is the starting material. I think anybody who has used embryonic stem cells knows that these cells can be tricky to maintain and you need to have very rigorous conditions to maintain them in an appropriate starting phenotype that's appropriate for the start of your differentiation process.

So not only do you have to characterize these cells, again as Melissa pointed out, for adventitious agents, but you need to maintain them in a state that can be an appropriate starting material for your differentiation process.

After you have established that procedures for growing the undifferentiated cells and characterizing them, you need to also characterize the differentiation process, each of the different unit operations that are required to step you through this process to take you from undifferentiated cells to differentiated cell population.

You need to characterize, for instance, cell density, the culture format, and the timing of induction with whatever differentiation agents you are using in order to make sure your optimizing and producing a relatively uniform composition of cells at the end. You also need to look at the storage conditions. How are you planning to store these for your eventual therapeutic use? Are you going to them in culture? Are you going to cryopreserve them?

Another important feature to look at, for instance, if you're going to cryopreserve them, what cells are lost during those cryopreservation procedures? What cells are retained?

And then importantly, looking at release testing. One of the, again, important features of looking at your final release testing is composition, in addition to sterility testing and mycoplasma testing and endotoxin testing, a very important feature really is looking at the what does your final population contain? And this is just an example of some of the different markers that we've looked at for this particular oligodendroglial progenitor product. And, obviously, during all of our release testing you're looking to establish what's the identity, the purity, the strength and the potency of that particular product.

There are some challenges, challenges in accomplishing this. And I think the message that I'd like to say is some of those challenges is that multiple markers are required. Again, as echoed before, these are populations of cells. You want to identify what types of cells, what is your targeted population, what are some of your extraneous phenotypes?

And in many cases there is no one marker to tell you that you have your population of interest or you're nontargeted population of interest. You are looking at multiple markers in some cases for every cell type.

Also have to be very careful about lineage specific markers. Again, for some particular applications in identifying some particular cell types, a lineage marker that has been well characterized in adult tissue might not, might have cross-reactivity in some of these more primitive populations. So it's really important to look at what are the markers that you're using.

In particular, also, the specific antibodies, we have found that in many cases we'll screen five to ten different antibodies for a particular marker before we find one that is specific enough and selective enough and sensitive enough to actually detect our marker of interest.

And in particular for your impurity assays, ones that are consequentials, it's really important to characterize both the limits of detection and the limits of quantitation for your particular marker of interest. Obviously, also, potency assays are required and are important, and, in some cases, if your cells have various pleiotrophic effects, that can be a challenge. But it is important to start early in looking at what your candidate potency assays would be.

So once you do have a population of interest and a population that you think will have some therapeutic efficacy, it's important, then, to consider what your nonclinical studies will be. And these are some questions that I think are important in establishing what those nonclinical studies are.

You're going to need to look at: what is the final product designed to do? Where do you need to inject it in order for it to have activity? Does the scale effect the composition? I can tell you from our experience that there are subtle differences in the way you culture these cells, that it's important to do these kinds of experiments with the product that you're thinking about taking to the clinic, that particular scale.

It's also important, again, to look at what your formulation is. Is it going to be a cryopreserved format? Is there selective cell survival? And how are you going to clinically administer these so that your models actually reflect both the site of administration, any particular effects on that site of administration, on the performance and the potential adverse events that you might see? Is there a need for immunosuppression? And what's the dose required?

That will lead to your pharmacology studies or your activity. What's the activity of the cells? For our product GRNOPC1, we actually looked at a couple of different activities. We looked at its in vitro activity looking at protein and gene expression for multiple different genes.

We've also looked at factor production. For instance, we know that these cells produce a variety of different neurotrophic factors and have looked at the effects of those neurotrophic factors in in vitro culture.

One important thing to consider also is whether these cells have any structural or metabolic effect, anything else besides factors that are important in the