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

(8:01 a.m.)

CHAIRMAN SALOMON: I want to welcome everyone here to the second day of Meeting Number 34 of the BRMAC, which of course doesn't mean anything to anyone, including me.

What does mean something is the topic for today, which is to update the Committee, and to provide advice to the FDA on retroviral gene therapies, particularly focused obviously on the cases of severe combined immunodeficiency disease. But certainly the discussions need to range somewhat beyond that narrow focus at some point. The nature of the topic is absolutely important and critical, and I think that does not need any introduction, so I will spare you that.

The day is relatively compressed, just because it is Friday, and with the snow and trying to get to airports, I would like to be done at 3:00. Otherwise, I am not making it back to California this evening.

So you have got a Chair who is engaged in finishing on time, which is usually a good thing. So without any other introduction at the moment. We will pick up with the FDA's introduction in a moment.

I would like to turn to Gail Dapolito, our Executive Secretary, to read a statement into the minutes.

MS. DAPOLITO: Good morning, everyone. This announcement is part of the public record for the Biological Response Modifiers Advisory Committee meeting on February 28, 2003. Pursuant to the authority granted under the Committee charter, the Director of the FDA's Center for Biologic Evaluation and Research has appointed Ms. Barbara Ballard and Ms. Abbey Meyers, and Drs. John Coffin, Kenneth Cornetta, John French, Warren Leonard, Stewart Orkin, Crystal Mackall, Thomas Murray, Bruce Torbett and Linda Wolf as temporary voting members for today's discussions.

Based on the agenda, it was determined that there are no products being approved at this meeting. The Committee participants were screened for their financial interests to determine if any conflicts of interest exist.

The Agency reviewed the submitted agenda and all financial interests reported by the meeting participants. As a result of this review the following disclosures are being made.

In accordance with 18 USC 208, Drs. John Coffin, Kenneth Cornetta, and Warren Leonard were each granted a waiver that permits them to participate in today's committee discussions.

Dr. Richard Mulligan was granted a limited waiver for today's discussions that permits him to participate in the discussion without a vote. We also note for the record that Ms. Alison Lawton serves as the non-voting industry representative member acting on behalf of a regulated industry.

She is employed by Genzyme, and thus has interests in her employer and other similar firms. With regard to FDA's invited guest speakers and guests, the agency has determined that the services of these speakers and guests are essential.

The following interests are being made public to allow meeting participants to objectively evaluate any presentation and/or comments made by the speakers and guests. Dr. Claudio Bordignon is employed at the Institute of Science in Milan, Italy. He is a researcher in gene therapy clinical trials, especially ADA-SCID, and has associations with firms involved with retroviral vectors.

Dr. Marina Cavazzana-Calvo is employed at the Necker Hospital in Paris, France. She is involved in retroviral vector gene therapy studies to treat patients with X-SCID.

Dr. Adrian Thrasher is employed at the University College in London, England. He is a researcher in gene therapy clinical trials to treat patients with SCID.

Dr. Cristof von Kalle is employed at the University of Cincinnati, and is involved in gene therapy research. Drs. Amy Patterson and Stephen Rose are employed with the Recombinant DNA Program, Office of Biotechnology Activities, NIH. NIH funds gene therapy research.

Members and consultants are aware of the need to exclude themselves from discussion involving specific products or firms for which they have not been screened for conflict of interests. Their exclusion will be noted for the public record.

With respect to all other meeting participants, we ask in the interest of fairness that you state your name, affiliation, and address any current or previous financial involvement with any firm whose product you wish to comment upon.

Waivers are available by written request under the Freedom of Information Act, and as a courtesy to the committee discussion and those in the audience, and we ask that you silence your cell phones and pagers. Thank you.

CHAIRMAN SALOMON: Before I turn to Phil and Carolyn to provide an FDA introduction, I just wanted to share with you sort of a brief strategy for today as far as I am concerned as Chair.

We have many of us sitting around the table right now, with a few additions to increase the expertise here, but most of us have heard the first round of this issue, with the first child that developed leukemia from the retroviral insertion in the LMO-2 site just before the first of the year.

And so I think it is very important that the focus of this meeting begin solidly on the framework that the committee set at that time and ask the major question; and that is, what has changed?

At the time, we talked about what would happen if a second case was found. How would that change the risk/benefit ratio calculations that we made? How does that serve the community and affect the stakeholders with severe combined immune deficiency disease? How does that affect the broader stakeholder community that now includes anyone who might benefit from various kinds of retroviral gene therapy and the companies and academia, et cetera, that are involved.

So I think the key thing here is to take the foundation of what we did last time. It is perfectly appropriate to say, well, we were right about this, and it incrementally changed in such and such a way, or that we were wrong, and we need to see it.

But I'd like to see the focus building on what we set up previously, rather than starting all over again and taking everything apart and taking a really long time to go over what I think -- we already plowed a lot of ground here, and I would like to see us go forward from this point onward. Phil.

DR. NOGUCHI: Thank you, Dan, and I would like to thank everyone who is participating today on behalf of the Center for Biologics. Our grateful thanks for coming here.

This is extraordinarily difficult and is one of a number of things that I would like to just introduce. I presented something like this at the last meeting, but why are we here?

We are here to acknowledge that there continues to be extraordinarily difficult diseases and the treatment of them remain hopeful, but are always a problem in terms of balancing risks and benefits.

And we are here to affirm that the way to get at this ideal is to do rigorous clinical trials. As we do this, and as we are seeing today, this is not a static evaluation, but it is a continuous balancing of risks and benefits.

We are here to further learn about adverse events as they occur, and as they are scientifically researched, and I think some of our discussion should be oriented towards taking what we now know and thinking of strategies for mitigation and elimination of these particular side effects, as well as strategies for side effects in general, where we know more on a molecular basis than we did even just a few months ago.

And really we are here to confirm that this is exactly where we should be, and we look forward to a very vigorous, a very timely and a very important discussion.

Dr. Wilson will now follow this and really step us through a little bit of the background from the last meeting and lay out the framework and the pathway for our discussion today.

DR. WILSON: While they are getting it set up, I am just going to introduce that what I will be doing today is to try to provide an update for the committee since our last meeting in October.

And that update will include, actually, a revisiting of the discussion that this committee had, so that analogous to what Dr. Salomon was just saying, we can remind the committee what we said in October, and then move forward from there.

(Brief pause.)

CHAIRMAN SALOMON: This is a new California thing where you kind of start the meeting off with a Zen meditation. I hope you appreciate that.

DR. WILSON: Thank you, everyone, for your patience. And so, again, what I am going to do for you this morning is to try to provide an update so that you can have all the information at hand in terms of what has happened since the October meeting, so that you can move forward in your deliberations on these important safety issues regarding the clinical use of retroviral vectors for gene therapy.

My update will include, as I mentioned, a review of the consensus points from the discussion of the October meeting, subsequent FDA actions that we have taken both in response to the October meeting as well as to the subsequent notification in late December from Dr. Fischer's group of the second child with the T-cell expansion.

I am going to provide a quick overview of active clinical trials that are under U.S. IND that use retroviral vectors, and then provide some information from other committees that have been deliberating on these issues, in particular the recommendations from the Recombinant DNA Advisory Committee that met earlier this month, as well as some of the reports from various international bodies as well.

And then finish with just a quick read-through of the questions for the committee so that you can have these in your mind as we move forward with today's agenda.

So to start off, the one general consensus of the committee in October was that the T-cell expansion seen in the X-SCID patient treated in France was likely due to an insertional mutagenesis effect of the retroviral vector used in the gene therapy.

With this as a premise in terms of a major conclusion, then you move forward to address the following question which we asked you in October; namely, are there additional data or measures that clinical investigators need to provide before future and present clinical trials in SCID patients should proceed in the U.S.?

Please consider in your discussion each of the following as they pertain to X-SCID or other forms, such as ADA. And we asked you to discuss six different issues: risk/benefit, informed consent, alterations to cell dose or vector dose, mapping of vector insertion sites, and alterations in vector design.

And what I would like to do in the next few minutes is just provide for you, again, your committee consensus on each of these points so that you know where you were in October.

So regarding risk/benefit, all agreed that, with HLA-identical donors for SCID indications, the benefits of this particular treatment far outweighed the potential risks of using gene therapy.

So the committee recommended that patients who have available HLA-identical donors should be excluded from gene therapy clinical trials. However, the case for haploidentical transplants was not as clear-cut, in that you do get 90 percent survival if the transplant is done in the newborn period.

Survival rates are lower, depending on the transplant center, when it is done later in life. But in both cases, you don't seem to get B-cell reconstitution requiring life-long administration of an IgIV.

And there was a general consensus that the quality of life for even those patients, quote, "who are surviving" is suboptimal, with recurring infectious episodes and other complications.

The other major point that was brought up in the discussion of risk/benefit issues was the analogy to cancer treatments, which often carry risk of secondary cancer, yet are very effective at treating cancer and would not be thought of as being eliminated because of the risk of secondary cancer.

As far as the risk of gene therapy, all acknowledged the obvious statement that if you don't get gene transfer, then, of course, it is a safe procedure. But in the case of the trial in France, clearly there is gene transfer and it is having a therapeutic effect.

And there is actually 100-percent survival at this time, even now with the second child having a leukemia-like illness. And the other important point that was brought out was that the success of Dr. Fischer's trial may be related to the fact that he treats patients de novo, and he is not using patients who had failed haploidentical transplants.

So the committee felt it was important that gene therapy in this context not be considered a salvage therapy but that families be given a choice to have a haploidentical or gene therapy.

And so, essentially, in October the committee consensus was that trials in SCID indications should proceed, but there were some caveats on that conclusion; namely, that clearly changes to informed consent documents needed to be made. You felt it was important that all retroviral vector clinical trials should have these revisions to reflect this event.

The informed consent document should be written in clear language. It should not include mitigating factors, such as issues regarding multiple hits, or number of patients treated, and that it should be clear in saying that gene therapy caused the leukemia, while still emphasizing the unknown quality regarding the risk to an individual patient.

On the third part of this question regarding cell dose, there was some information provided by the committee regarding, for example, use of cord blood, that you may be able to reduce the numbers somewhat and still maintain engraftment.

The number that was thrown out by the committee was one times ten to the fifth CD34 positive cells per kilogram. I recognize that you had extensive discussions of this yesterday, so you probably are much more informed on this, at this point, than I am, and so I will just leave it at that.

But, essentially, after going about this issue, the committee felt that based on 30 years of experience that there are certain acknowledged minimum cell doses that are required in order to get engraftment, and that to go below those doses would provide -- would put these children at risk of additional infectious disease complications.

So, actually, the committee recommended moving forward on researching how to better target the true hematopoietic stem cell, and that would be a more effective way to reduce the cell dose.

Regarding vector dose, this was really felt to be a research question. People also felt this wasn't a huge issue at that time because most retroviral vectors hit about one copy per cell.

However, CBER does want to point out that we are starting to see novel vector systems that reach significantly higher copy numbers per target cell, and that this is an issue that the committee may need to revisit.

In terms of insertion site mapping, the way that we had phrased the question was regarding lot release of ex vivo transduced cells. The committee quickly pointed out that this was not scientifically or technically feasible, and this was obviously rejected.

However, there was a strong recommendation from the committee that patient samples should be monitored closely for outgrowth of single clones, and that if a monoclonal integrant is observed, that it should be sequenced, following with additional phenotypic analyses. And it was pointed out that this information regarding the integration site may inform clinical treatment, and it may allow earlier treatment.

The committee recommended that this type of monitoring be performed at approximately a 3- to 6-months interval. It was thought that this should be done on an active basis, rather than archiving of samples. And, obviously, if there is no vector positive cells in the peripheral blood, then there is no need to perform this monitoring. But the committee recommended that each protocol needs to develop a monitoring plan, including trigger points, for additional analyses.

And, finally, the committee left a caveat that if a particular sponsor felt that it wasn't justified to do this monitoring in their clinical trial, then the FDA should consider this, obviously on a case-by-case basis.

Regarding vector design, the question here was whether or not you could increase the safety of retroviral vectors by minimizing the effect of an enhancer to activate neighboring or distal genes. And while everybody on the committee felt this was an important research question and made the recommendation that it should be further studied, nobody felt that this was something that needed to be changed now for clinical trials to proceed.

And, in addition to that, it was strongly recommended that preclinical models be developed to assess the risk of vector insertion for new vector designs.

Now, following the meeting in October, CBER then sent out letters to sponsors requesting first of all that all sponsors using retroviral vectors revise their informed consent, and we actually have specific template language that we recommended in that letter.

In addition, we asked that, in a subset of those trials using CD34 cells, sponsors submit plans for monitoring for integration clonality, and the interval that we recommended was every six months for the first five years, and yearly thereafter for the next ten years.

And then when a predominant clone is identified, that there be a second test within three months from the first, that the sequence be determined, and that you monitor the subjects closely for signs of malignancy.

So to summarize then, there were three sets of letters that were sent subsequent to the October meeting. In the clinical trials in SCID, which were the subset which were on hold at the time we met in October, we required that they revise their informed consent and develop plans to monitor clonality in order for those trials to proceed.

In indications that use hematopoietic stem cells or other stem cells, these two conditions were recommended, but not a requirement. And in all other retroviral vector clinical trials, we requested only that they revise their informed consent, and at that time did not recommend developing plans for monitoring clonality.

And we confined this to hematopoietic stem cells, this issue of looking for clonality, based on the idea that stem cells are by their very nature long-lived, with high proliferative capacities, and therefore more at risk to the effects of vector integration mutagenesis.

Now, as you know, in late December, we received a second report from Dr. Fischer that there was another subject in his X-SCID gene therapy clinical trial that developed a leukemia-like illness at 34 months post treatment.

And subsequent to that initial report, preliminary data was provided to us by Dr. Fischer that there seemed to be, again, a monoclonal retroviral vector integration 5' to the LMO-2 locus, which I am sure you recall is the same locus, but a slightly different site within that locus, that was observed in Patient No. 4.

And so these preliminary data suggested that we were probably looking again at an incident where retroviral vector integration was likely playing a role in the T-cell expansion.

So in response to these data, we took the following actions: We again sent three sets of letters to sponsors. In this case now, we changed the issues of informed consent and clonality -- monitoring clonality plans to a requirement.

In other words, all clinical trials that used hematopoietic stem cells as their target for ex vivo transduction were now put on hold until they met these conditions.

Inactive trials that use hematopoietic stem cells were told that if they ever wanted to resume their trial, that again they would need to meet these conditions. And then all other retroviral vector clinical trials, we didn't put these on hold, but we made a recommendation that they now would also need to develop plans to monitor for clonality.

Now, what are the clinical trials that we are talking about? In terms of target cells, you can see that really fully half of the current active clinical trials are in either CD34 hematopoietic stem cells or bone marrow cells.

The next largest category are in lymphocytes of various types. Then we have a few in fibroblasts. The other category includes smooth muscle cells as well as a number of different tumor cell types.

And then we have two each that are either direct administration of vector producer cells, or of the retroviral vector itself. In terms of indication, I have broken this down for you two ways. One by total active clinical trials using retroviral vectors, and in the blue bars, this shows that subset that are in the CD34 bone marrow target cells.

So in cancer, we have 8 out of 28 total trials using CD34 cells for the ex vivo transduction, 8 out of 12 in HIV. The miscellaneous, these two are osteogenesis imperfecta and multiple sclerosis.

And almost all of our trials in genetic disease target CD34 cells. And these include, of course, the SCID indications, chronic granulomatous disease, and Fanconi's anemia.

So at this juncture then, it is obviously prudent to ask whether or not there are additional modifications that we need in order for these types of clinical trials using retroviral vectors to proceed safely.

And what I would like to do for you now is to just outline some considerations, some which are more theoretical and would require more research before they could be immediately implemented in clinical trials, and others which are potentially more practical and could be implemented more directly.

So the first, which is a more practical solution, is again to revisit the issue of dosing with the idea that if you reduce the total load of vector integrants that this reduces the risk of integration into a potentially, quote, "bad locus"; a locus that might be tumorigenic.

And one could envision doing this by reducing the dose of vector used in the transduction, reduce the dose of cells given back to the patient, or potentially changing the dose so it is based on a total number of vector integrants.

In addition, as was recommended by the committee in October, we want to consider whether or not we might need to have additional preclinical studies to assess the carcinogenic potential of a particular vector backbone-transgene target cell combination.

And there is one of two ways that one could envision doing this. One is to perform traditional carcinogenicity testing at an earlier stage of clinical development. This type of testing is usually done at the time a sponsor gets to licensure.

And the reason for that is because these are multi-year, very expensive studies. So we don't usually ask people to do this for a Phase I trial.

Alternatively, one might consider some of the newer accelerated models of tumorigenesis, such as transgenic models carrying oncogenes or knockout models of tumor suppressor genes. One might also consider a complementary approach of both assays. One might consider doing these concurrently with an early phase clinical trial.

In terms of cell target or culture conditions, again these are much more on a theoretical, but again coming back to if we could identify what is a true hematopoietic stem cell.

One issue that is acknowledged regarding retroviral vector or retroviral integration in the genome is that it tends to occur in sites of transcriptionally active genes.

So if you could identify a transduction protocol, for example, by studying gene expression by gene or protein microarray that has a fewer number of transcriptionally active genes, this might reduce the number of integration targets.

And then, ideally, if you could actually identify cells with vector integrants into known tumorigenic sites, this obviously would be nice to be able to do, but clearly on the realm of way theoretical.

In terms of modifications of the vector, again, these are things that are in the research phase. People are starting to look at the addition of insulator sequences.

These have really been done more to look at the ability of these sequences to insulate from the effect on the enhancer becoming silenced, and it has not really been studied whether or not these elements would also have the effect of blocking enhancer activation of neighboring genes.

But that is something that probably should be studied. Deletion of retroviral vector enhancer elements within the LTR would clearly reduce the risk of enhancer activation of neighboring genes, and if you could develop a vector with targeted integration, this would be ideal, but again much more on the theoretical. And it is not known at this point whether or not specific transgenes might also play a secondary role in the tumorigenesis.

So I want to finish then with recommendations from other advisory committees. The first is the NIH Recombinant DNA Advisory Committee.

They have had two meetings on this topic. The first was in December of last year, and the second was just earlier this month on February 10. And they made the following observations: The majority of children in this X-linked SCID gene transfer study have had major clinical improvement to date.

The gene transfer was a cause of both leukemias and the occurrence of leukemia in this protocol is not a random event and constitutes an inherent risk in this study.

So based on these observations, the RAC recommended the following two main points: Pending further data or extenuating circumstances, retroviral gene transfer studies for X-linked SCID should be limited to patients who have failed identical or haploidentical stem cell transplantation.

And then secondly, there are not sufficient data or reports of adverse events directly attributable to the use of retroviral vectors at this time to warrant cessation of other retroviral human gene transfer studies.

Such studies may be justified contingent upon appropriate risk/benefit analysis, accompanied by an implementation of appropriate informed consent and monitoring plans.

The French gene therapy working group met at the end of January of this year. They have recommended that the X-SCID clinical trial maintain its clinical hold status pending additional scientific investigations of the most recent event, and they also recommended that safety of retroviral vectors should be improved by, for example, deletion of enhancers, addition of insulators, and also by development of relevant and predictable animal models.

Germany's Commission for Somatic Gene Therapy has also met twice to consider these issues. In their most recent meeting, they went case by case through the active clinical trials using retroviral vectors.

Based on that analysis, they recommended that one protocol using hematopoietic stem cells for the treatment of chronic granulomatous disease remain on hold pending additional data, while the other four active protocols were recommended to proceed, obviously requiring changes in informed consent to make patients and their families aware of these events. And in some specific cases, additional protocol changes were recommended regarding patient inclusion criteria and so on.

Italy's National Health Institute has recently come out with a continued ban on all gene therapy for an additional four months, but they do say that they may approve trials on a case-by-case basis. Likewise, the U.K. Gene Therapy Advisory Committee is meeting in March of this year. They are currently looking at trials on a case-by-case basis, considering particular circumstances for each situation.

So, finally, I just want to quickly go through the questions for the committee, and the first question is specific to the subset of trials in SCID indications.

Please discuss under what conditions clinical trials using retroviral vectors to transduce CD34+ hematopoietic stem cells for the treatment of SCID may resume.

Currently, before sponsors of these trials may proceed with clinical trials, they need to provide a revised informed consent document and plans for monitoring of peripheral blood cells for the clonality of vector integration.

So the specific question then we have, is this sufficient, or should additional conditions be placed on these trials. We ask you to consider whether dose should be altered, whether we need to include additional pre-clinical studies, whether we need to look at the particular target cells, or have alterations in the vector design.

The second question is identical to the first, so I won't read it, with the notable exception that we are now asking you to consider really these same issues in other clinical indications where retroviral vectors are used to transduce CD34+ hematopoietic stem cells.

And then the third question regards the use of retroviral vectors for marking studies where no direct therapeutic benefit is possible, whether or not these should be allowed to proceed, again, considering issues regarding risk/benefit, a particular vector design, or target cells.

And then the final question, if time permits, that we would like you to consider is: Given the properties of lentiviral vectors to transduce and integrate into non-dividing cells and their increased efficiency of transduction, please compare the current requirements for clinical use of gamma retroviral vectors to those that should be in place for use of lentiviral vectors.

We currently are requesting that sponsors revise their informed consent and monitor peripheral blood cells for the clonality of vector integration. Is this sufficient or should additional conditions be placed on these trials?

So, with that, I thank you for your attention. I am going to ask that the committee hold questions after my talk so that we can continue with our guest speakers, and there will be plenty of time in the afternoon to ask myself or other members of the FDA questions about what we have done. Thank you.

CHAIRMAN SALOMON: Thank you very much, Carolyn. That was excellent. Before we start with the guest presentations this morning, I would like to go around the room and have everyone briefly introduce themselves.

I apologize for not having done that earlier. It has just gotten a little mixed up, as we were here yesterday in the same room, but it was a different group. So, Rich, could you start us off.

DR. MULLIGAN: I am Rich Mulligan from Harvard and Children's Hospital, and my expertise is in gene transfer and in stem cells.

DR. ALLAN: I am Jon Allan from the Southwest Foundation for Biomedical Research, and my expertise is in retroviral pathogenesis in non-human primates.

DR. KURTZBERG: I am Joanne Kurtzberg from Duke University. I direct the Pediatric Bone Marrow Transplant Program, and my expertise is in stem cell transplantation and cord blood transplantation.

DR. TSIATIS: I am Butch Tsiatis. I am from North Carolina State University, and my expertise is biostatistics.

DR. MURRAY: I am Tom Murray. I am from The Hastings Center, and my expertise is in the ethics of research with human subjects.

DR. MACKALL: I am Crystal Mackall, and I am from the Pediatric Oncology Branch of the Intramural Program of the NCI. I practice pediatric oncology and study transplantation immunology.

DR. COFFIN: I am John Coffin from Tufts University and the National Cancer Institute. My research interest is in molecular retrovirology.

MS. LAWTON: I am Alison Lawton. I am the Industry Rep., and I am head of regulatory affairs for Genzyme Corporation.

DR. RAO: I am Mahendra Rao from the National Institute on Aging, and my interest is in stem cells.

CHAIRMAN SALOMON: Dan Salomon, from the Scripps Research Institute. I am in organ and cell transplantation and xenotransplantation. We do gene transfer studies as well.

MS. DAPOLITO: Gail Dapolito, FDA, Center

for Biologics, Executive Secretary for the committee.

DR. HARLAN: I am David Harlan. I am an intramural investigator with the National Institute of Diabetes, Digestive and Kidney Diseases, and my interest is in transplantation immunology in diabetes.

DR. HIGH: I am Kathy High. I'm at the Children's Hospital of Philadelphia, and my interests are in gene transfer for hematologic disease.

MS. LEONARD: I am Warren Leonard in the National Heart, Lung, and Blood Institute. My lab discovered the genetic defect in X-linked SCID, JAK-3 deficiency, and IL-7 receptor deficiency, and my expertise relates to gamma-c-dependent cytokine interactions.

DR. ROSE: Stephen Rose. I am in charge of the recombinant DNA activities in the Office of Biotechnology Activities in the Office of the Director of NIH, and Executive Secretary of the NIH Recombinant DNA Advisory Committee.

DR. RASK: Cynthia Rask from FDA, in the Clinical Evaluation and Pharmacology/Toxicology Division, in the Office of Cellular Tissue and Gene Therapies.

DR. WILSON: Carolyn Wilson in the Division of Cellular and Gene Therapies.

DR. NOGUCHI: I am Phil Noguchi, Acting Director of the Office of Cellular, Tissue and Gene Therapies.

DR. PURI: I am Raj Puri. I am the Acting Director of the Division of Cellular and Gene Therapies.

DR. CORNETTA: I am Ken Cornetta from Indiana University. My interest is in retroviral gene transfer, and I coordinate the National Gene Vector Lab for the NCRR, and I am also a clinical bone marrow transplanter.

DR. FRENCH: Jeff French from the National Institute of Environmental Health Sciences. My interest and expertise is in alternative models for studying carcinogenicity.

DR. WOLFF: Linda Wolff from the National Cancer Institute. My expertise is in retroviral pathogenesis in small animal models.

DR. TORBETT: Bruce Torbett from the Scripps Research Institute. My interest is in HIV vector design as well as myeloid transcription regulation.

MS. MEYERS: Abbey Meyers, President of the National Organization for Rare Disorders. I am the Consumer Rep.

MS. BALLARD: I am Barbara Ballard. I am here to represent the SCID families, and I am also on the Board of Trustees for the Immune Deficiency Foundation.

CHAIRMAN SALOMON: Just a point for those of you who are not used to it: when you are done speaking, if you will shut the mic off, it prevents feedback from happening. Thanks. So if you see me doing something like this, it is not, you know, be quiet, it's to shut your light off.

Okay. Let's get started then. It is my pleasure to introduce our first guest speaker from the Hospital Necker, Marina Cavazzana-Calvo.

DR. CAVAZZANA-CALVO: Good morning, everybody. First of all, I would like to thank Dr. Carolyn Wilson and Ms. Dapolito to give me this great opportunity to share with you our updated results concerning all the trials and the two side adverse events.

Nevertheless, I would like to just make one comment before I initiate my talk. This protocol is under the authority of the French government, and each public disclosure of this data must be made with our approval. Thank you very much.

So let me remind you that X-linked severe combined immune deficiency is characterized by the complete absence in the peripheral blood of natural killer cells and T-cells. This is due to the fact that there is a gene mutation on the gamma-common chain that is shared by IL-7 and IL-15.

The gamma-common vector, retrovirus vector, we used is described here. The gamma-common cDNA is under the transcription counter of LTR of the retrovirus, and we used an MFG vector kindly provided by Dr. Richard Mulligan, and the classical psi-crip packaging cell line set up by Olivier Danos.

The envelope of this retrovirus is an amphotrophic one, and no other relevant characteristic of this retrovirus vector can be underlined.

So I resumed here the pre-clinical studies that we conducted in order to obtain the approval through the clinical trial.

First of all, we have demonstrated as have two other groups in the world, that the B-lymphocyte transformed by EBV, derived from the patient and the gamma-c negative, can be efficiently transduced by our retrovirus vector.

This demonstrates the feasibility of protein expression and the function in terms of gamma-c transduction signaling and the stability of the retrovirus expression over time. This cell line has been followed over six months.

The second step was to demonstrate almost in vitro that the CD34 positive cells gamma-c negative, obtained from the patient at the time of general anesthesia through the central line, can be efficiently transduced in vitro and can restore, almost in vitro, NK and T-cell differentiation.

In parallel, Gemmady Santos in our laboratory set up a knockout mouse model without the gamma-common chain, and this model was indispensable to demonstrate no toxicity. This is correction selective advantage.

Just some details on the animal model to implement the discussion. The bone marrow site was derived from the donor gamma-c deficient animals. The animals had been pre-treated by 5-FU and the whole bone marrow has been transduced with a retrovirus vector very similar to that used in the clinical trial, and for this model he carried a marking gene that is the human CD2 molecules.

The recipient of the double transgenic --double knockout mice gamma-c deficient animals was low irradiated.

This is the survival curve of treated animals in comparison with non-treated animals. Non-treated animals in the standard animal facilities died within 15 weeks. Conversely, the treated animals are 100 long-term survival course.

How long have these animals been observed before sacrifice? For the primary transplantation up to 47 weeks and the animals show no toxic effect. The same cells have been used from the primary transplantation of eight-weeks-old aged animals to transplant secondary mice to demonstrate again the efficacy of the transduction of the stem cells, and here again we obtained a correction of animal disease with no toxic effect.

So on the basis of this preliminary design completed by the translation medicine to make a scale-up study, we obtained the approval from the French authorities to go to the clinical study with this protocol.

Patients were eligible for these trials when they have gamma-c gene mutation, of course, and in the case where they have no HLA gene-identical familial donors, and the presence of informed consent from the family.

The protocol is the following one. Marrow is harvested under general anesthesia and very small quantities of bone marrow cells are harvested, around 50 up to 150 ml per child.

CD34 positive cells are selected by immunomagnetic microbeds, and pre-activated in vitro for one day in the presence of the stem cell factor FLT-3 ligand, MGDF and IL-3. I underline the fact that we used very high doses of stem cell factor FLT-3 ligand following the protocol described previously by Eve Sconis in the Vancouver laboratory, 300 milligram per milliliter.

MGDF was used at a nanogram per milliliter and IL-3 in very low doses.

After this 24 hours of preactivation, the cells are transduced with a supernatant containing the vector in bags precoated with fragments of fibronectin. And it is made each day for three consecutive days. At the end of the treatment, which began on Monday morning and finished Friday evening, the cells are extensively washed and transfused intravenously into the patient without any additional therapy.

These are the 10 patients, the 10 newborn patients with classical SCID form, that have been enrolled so far in this clinical trial. As you can see the age is under one year, and the severe adverse event occurred in two patients, the youngest. One was one month, and the three-months was without any infection at the time of treatment. Conversely, all the other patients had severe infection at the time of diagnosis

And I would just like to stress that these two patients -- because I think it is fruitful for discussion -- these patients had been admitted to the hospital with a disseminated varicella-zoster infection which affected the central nervous system.

And we know there is no way to obtain a definition that this varicella also disseminates infection with the classical haploidentical bone marrow transplantation.

In the second case also there is lympthoproliferative disease with the spleen and the lung affected, with a very, very poor prognosis for the haploidentical bone marrow transplant.

Four patients had maternal T-cells at the time of diagnosis, with graft vessels soft-like lesion on the skin: this patient, and this one, and these two.

And two patients had an endogenous gamma-C process expression, and we tested before gene therapy for the presence of transforming into a negative effect.

This is the characteristic of infused cells for these nine patients, and so this is the mean and the standard deviation.

Each patient received, except for one patient that I will describe in detail, a huge number of CD34 positive cells, 24 million per kilogram with a large variation between the patients.

The total gamma-C transfused cells is also very huge, with a mean of 17 million per patient with a huge standard deviation, and the double-positive CD34 gamma-C per patient is 9 million plus-minus 7, because there is one patient, Patient 6, that received a very low number of CD34+ cells.

But you can see that the B cell reconstitution is less good and the percent of CD34 gamma positive cells is finished one year after transplantation. So a discussion of the dose of this patient can be very interesting.

The proliferation rate for bone marrow cells is very high, and then you have a fold increase in the number at the end of a manipulation of 6 up to 10-fold. And in particular, I'll describe the two patients who experienced the side effects, P4 and P5 patients.

And in these cases, the cells proliferated even more. The proliferation rate was up to 8 and 9.5 for these two patients who were the youngest ones, and I think that for these newborn patients, aged similar to these ones, that the biological characteristics of stem cells are more close to cord blood than to bone marrow cells. And it is important to stress that for this discussion.

The number of CD34 positive cells that they received was very high: 44 million per kilogram for the first one, and the 42 million for the second one. The total gamma-common positive cells was even more huge, 31 and 24. You can have here CD34 negative cells, but active lymphoid precursor cells.

And the double-positive hematopoietic precursor cells is up to 20 million per kilogram for this patient, with a high proliferation rate.

What about the kinetics of lymphocyte T-cell reconstitution in this different case? This is the patient that we followed actually, and this is P4 and P5 patients that have experienced the side effects.

In this case, the kinetics of T-lymphocyte recovery is quite light, and it begins 75 days after transplantation of the gene-modified stem cells. And they grow within the first three months. They can recover a normal number of T-cells that permitted the discharge of this patient from the hospital.

Recovery increases for the first year, and then it goes down, as is usual for patients of this age.

Strangely, for these two patients, the kinetics of the T-lymphocyte recovery is quite different. First of all, the appearance of T-cells in the peripheral blood is very fast. Within the first month, they recovered a high number of T-cells, and the other difference is the shape of the curve. This table remained high over the time, and at this moment we have no worry about this.

And, abruptly, pretty much at the same time the number of T lymphocytes goes up, and the patient is cured of the lymphoproliferation.

So the characteristics of the transduced T-cells for all the patients enrolled in this trial are the following ones: normal count and subset distribution; normal repertoire, analyzed by two technologies, immunofluorescence and immunoscope; and normal phenotype. The T-cells are the naive phenotype, not a memory one. This is normal for a new appearance of a neurological system.

They present a good thymopoiesis, with the presence of recent thymic emigrants in the same number as normal patients. This is another element for discussion for the comparison between haploidentical bone marrow transplantation and gene therapy.

They have a normal function, and by Southern blot analysis, they contain one copy of a provirus per cell. NK cells after gamma gene transfer, you know that this lymphocyte subset is missing in this population.

And we observed the correction for the first year after gene therapy, and then we observed a decline over the time. And it is quite comparable to that observed after haploidentical bone marrow transplantation.

So from a natural killer cells point of view, this result is very similar to that obtained by bone marrow transplantation. The good news from this clinical trial, and which was not expected at all, is the correction of B cell compartment.

For these children, we had stopped at six months after gene therapy the immunoglobulin substitution, and they recovered a normal level of serum IgM ,IgG, and IgA antibodies.

This is the follow-up data as of February 1, 2003. These are the two patients with the adverse effects. The other one, the follow-up, is closer to four years for the first two patients treated.

This is the failure of gene therapy treatment. This patients never restored a T-cell compartment. He has been grafted with a match unrelated donor, and he is doing well. All the other patients, the follow-up is variable between 1.7 years and down to less than one year after gene therapy. All of them are doing well, with a complete restoration of immunological function, and at the end of this talk I will give you much more information about the immunological state of all these patients.

Questions about the gene therapy are how many cells were efficiently transduced to obtain a T-cell compartment, the characteristics of the transduced cells, and the duration of the immunodeficiency correction.

For the first question, it is important for us to have a diverse repertoire of T-cells because a restricted T-cell repertoire does not permit these children to defend against infection. So the T-cell repertoire after gamma gene transfer for all the children that restored a T-cell compartment has been a diverse one, with no abnormalities detected so far for the treated children, either by immunofluorescence or immunoscope analysis of the CDR3 lengths.

The other point is the integration of the transgene more than one year post gene therapy. You can observe here that we demonstrate once again that there is a strong selective advantage for T-cells and NK cells over the other hematopoietic cell lines.

One hundred percent of T-cells and NK cells have integrated and express the transgene, and less than one percent, or around one percent of B-cells, but this can be important for B cell reconstitution, and it is variable between 1 percent and 5 percent, and less integration for monocytes and polymorphonuclear cells.

I would like to stress that the polymorphonuclear cells that express the gamma-common chain persist over the time at the same level that you can detect a few months after gene therapy. So stated as a concept, probably we have obtained a transduction of CD34 autorenewal of stem cells, or a very primitive one in any case.

Worries about the T-cell reconstitution of these patients. This is a magnetic resonance imaging of the P5 thymus one year after transplantation, kindly provided by Dr. Sorensen.

And you can see, not very well, but there is a thymus, a sizeable thymus that is of normal size. The other point is the function of the thymus that we have studied by the investigation of the TREC. That is that an episome drives it from the DDG recombination in the nucleus that can be measured and tracked for the patient.

And all these patients have normal thymopoietic activities. This is in contrast with the decline of thymopoiesis over the year for the patient transplanted from a haploidentical donor.

About the duration of the correction of this immune deficiency, we have performed some studies on the CD34 positive cells recovered one year after transplantation for some of these children.

Here we see the LTC-IC frequency in P2 and P4 patients 22 months and 30 months after gene therapy. We have a normal frequency of CD34 immature cells. And five weeks after long-term culture and limiting dilution analysis, we have detected a high number of gamma-c expressing CFU-GM up to five percent. This is another proof of the idea that we have transduced some stem cells.

This is the study about this patient, realized thanks to the great collaboration that we have with Christof von Kalle's group. We can see here this mirror of the high number of integration sites for CD3 cells at 13 months after gene therapy for the P4 patient, and he has been able to track the same integration site in the CD3, CD15, and LTC-IC on the basis of the diagram that I presented previously.

So we can say that in this trial we have transduced some T-cells, polymorphonuclear cells, and the monocytes. So the idea is that we can transduce the common lymphoid precursor cells with common myeloid precursor cells.

And if we can track the same integration site, this can lead us to conclude that we have transduced pluripotent progenitor cells.

The persistence of these effects up to two years after gene therapy and the fact that Christof von Kalle has been able to detect the same integration site permits us to conclude that we have transduced hematopoietic stem cells with high proliferative capacity and probably with self-renewing ones.

Limitations of gene therapy. We have had a failure in a child with an enlarged spleen that received the VZV vaccination before diagnosis, and the two side effects in two children.

This is the situation for the first patient up to month 30 after transplantation. He showed rapid T-cell development with the polyclonal repertoire, and the development of T and B cells immune responses, including to varicella-zoster infection at month 30.

Christopher can detect about 40, up to 60, integration sites in T-cells, and is doing well up to this date. What happened after? He showed an increase in the T-cell count, overall for the gamma-delta T-cell compartment and the TCR, this particular clone of gamma delta positive cells, had a slight increase of 7,000 per microliter over the time and overall after the chicken pox infection.

Nevertheless, after he cured this infection, the number of these clones continued to grow during the summer, and at the end of August of last year, they went up to 300,000 lymphocytes per milliliter with evident clinical signs. At this moment we began the chemotherapy, the classical one for high-risk acute lymphoblastic leukemia because we don't have available at this moment the monoclonal actibodies against this clonal proliferation, but now we have one for him.

The plan for this child is that he has been treated with a conventional conditioning regimen this week, and he is to be transplanted next week with a match unrelated donor, ten-out-of-ten antigen identical.

The analysis of this monoclonal proliferation -- and you know some data already -- is that he is a monoclonal gamma-delta T-cell clone. And he has a signature for the coming of this clone from the proliferation of one single cell, because, both by immunoscope and by TCR sequences, all the cells have the same TCR. The cells have the appearance of blast cells. They are extremely mature because they have only the marker of gamma-delta mature T-cells of a memory phenotype.

They are CD3, 5, 7, 28, and 45RO positive, and they all express gamma-c, and they don't express any antigen for immature cells. In vitro, they proliferate in the presence of IL-7 and IL-15. And these cells present chromosomal abnormalities with a translocation of 6 and 13 at the time of the treatment.

And the first question raised by the authorities and by everyone in the scientific world of gene therapy was the detection of the replication of competent retrovirus. There is a contamination. And we have performed -- these children are forwarded for six months for the detection in the biological fluid of RCR, and all the detections made are negative, of course.

And so we performed the classical mobilizing test on the master donee, and Philippe Leboulch in Boston helped us with the match with other tests because we were afraid that the test was not sensitive enough to detect some competent retrovirus.

So he did a Southern blot analysis. This is the patient, the P4, and this is the B cell count. We then entered a probe and an RT-IN probe, and in all the cases, he was unable to detect any RCR.

And the question arose from our authorities and even in the first meeting that we had with you about the possibility that at the time of the production of a retrovirus some murine VL-30 retrotransposon are packaged with the retrovirus.

So, again, Philippe LeBoulch in Boston helped us with this test. We sent him the cells, and he was unable to detect VL30 in the blast cells of this patient with this probe.

So these cells have one provirus integration site detected by LAM-PCR in chromosome 11 in the short time within the LMO-2 locus. There is another expression of LMO-2, and this integration site is detectable at least from month 13 and it is not detectable at all in the CD34 positive cells, but it is normal.

We have freed some of these CD34 positive cells just at the end of the transduction protocol, and we have sent some cells to Christof. But probably the technology is less sensitive, or is not sensitive enough to detect this integration site on the CD34 positive cells. We can conclude that the analysis, immediately after the transduction, is unable to detect the "dangerous" integration sites.

This is the structure as it is known so far for the human LMO-2 gene, and this is the integration site for the P4 patient in the first intron in the reverse position.

So the question was: What is the mechanism of hyperexpression of the LMO-2 gene? And before speaking about the aberrant expression of this gene, this is the sequential immunoscope study of the TB delta one population for this child.

And you can see here that for the SCID disease where we are required to reconstitute or to restore the immunological compartment, you can follow this patient, or this type of patient, by immunoscope.

You can see here the growth and distribution of the gamma-delta T-cell repertoire for the control people. This is the situation of this patient at month six, and you can see here that retrospectively, we can detect a very, very mild incrementation of the size of these TCR, 11 percent.

But this clonal type increases much more between the 6th and 13th month after gene therapy, up to 70 percent. But much more than an increase of this clone, we can observe the decrease of polyclonality of the gamma-delta T-cell population.

So the much more striking data is the loss of polyclonality, more than the appearance of a clone. And over time this clone increases, at month 17 up to 52 percent, and you can follow this clone. This is the only one at month 34 after gene therapy.

And these results by immunoscope are strictly concordant with the analysis made by Christof von Kalle of this type of integration site. So I think it could be alternate way to follow the immunological reconstitution of this patient. And we can roughly build this type of curve where we can see that this clone increases over time in a linear way, and that it abruptly goes up after the chicken pox infection.

If you come back to the aberrant expression of this oncogene, we have made a quantification of the LMO-2 RNA messenger transcript in the patient's peripheral blood leukocytes by real time PCR.

The LMO-2 transcripts are ten times greater in patient cells versus normal control cells. You have here the patient, and this is the normal control industry standard. And as you can see, slight contamination by monocytes in the peripheral blood leukocytes can give a positive result.

This is -- the question was: Is LMO-2 hyperexpressed by the same allele that has been integrated by the provirus? And the answer is yes. This is the RNA-FISH study for the P4 patient, and there is a co-localization of the gamma-c probe and the LMO-2 probe.

This is the gamma-c probe in the red, and this is the LMO-2 detected with the green fluorescence probe, and there is a co-localization of the two in the same cell.

The other question was: Is there an aberrant splice of the RNA messenger from this integration? And the answer is no. Only one significant RNA species of the respective size for the provirus on SCID is detected by Northern blot analysis with two different probes.

And the reason for the normal splicing of the RNA messengers, despite the integration of the provirus within Intron 1, is that in this site there is the same splice locus that is present in the physiological intron.

This study has been performed by Philippe Leboulch with our help.

We did help out with the greatest research in the world that has studied the LMO-2 gene. He is permitted to us to be with him in the UK. We have conducted a study on the LMO-2 protein expression by Western blot analysis, and you can see here that we have a tremendously high expression of this protein on the patient P4.

And it is equivalent to what we can detect in murine erthyroleukemic cells and very comparable to that obtained in the CHO cell line, transected with the LMO-2 plasmid.

So for the mechanism of LMO-2 activation by integrated provirus, we arrived at this conclusion: RNA of the normal size contains both the first and last exons and the correct junction.

LMO-2 RNA is monallelic. There is Cis-activation by provirus of a normal endogenous LMO-2 promoter with a normal splicing of Intron 2. Two possibilities still remain, but it is much more likely that the first one is the correct one, that there is a Cis-activation by the proviral LTR announcer.

And even if at this point we can't exclude it completely, there is a disruption of a provirus LMO-2 silencer with normal slicing of Intron 1.

This second possibility is under study and in collaboration with Dr. Kathleen Anderson at Cincinnati University.

So what about the overall interpretation of the side effects for Patient 4? The crucial point is insertional mutagenesis with the aberrant LMO-2 expression. But we are always looking for additional factors, such as aberrant clonal gamma-c signaling, the role of varicella-zoster infection, and genetic associativity to the cancer.

What about the gamma-c expression? This is the same as we obtained in the knockout animals treated by gene therapy, and you can see here that it has been deeply investigated by us that the expression of this protein on T-cells, B lymphocytes, and the natural killer cells is strictly normal.

And that there is no hyperexpression in the membrane of this protein. What about -- but we can have normal expression of the gamma-c chain and the hyperactivation of the signaling pathway, but we have no JAK-3 phosphorylation detectable in vivo.

This is the control cells, and we can have the phosphorylation of the JAK-3 tyrosine kinases exclusively after activation with the appropriate cytokines IL-7 and IL-15. This is the control cells for the patient, and if we have no activation in the steady state, just when we bring out of the body the cells, and we have phosphorylation of the JAK-3 tyrosine kinases after simulation. And this is the control.

So far we have no overexpression of gamma-c common chain. The sequence of the provirus of the cDNA is strictly normal. We can't detect any abnormal activation of JAK-3 Stat5 activation, and the study of Stat5, that it could be much more sensitive than JAK-3, is pending. And in the blast cells of this patient, apoptosis is strictly normal.

So at the first meeting that we had with you and with Alain Fischer, we described to you the pedigree of these families. And something that everyone around the table found very interesting was the fact that this family has two cancers, and one sister and one cousin of the affected child. And we investigated at least for this child the possibility that there is a genetic associativity to the cancer.

So we sequenced the p53 protein, the complex interesting in reparation of DNA as the MLH1 gene, and we sequenced thoroughly the ATM gene, and everything is normal as expected.

What about the follow-up of Patient 4 because we are much more interested in the clinical causes for this side effect. This is the bone marrow of this patient at the end of January.

And we can consider that this bone marrow is in complete remission, even if we can detect some cells in the bone marrow after Fikol that still express the gamma-common chain. And we can detect a very low percentage, 0.1 percent, of the TCR gamma9. This is the TCR that is responsible for the clonal expansion at the beginning.

We gave these samples to Christof von Kalle, and he provided us with these extremely interesting results that show that this child recovered a polyclonality between the few T-cells he has. He has no immunological reconstitution, of course.

This child is now conditioned for his transplantation and he is doing well, but we decided that theoretically the persistence of gamma-c positive cells in bone marrow can be expected to restore or it could restore an immunological system by these few cells.

But, naturally, everybody prefers not to run any risk for the clinical status of this patient. And on the basis of the consideration that the appearance of this lymphoproliferation with the characteristics of a high-risk one, we recommended to be as cautious as possible and to perform the bone marrow transplantation.

CHAIRMAN SALOMON: Was that the peripheral blood, the flow cytometry you showed?

DR. CAVAZZANA-CALVO: It is the same, 0.1 percent of this TCR gamma-9 positive.

CHAIRMAN SALOMON: Is that a bone marrow sample?

DR. CAVAZZANA-CALVO: This is a bone marrow sample.

CHAIRMAN SALOMON: But it was the same?

DR. CAVAZZANA-CALVO: The same. We have some more details here. The malignant characteristics for this patient are blastic appearance at the time of diagnosis, the presence of translocation of 6 to 13, and the detection of the hyperexpression of the LMO-2 protein.

This is all present at high levels at the time of the occurrence of this disease. Two months later, after high-dose chemotherapy, we have a complete disappearance of the blastic clone, the disappearance of the translocation, but the persistence of a significant number of T-cells that are hyperexpressing LMO-2 protein.

So we decided to reinforce the obtaining of a complete remission because one month after, the blastic cells can appear again, and we can again detect the translocation. So we reinforced the obtaining of complete remission with three more months of chemotherapy.

Actually, we are in February, and this is 40 months after gene therapy and six months after the occurrence of a leukemic effect, and we have no blastic cells, no translocation either detected with a very accurate method.

We have not done this exploration on the basis that we have 0.1 percent of T-cells, and they are polyclonal in terms of the integration site. So even if we think there is a very, very small quantity of gamma9 positive cells, this is less than the residual leukemic cells normally detected in a leukemic patient.

The therapeutic plan. You know we are at month 40, the patient is under a conditioning regimen, and we have prepared over six months monoclonal antibodies of a clinical grade in order to be able to treat the residual leukemic cells eventually persisting in bone marrow.

We have obtained this clone through a very kind company in France, and we have produced in real time monoclonal antibodies against the TCRV delta-1. These antibodies are now finished, and we have enough quantity to treat the patient in the next month.

We have approval that these monoclonal antibodies can target this specific TCR, and we plan to inject it as soon as we have received the test for eventually the viral contamination of these monoclonal antibodies.

Patient 5 had gene therapy at three months of age, had a huge number of CD34 positive cells infused, and is alive and well up to month 31 with multiple integration sites. He came to the hospital at 34 months with splenomegaly, an enlarged mediastinum, and a huge number of white blood cells, 80 percent of blast.

And chemotherapy has been initiated by Ricardo Sorensen without any delay. The marker for this Patient 5 is they are CD8 TCR alpha-beta positive cells, gamma-c positive, no myeloid markers, and no NK or B cell markers.

We have -- and this is the greatest difference between the two patients -- 3 TCRV beta peaks, with only one integration site. No proliferation is present with the gamma-c dependent cytokines, another difference with the previous patient. And the chromosomal abnormalities appear at the time of diagnosis at trisomy 10.

I would like to stress the fact that these chromosomal abnormalities are not typical at all, even in the first cases, with the chromosomal abnormalities that we can detect in typical ALL.

This is the repertoire of the beta chain for Patient 5 at month 34. The peripheral blood still remains polyclonal, but we have the emergency of three clones, Vbeta1, 15 percent of the total and Vbeta2, 64 percent, and Vbeta23 that we can't detect at the surface is 70 percent.

So these three Vbeta clones make up the totality of the T-cell count in the peripheral blood. And we have tracked this clone by immunoscope in direct respective study, and you can see very easily that at month 30th that we have a very diverse repertoire for these clones. Vbeta1 and Vbeta2, there is no worry about.

At month 31, Vbeta2 is also polyclonal; no increase. Month 31 is September. The leukemia occurred in December. And we just -- this alteration in the immunoscope analysis lost the polyclonality of the beta-1 family.

And at month 34, we have 50 percent of this clone and 64 percent of this one. And thanks to the analysis performed by the Funkel group we can detect the presence of the provirus upstream of Exon 1 in sense orientation in these cases.

So here again we have a normal hyperexpression of LMO-2 gene, with exactly the same result in the P4 patient even. We did an RNA-FISH study, where we can see merge between the probe for gamma-c and the probe for LMO-2.

And the other red point that you see is the gamma-c physiological gene, the two alleles.

The evolution of the P5 lymphoproliferation. It is blastic, there is a trisomy 10, there is hyperexpression of an LMO messenger, and he has three clones, two of them detectable by immunofluorescence.

Everything is present at month 34, and there is a disappearance just one month after the chemotherapy. This is another difference with the P4 patient: this second case is much more sensitive to chemotherapy than the first one.

And the medical doctor in charge of this child detected a disappearance of the blastic cells, and a complete disappearance of the chromosomal abnormalities. We can detect in significant quantity the clone involved in this lymphoproliferation.

Now we are up to 36 months after gene therapy, two months after the beginning of the treatment. There is the disappearance of the blasts and the chromosomal abnormalities.

We can detect some Vbeta1 and Vbeta2 cells that have the same integration site. There is a very significant decrease in terms of the quantity of cells, but we can't formally conclude that this patient is in complete remission but has a normal course for a child who is treated for acute lymphoblastic leukemia.

Just to finish this talk, you have here the immunological analysis of patients with T-lymphocytes, the other one that does well so far.

And you have the months of follow-up here. The first two patients are more than four years after gene therapy, and this is the number of T-lymphocytes for the first patient treated.

And the last determination was made yesterday, and he has 1,000 lymphocytes so strictly normal, probably slightly decreasing over the time. This is the curve that you probably noticed in the first kinetic T-cell reconstitution, that it dropped down much faster than the other patient.

For the other patient the lymphocyte count is strictly normal. This patient even -- this is the patient with the sort of lymphoma proliferation at the time of the diagnosis that I stressed to you at the beginning, and he had a varicella infection two months ago.

We hospitalized him because we were very worried about the varicella-zoster complication, taking into account the first cases, but he was cured completely from his varicella-zoster without any complication, and with a very mild increase of the gamma-delta T-cell compartment that decreased over time after the resolution of the infection.

And this is the blood cell count for the other patient. In terms of activity, these cells proliferate normally to the T-cell stimuli, such as the PHA anti-CD3 monoclonal antibodies, or antigens to the tetanus toxoid and Candinine.

I think that there has been trouble with the patient called by Christof Patient A, and some worries in the scientific journals. This is the patient -- the follow-up of this patient as of February 26th. He had a normal physical examination, normal peripheral blood cell count, normal subset distribution, and normal repertoire.

So to conclude this long talk, just two or three slides on haploidentical hematopoietic stem cell transplantation versus gene therapy for SCID-X1 patients.

Two points to consider from my point of view. Kinetics of immunological reconstitution: twelve months to obtain a protective number of the CD3 positive cells versus three months for the gene therapy-treated patient.

This is extremely important, because in the case of a patient that arrive at the time of the diagnosis with a severe infection disease, this is the sort of run gain where we normally lose because of the viral infection in comparison with that.

So a faster immunological reconstitution for CDF-6 patients is a very important parameter to consider. The second one is the quality of the immunofunction, and always a long term follow-up can permit us to conclude about this.

I would like you to remember the European result of haploididentical hematopoietic stem cell transplantation in an identical condition in a SCID patient.

We have as each one in the world an extraordinary improvement in the long-term survival curve for patients with a B plus SCID form over the time, and this is the time of the first transplantation in Europe.

And we are now up to 18, and we can say 70 percent survival. All patients are confused, and so even the patients that go at the time of transplantation and the patients have a severe infection.

In any case, you can have a poor T-cell function in some case with a partial T-cell reconstitution and as reported by Dr. Blakely, we have in fact the same result, and a long-term decline in T-cell function, and low NK cell count, and infrequent T-cell immunity.

The quality of the immunofunction, I feel we can make a comparison between hematopoietic stem cell transportation and gene therapy. Or regards of the T-cell function, the time created over 10 years is a decline in this patient, and we don't know exactly why because we don't observe, so far, any decline in T-cell reconstitution for SCID treated patients, is probably leaked infection from one side, and incompatability, allogenic response against chemokapithalium thymopoiesis in the other cases.

This is the situation for gene therapy. The question is, are we able to transduce stem cells to guarantee a long term restoration of immunological function in this case.

And each case the restoration is the same in all the cases, and low for hematopoietic stem cell transportation, and low for gene therapy.

In terms of B cells reconstitution, it is infrequent in this case, and frequent, but, and as you know, we have a low figure.

This is the gamma-c transgene persistence in a patient with CD34 positive cells, at a different follow-up. For Patient 1, we have only one data that amounts to six after gene therapy.

For Patient 2, we have a two determination in the bone marrow, and positive at 5 months, and positive at 21 months after June 30.

Patient 4 is still positive in spite of the chemotherapy, and so I think we have some doubt on the fact that we have transduced some stem cells, and these doubts are now finished. He has sample cells detected at the same extent after 6, 13, and 39 months after gene therapy.

For Patient 5, we have just one detection one year after gene therapy and not evaluable at this time. We preferred to privilege the detection of the malignant clone by integration site.

P6, this is a patient that had an extensive varicella infection at the time of diagnosis. He was restored to complete immunologic function, but the very cells that destroyed the -- we know that what fights the infection are very toxic for bone marrow, because of cells, and at the time of the harvest of bone marrow, we can have very few CD34 positive cells, in comparison to the other patients who are free of the virus infection at the time at the time of diagnosis.

And in this case the kinetics of the immunologic reconstitution was very slow, and the patient has a very slow increase in T-cell number, and the CD34 detection up to one year after gene therapy is negative.

So in the debate on how many cells you must transfuse, these data can be extremely important.

Patient 9, we only have one determination 5 months after gene therapy, and the patient then is not evaluable because it is in Europe but not in France. And why there is an apparently increased role of insertional mutagenesis in a SCID patient, we can assess the gamma-c more than the interaction of the gamma-c with LMO-2, and that the role of every block in that T-cell differentiation pathway with the accumulation of immature CD34 cells ready to go.

There is a massive proliferation of a transducer because of cells tend to selective advantage, and probably we have an induction of gamma-c expression of function in pre-CLP cells.

In all of age, I think it could be very important in these cases with a distant pattern of hematopoiesis, and the accessability of active gene loci. So different -- gets a function of the disease and their agents. So phenotyping of gama-c negative cells between CD34 positive cells between 3 months and over 3 months is under investigation now in our laboratory.

And gene inspection pattern is ongoing on the inspection of an proto-oncogene as a function of the age, and integration sites as part of a falling gamma-c transfer, and we are coming back to the gamma-c negative murine mammal with new experiments in newborn mice.

This is the possible identification that everyone would like to know, and lastly, I would like to thank the large quantity of people that are studying in detail these two side effects, and the follow-up of the patients treated.

First of all, at the Necker Hospital, with Peter Cooper, Alain Fischer, Salima Hacein Bey, and Francoise Le Deist, who permitted us to sort the immunological reconstitution of these patients.

Out of Paris, we have a great collaboration of LMO-2 expression studies with Dr. Terry Rabit. And we have performed some experiments in transgenic mice carrying this transgene; and Karen Osbourne and Peter Fraser for the RNA-FISH analysis.

And Philippe Leboulch and Roberta Powlack in Boston for the study of eventory presence of replication of recombinant retrovirus, and RNA splicing of the LMO-2 protein.

And the VSV studies by Jeffrey Cohen in Bethesda, and Erika Avivi in France. And we have made, but have no time, to explain all the data and transcription for the phy-level gamma delta T-cell clone in relationship to the gamma-delta control T-cell made by Francois Cigaud and Delain Rigas in Paris.

And we make also the immunoscope analysis that doctor permitted to follow very tightly, the patient treated by Annika Lim in Paris.

And all the medical doctors that followed up the patients that were treated in Paris, the patients elsewhere in the world, especially at Alexander in Melbourne, Australia; and at Winterstein in Munich. And Ricardo Sorenson, that which the collaboration with whom is very fruitful.

Of course, I would like to thank the group of Christof von Kalle and Manfred Schmidt on the work on the integration site. And we are performing some studies on the cell cycle of B-cells in Paris with Papadapaluas involved in gene study and the reparation of genes. Thank you very much for your attention.

(Applause.)

CHAIRMAN SALOMON: It is obviously impossible to cut discussion off at this point and go on to a break, but what I would like to do is keep it focused for 15 minutes, and then go to a break at 10:00.

Otherwise, we will not get to the important part of the discussion this afternoon, and I feel responsible for that. So I would like to open this very informative presentation on the patients for discussion. I had one -- just to start one question. Is it Patient One then that had the -- is it the third patient with an LMO-2 integration site?

DR. CAVAZZANA-CALVO: Yes, I think it is Patient One in our service.

CHAIRMAN SALOMON: And that patient, if I remember right, is around 3 to 3.9 years?

DR. CAVAZZANA-CALVO: Four years.

CHAIRMAN SALOMON: Four years. Okay. John.

DR. COFFIN: A couple of things. I guess I just lost it as it went by, but could you quickly indicate the relevant integration sites of the vector in these three patients that have --

DR. CAVAZZANA-CALVO: For the -- I know in-- the question is can I indicate very precisely the integration site for the three patients, and this is the question?

DR. COFFIN: Yes.

DR. CAVAZZANA-CALVO: This first patient, and I speak under the control of Christof von Kalle, who performed the study, but for Patient 1, it is within the intron, and in spite of this has a normal LMO splicing.

DR. COFFIN: That is a common feature in MLVs. It's not that unusual.

DR. CAVAZZANA-CALVO: But we have it two sites.

DR. COFFIN: Yes, of course.

DR. CAVAZZANA-CALVO: And for the P-5 patient, it is extreme, about 5 kb before the first exon, and the third patient I know nothing in detail.

Dr. Von Kalle: Okay. Maybe I can briefly comment on this. As I have stated previously, we have no indication that there is any lymphoproliferation associated with LMO-2 integration in any other than the two patients that Marina described.

The data that we have presented was with regards to the chance of an LMO-2 integration into the vicinity of the LMO-2 locus and we have specifically searched for such integrations, and we have found a couple of instances.

One was located 40 kb upstream of the locus, and the other one was in reverse orientation, about 2 kb upstream of the start codon of the distal promoter.

And again we have no indication whatsoever that there is lymphoproliferation from either of these clones. Of course, we are looking closer with regards to patient safety in the other patients.

DR. CAVAZZANA-CALVO: But I have a comment. I think you can have one, only one integration site near to the dangerous site. But the much more important criteria is the clinical one; the physical examination of the number of peripheral blood cell count, and immunoscope analysis, and the other characteristics because if I remember correctly from the study of Don Kohn for ADA patients, there is only one integration site, and there is only one in one patient with an extremely large T-cell repertoire.

DR. COFFIN: Yes, I would completely agree with that point, but I think it is important I think to get on the table exactly what we are seeing in the total set right now, so that we can get the back of the envelope calculations if nothing else.

CHAIRMAN SALOMON: Crystal.

DR. MACKALL: Two clarification questions. First off, I thought there were 11 patients in the initial study.

DR. CAVAZZANA-CALVO: Which patient?

DR. MACKALL: I thought that there were 11 patients initially treated that were discussed last fall.

DR. CAVAZZANA-CALVO: I think on the 11th patient there was a compassionate treatment for a 15- year-old boy, because he had an affected biology form, and he had a little T-cell repertoire that permitted him to live in between a lot of virus, and parasitic infection, and bacteria infection.

And we described this and discovered that he had a gamma-c mutation very recently, two years ago or something like that. But the gene therapy treatment has been a failure because the child has been largely infected. So he has no restoration of T-cell function.

DR. MACKALL: Okay. So 9 out of the 11 patients that received the treatment had initial benefit at this point?

DR. CAVAZZANA-CALVO: I think I prefer to consider the newborn infants as a homogeneous group, and 9 of these were successful. And in one patient, who was 15 years old, was also a failure.

DR. MACKALL: Okay. And a second question. When you were showing the T-cell numbers on the two groups, separating out Patient 4 and 5 from the others, it seemed to me that those patients from 10 months or so on had about 10,000 circulating T-cells. Is it clear that they did not have super-physiologic levels of T-cells before the development?

DR. CAVAZZANA-CALVO: It is very difficult to answer, because as you probably know, a child less than one year has a physiological lymphocytosis. So after to 10,000 lymphocytes, an upper airway infection can determine lymphocytosis, and this is considered normal.

And even a small infection, a virus infection, upper airway infection can determine a lymphocytosis. And these only after one year that they showed the classical version of the cell hematological number, with an increase in the neutrophil count and a decrease in lymphocyte.

This is -- if you go see the identification tabs at 6 months late in the treatment, you have a 12 months plus 6, and you can consider that after or around 2 years, we can tolerate a number of lymphocytosis, mild lymphocytosis, but it is normal in comparison to the age.

DR. MACKALL: But they certainly had higher lymphocyte counts than anyone else in this study even early on looking back now?

DR. CAVAZZANA-CALVO: Well, looking back now, yes, you can say yes. But if you have a newborn with 8,000 lymphocytes, what do you do with it.

CHAIRMAN SALOMON: Rich.

DR. MULLIGAN: Marina, two questions. On the kinetics of the T-cell recovery, did I get that right; that those two patients, that they recovered their T-cells much more quickly?

DR. CAVAZZANA-CALVO: Yes. This is impressive.

DR. MULLIGAN: And is that different than a person given a transplant, a bone marrow transplant?

DR. CAVAZZANA-CALVO: Completely different.

DR. MULLIGAN: Okay.

DR. CAVAZZANA-CALVO: Never can we get two cervical -- with haploidentical bone marrow transplants.

DR. MULLIGAN: Okay. Secondly, the issue of -- is there a different target cell for infection? You mentioned potentially that there were CD34 plus cells, quote, ready to go.

Can you tell us anything you now know about the CD34 plus cells from those patients before infection, versus normal patients? Is there any sense that they are more replicating CD34 plus cells?

Is there any strange characteristics? Do you have a sense that after they are cultured under the conditions for infection that they give different counts?

And particularly LMO-2, I think, is supposed to be pressed in primitive hematopoietic stem cells. Is there any sense that these patients have more LMO-2 expression in their CD34 plus cells?

DR. CAVAZZANA-CALVO: Yes, the question is crucial and very interesting, but unluckily I have no clear answer to give you. Lastly, I can answer that in this patient the number of CD34 positives out of a viral infection is higher than other bone marrow cells.

And other bone marrow cells in the mononucleated fashion, and you have 3 percent of CD34 positive cells, and in newborns, we had upwards to 10 percent of CD34 positive cells. So this is the first great differences without, and I can't underline any biological differences between the two.

And the other one is the proliferative analysis, and for the scale-up protocol, we have set up an experiment in cord blood. And roughly for these two patients, the biological characteristics of the bone marrow is much more closer to cord blood than the bone marrow cells.

And we can also say that to the other patients that are oldest have a less proliferation during the transaction, but I am not sure that the difference is significantly relevant.

I have no other data to share with you about LMO-2.

CHAIRMAN SALOMON: I need some help with two technical questions. The first is, and it is just me being ignorant, but I don't understand this proliferation rate that you referred to. It had no units, and so there was a 7.7 and a 9.5 in two patients.

So, I apologize, but I just don't understand what that is.

DR. CAVAZZANA-CALVO: It is a stupid calculation, but you have 1 million cells and you have 10 million at the end, and I say that you have a 10-fold incremental of the number itself.

CHAIRMAN SALOMON: And so that is over what period of time in culture?

DR. CAVAZZANA-CALVO: The five days. I explained that we began on Monday morning to purify the cells and injected back the cells on Friday evening, Friday afternoon. And so during this period you can from 5, up to 10-fold expansion of the initial number of cells that are put in a culture.

CHAIRMAN SALOMON: Okay. So maybe we will come back to that.

DR. CAVAZZANA-CALVO: CD34 positive.

CHAIRMAN SALOMON: Yes. And the other question that I had would be if I -- and again if I got this right, these two patients got 44 to 40 million CD34 per kilo? I realize that they are small, and that these kids were infants, but that seems to me an extraordinary high CD34 dose.

DR. CAVAZZANA-CALVO: Yes, but per kilo. But we come from the experience in identical bone marrow transportation, and you must remember that during this period that we know nothing about gene cells. And all of the clinical trial so far realizing that the world was negative in terms of results.

And from the haplo condition we know that T-cell constitution is very slow. And that even when we injected a mega dose of CD34 positive cells, in some cases, although not in identical situations, you have a normal constitution.

And so we were very worried about efficacy than toxicity at this time. So this is the reason that in face of all of these proliferating cells that we injected back everything without knowing anything about toxic effect.

DR. COFFIN: Could I follow that up a little bit? So do these two patients then get more? I mean, how do these two patients compare to the others? Were they the ones that got the most cells?

DR. CAVAZZANA-CALVO: Yes, sir.

DR. COFFIN: And had the most rapid rates of cell reconstitution?

DR. CAVAZZANA-CALVO: Yes, sir. Correct.

DR. COFFIN: Did they also get -- could you compare the numbers of transduced cells?

DR. CAVAZZANA-CALVO: They are much higher.

DR. COFFIN: I'm sorry?

DR. CAVAZZANA-CALVO: The higher --

DR. COFFIN: These two were the two that had the highest?

DR. CAVAZZANA-CALVO: Yes.

DR. COFFIN: So quantitatively how many transduced cells were in these two patients compared to what in general was in the others?

DR. CAVAZZANA-CALVO: I can say that they have two times the number of the other patients roughly.

DR. COFFIN: So the average for the other patients --

DR. CAVAZZANA-CALVO: Except for P6, because P6 is a strange patient -- We know the toxicity of the various viruses on born marrow cells, and so we could not have very few CD34 positive cells at the time of treatment.

And this patient received around 1 million of transduced 34CD positive cells per kilogram. And so 10 times less than the other, and has a very slow T-cell reconstitution probably. And we are looking attentively for B cells because the level of immunoglobulin is going down, and it is not perfect.

And they have no CD34 positive set transduced detectable today.

DR. COFFIN: I'm sorry to keep hitting on this, but these numbers I think are going to become very important in our discussions later on, and so I wanted to make sure that we were very clear on them.

So you said before the average total number of transduced cells per patient on average is 17 million per patient? I took that number.

DR. CAVAZZANA-CALVO: No. There is 10 patients, and so we have 8 million and 20 million for P4 and P5, okay?

DR. COFFIN: Okay. Total transduced cells.

DR. CAVAZZANA-CALVO: Gamma-c transduced set cells. The total transduced cells are much more. I think it is about double, about 40.

DR. COFFIN: But successfully transduced cells.

DR. CAVAZZANA-CALVO: CD34.

CHAIRMAN SALOMON: John, I calculated about 70 percent transduction.

DR. CAVAZZANA-CALVO: No, it is not 70 percent. It is 40 percent.

DR. COFFIN: I don't want twice as many. I want to know what the numbers are. That's what I am trying to get at.

DR. CAVAZZANA-CALVO: But I think that your comment is right, because in reality they received a huge number also of gamma-c positive cells, CD34 negative. And we don't know if they are committed for the precursor cells that the other patients received less then.

DR. COFFIN: All right. So just for either patient, either Patient 4 or Patient 5, once more, can we get the numbers. So the total for that patient, the total number of gamma-c positive cells transduced in the whole population of transduced cells was what? Was it 20 million, or --

DR. CAVAZZANA-CALVO: Forty million.

DR. COFFIN: Forty million.

DR. CAVAZZANA-CALVO: The total number.

DR. COFFIN: The total number of gamma-c positive cells.

CHAIRMAN SALOMON: John, just for the interest of getting done here, can we break on that, and then during the break, which is going to come in a minute, can we make some calculations together.

DR. COFFIN: Yes.

CHAIRMAN SALOMON: And then present it to the group?

DR. COFFIN: That's fine.

CHAIRMAN SALOMON: So before the break, I would like -- Joanne, did you -- and Warren, and Kathy, and Ken, you get the last. And can we keep these kind of short and brief, because I don't want to completely lose the track here, but I do certainly appreciate how important this part of the discussion is.

DR. KURTZBERG: I have a couple of questions. One is did you change anything about the separation of the marrow cells to select the 34 cells between any of the patients?

DR. CAVAZZANA-CALVO: Nothing. Everything is reproducible and we have made the same conditions, exactly the same for all of the patients treated. And we obtained in term of a transaction of a CD34 positive cell exactly the same results.

DR. KURTZBERG: Did you look for EBV infection in any of these patients?

DR. CAVAZZANA-CALVO: Yes, negative.

DR. KURTZBERG: Okay. And what happened to Patient 8? I lost track of that.

DR. CAVAZZANA-CALVO: Patient 8?

DR. KURTZBERG: Yes.

DR. CAVAZZANA-CALVO: What is the problem that you have with Patient 8, because I don't remember the all the patients in this way.

DR. KURTZBERG: I just thought that I saw Patient 8 missing from a lot of the graphs, and maybe it was me. I can look after.

DR. CAVAZZANA-CALVO: It is narrow, but he is doing well. In the kinetics -- well, it is a patient that is not in France, and so it is impossible for us to solely monitor the kinetic reconstitution.

CHAIRMAN SALOMON: Kathy.

DR. CAVAZZANA-CALVO: And also for Patient 5, the approximate kinetic strength is not so precise as this patient does not live in France, and we cannot follow at the same time period of the patient. This is the reason. Kathy.

DR. HIGH: I wanted to ask you one question about gamma-c subunits. And so I think you showed one slide where you showed that the levels of expression, of protein expression, were essentially normal of gamma-c. Is that right?

DR. CAVAZZANA-CALVO: Well, even knowing that for the other patients, normal or low.

DR. HIGH: Normal or low for all patients.

DR. CAVAZZANA-CALVO: I decided that the best one is normal.

DR. HIGH: Okay. And is the gamma-c subunit usually limiting for assembly of the cytokine receptor, or is it the other subunit?

DR. CAVAZZANA-CALVO: I have not performed the study of this type, and I don't --

CHAIRMAN SALOMON: Warren, why don't you answer that? You have the next question.

DR. CAVAZZANA-CALVO: Well, just a second. In the EBV study, we quantified the number of gamma-c receptor on each cell, and it was 150 gamma-c receptors present.

And the normal EBV cell line has up to 200 gamma-c receptors present. So this is the only answer we can give you, and Dr. Leonard can give more of an explanation.

DR. LEONARD: Yes. My question had also related to the level of gamma-c expression, and whether perhaps Patients 4 and 5 had any more than the other patients.

But related to the level of gamma-c versus the other cytokine receptor chains, I don't think that anyone really adequately knows the answer to that, because no one has ever done rigorous scat charts comparing with mononuclear bodies comparing the level of each of the various components.

And one of the underlying questions is whether gamma-c is overall limiting or whether it is available in excess, and is there enough to go around for all of the different cytokine receptors simultaneously, or is it perhaps limiting some situations which would allow for competition?

And the answers are really not rigorously known for primary cells, and particularly if you are considering thymic precursor cells or early pre-genitor cells, and certainly no one has done any experiments on those sorts of cells that would rigorously address that sort of question.

DR. MULLIGAN: Can I just ask him something about this point, which is that there is a point where gamma-c isn't normally on in the very earliest cells I would assume, and so wouldn't this -- there definitely is a novel case here in principal, where the addition of the gamma-chain to a cell that might be making or using another beta-chain of these other cytokine receptors. Isn't that the case?

So you are introducing, you are having gamma-c at a point where it wouldn't normally be.

DR. LEONARD: All I can really say is that when we were making our own gamma-c knockout nest with Paul Love, and looked at early thymic precursor from wild type mice, they made a Northern blot with identifying thymic promordial cells at the earliest points they could look at, gamma-c was expressed everywhere.

So we really never found in the thymus an early enough place where it was not evident in the mouse.

DR. CAVAZZANA-CALVO: And just a piece of information. I don't know if everyone knows that a transgenic mouse does exist for gamma-c that is under a CD2 promoter. And that is a little bit later than ours, and the mice did not develop any leukemia. That is information just to complete your information.

CHAIRMAN SALOMON: Ken.

DR. CORNETTA: I actually have a question about the animal stuff that you just presented earlier. You said that you had in your animal model study out to 47 months.

DR. CAVAZZANA-CALVO: No, 47 weeks.

DR. CORNETTA: So, 47 weeks. Okay. 47 months, I wish. How many animals and what kind of analysis was done at the end to have confidence that there was not a leukemic process going on?

DR. CAVAZZANA-CALVO: How many animals we studied after the 47 weeks, I don't remember, but I can check when I come back home. And conversely I know what we studied extensively is the thymus, spleen, gastrointestinal tract, and the peripheral lymph node for each animal after the necropsy.

And we studied these because as you know in gamma-c knockout mice, there is -- the gastrointestinal lymphoid compartment is depleted. And the restoration of this compartment is a sign of good restoration. So it permitted us to see very small histological modification for the mice.

DR. COFFIN: Just very quickly. The real question there would be how many transduction events, how many integration events did you actually study in total in those mice as compared to what you see in these patients?

It would be interesting, and you probably don't have that information, but it would be interesting to dig that back out again and see how that compares.

DR. CAVAZZANA-CALVO: Yes, but there is a gap in the sense that we performed no study on the integration of profiles in the mice. I have no information. When we performed the study on the mice, we didn't think to look at the integration site. Sorry.

CHAIRMAN SALOMON: Okay. The last question before the break, and then we are going to have a break.

DR. TORBETT: I just would like to ask Dr. Leonard a question. He addressed the question of where gamma-c is expressed, but I would like to know how early is it expressed, and is it expressed in hematopoietic compartment? Getting back at Richard's question, because what you are doing is transducing 34 cells at or in supposedly a very primitive stage.

And what I would like to know, getting back at this question of inappropriate expression, at a stage that might be critical to send it on its way to an oncogenic event.

DR. LEONARD: We have not looked, and I am not sure that anyone has looked really early in bone marrow derived cells. So I can't really answer.

DR. MACKALL: I want to make a comment. I mean, early T-cell progenitors in the bone marrow are known to be IL-7 receptor positive. And so I presume that means that they are gamma-c now. Those are already somewhat committed to a lymphocyte lineage.

DR. CAVAZZANA-CALVO: I know that --

CHAIRMAN SALOMON: We will get back to this. This is kind of like holding back a hundred thoroughbred horses. I mean, there is just a lot of brainpower around the table, and I respect that.

And I wish that we just had hours and hours, but this is really where it ought to go. I just think from the whole committee and everyone here that I want to thank Marina for coming here.

I mean, you had an adverse event, and you really handled it well. That is me speaking personally. You presented your data, and you shared it with us in October, and you are sharing it with us now.

I hope that you hear directly from me, and I just have a lot of respect for how well you have handled it, to me it has been an example. To the whole world, how gene therapy ought to deal with its problems, as well as its successes. Thank you for being here.

DR. CAVAZZANA-CALVO: Thank you.

(Applause.)

CHAIRMAN SALOMON: So, 10 minutes, guys.

(Whereupon, at 10:08 a.m., the meeting was recessed and resumed at 10:29 a.m.)

CHAIRMAN SALOMON: Okay. I would like to get started again. What we have planned now are two 25 minute talks to further expand our database, and then we will go into a Q&A period, and then an open public hearing.

My recommendations would be to everyone to stay hydrated. We have been known to forego lunch, and I have sort of the sneaking suspicion that may be necessary today as well.

So, getting started again, I would like to introduce Dr. Claudio Bordignon from the Institute of Scientifico of Raffaele.

DR. BORDIGNON: Thank you. Thank you for this opportunity of sharing our data. I have really preferred a talk that is more oriented to analyze the overall data that we have available, rather than getting into the scientific details.

Of course, all the scientific details are here, and so if there is time, we can go back to the specifics during the questions. The scope of my presentation is really to review all the data that are in our hands, meaning my group and in collaboration with those that are directly working on some of our projects; retroviral vectors, hematopoietic stem cells, and peripheral blood lymphocytes.

And to share the results of the initial analysis of the integration process and to take a look at whether or not this is meaningful, and how much we need, and what kind of direction we should go.

I will be reviewing data from 21 independent groups, including pre-clinical studies and clinical data. Two main fields, clinical trials of gene therapy of ADA-deficient SCID. Our group, the collaborators in Brescia, Luigi Notarangelo and others; and the collaborators in Israel, Shimon Slavin and co-workers, and I have one slide from Fabio Candotti, who has been so nice to give me one of the long term follow-up of the lymphocyte ADA studies and making one of the points that I am trying to make at the end.

The second set of group are pre-clinical and clinical studies of the use of the suicide genes. This is an independent group and there is a large European consortium founded by the European Union, and that is allowed to study very similar parameters in different models and different animal models for clinical and patients.

The reason why I decided to include this data, if there is time, of course, is because it is probably the largest communitive experience in the use of retroviral vectors in hematopoietic stem cells and lymphocytes.

And in addition to the large European consortium, there were collaborators from the U.S. and from Israel, and from Japan.

The first is ADA-deficient SCID. Of course, it is a different disease than gamma-chain, but definitely we don't have the time to review all the differences.

It is sufficient to say that probably that the selective event that is described in this type of disorder is probably not as pressing, as strong as it is in the gamma-chain deficiency. Here the defect is a housekeeping enzyme adenosine deaminase.

Now, in order to have a reasonable evaluation of the risk benefits and this was one of the points raised at the beginning of the presentation today, and in the first discussion today, I would like very briefly to come back to the point by Marina discussed in relation to the gamma-chain study.

Here the available treatment are HLH matched transplant, and I will not discuss this, but I will come back to this point.

Initially, no identical transplantation, and the range in Europe has been updated fairly recently in the EBMT in 2002. And the results are really not encouraging. Twenty-three percent overall survivor from the long experience. Of course, this has improved over time, but in any single selected center studies, you don't get higher than 60 to 70 percent. And if this is as a result of transplantation with conditioning.

If you go to non-conditioning, without conditioning, then the largest series comes from Dr. Buckley in this country, and it is about 53 percent.

Of course, there is PEG-ADA available, but this I think is also somehow a misconception in the general perception, because PEG-ADA doesn't really work in every patient, and does not offer the same level of reconstitution.

And if you look at the last survey by Mike Herschfield, who presented ESID in 2002 -- 83 percent is the survival of this patient, and 73 percent if you include the patient with failing PEG-ADA, and then go on to receive the transplant.

So these are the reference points, and what we are going to review as quickly as possible is our ten, eleven years experience of gene therapy and deficient bodies recombining immune deficiency, and this was mainly down at our center in collaboration with the group in Brescia more recently, and with Shimon Slavin in Jerusalem.

What were the main milestones of this research? In 1990, we published the pre-clinical data for what was available at that time as animal models, and I will not discuss any of those data today.

In 1992, we started a combination of stem cells, meaning total bone marrow actually, and the peripheral blood lymphocytes transduction, and this was published back in 1995.

And in '99, we published the first discontinuation of PEG-ADA in one patient. And in 2002, the more recent study based on CD34, and so when I say stem cell here, it is selected CD34 in combination with non-myeloblative conditioning, and this was also reported last year.

And these are the 12 patients that are enrolled in the study, and with a different protocols -- the PBL. There is not much that I want to make in relation to this, with the exception of a couple of relevant points.

The patient who received the treatment of multiple injections, and we don't have the total number here, but as Fabio conducted for this patient, it is still a huge number of lymphocytes and a huge number of integration. This is relevant and we will come back in a minute to the analysis of integrations.

The patient discontinued PEG-ADA, and what happened in that sense went from about 10 percent, less than 10 percent, to nearly a hundred percent of the transduced lymphocytes, and the curve continued.

And together with this effect, the majority of the lymphocytes that were present were transduced in increased numbers to increase the level of ADA, and the ADA present in the transduced population and the reconstitution of immune functions.

If you look at this, this is the first comparison of the response, and if you take a vaccine in this patient, and you analyze the response, you see that it is definitely better than the HLA mismatched transplant and comparable to the HLA matched transplant.

And this the slide that I made and referenced several times. It is important for the total number of lymphocytes that you see down here, and for the people that are not close in the room, this is 10 to the 11th, and I am sure that Fabio will be happy to comment on the duration, the number of integration and the survival of such cells for 13 or 14 years now.

Of course, please consider Patient 1 is on the original slide, and so this is Patient 1 of the overall experience, and it is not Patient 1 in our study.

Conclusions that I can draw at this point, and I apologize because I only gave you a quick touch of what was studied, but all patients had a total wealth of normal growth and development.

Gene therapy with PBL stands for safe and efficacious and optimal, as compared to hematopoietic stem cell gene therapy. This is mostly due to the failure of the total population of the peripheral blood lymphocytes to toxify the whole system.

So although you have an immune constitution, you have a number of other elements that are not optimal, at least if you compare to the data that you will see with CD34 purified cells.

And you have the restoration of lymphocytes polyclonal repertoire, and so it is included in the repertoire, the protection of infection is pretty good.

A very large number of transduced lymphocytes have been injected and have survived for more than a decade, both in the NIH study and in our study. No apparent safety reasons are present to reduce or contain the number of transduced lymphocytes that are injected into patients.

I wanted to make this point because the discussion on those integration events and the total number of integration per cell and so on, came up into the discussion, and if you look at the number, apparently for lymphocytes there is no such concern.

The CD34 positive study is very, very similar to what you have seen from Marina, and the protocol was inspired by their study. There are two main differences. One is technical and it is related to how we do the transplant, and the cell dose is much lower, the cell dose that is injected.

And the patient -- well, it is about one log less, between 1 and 2 logs less, depending upon the patient. The patients are also conditioned with non-myeloablative conditioning with Busulfan. You will see that this is an ambitious definition, because actually one of the patients went into myeloablation.

This is the story of the last four patients of the 12 that I indicated in the summary slide. And here is present are the age of gene therapy, in months, and so you see the patient, with the exception of the first patient, is slightly older.

And this is the follow-up available as of today, or well, yesterday, or the day before; 29, 23, 9, and 4 months. So the full follow-up is not as long as the French study.

As you see, some of these patients had already failed after transplantation. This is the total number of cells infused total. And this is the dose per kilo, and so you can have your idea. This was particularly low and it is important to keep in mind that this is Patient 2 of the hematopoietic stem cell, and you will see that this shows in integration a number of other issues.

So this is probably, I would say definitely the maximum dose. The percent of transduction ranges around 13 and 40 percent, and the degree of myeloablation.

As I said, essentially all patients were maintained on the same dose, and at least we tried our best to maintain the same dose, but the level of myeloablation, depending upon the condition of the patient, was very different, and actually this patient received myeloablative treatment. This is the patient, and you see here that there is a profound difference.

The other patient actually recovered very quickly, and also this patient, I have not updated the slides, but none of them needed to be transfused for platelets, with the exception of these patient obviously.

I think we figured out what was the clinical reason for this, and it is not too relevant for today. How about the correction of the immune defect? Well, this has a level of a total account of lymphocytes, and the different lineages, and as Marina described, in about 3 months you see the reconstitution of lymphocytes.

Of course, all these studies, the hematopoietic stem cell studies are done without PEG-ADA, and so only gene therapy is responsible for this reconstitution.

This occured B-Cells and T-cells, NK-cells, the different sub-population and so on, and I will not show any of the other patients, because the story is essentially the same. Naive cells, thymic emigrants, and that is an indication of thymic function, and you see there is nothing before, and then you go into the range of the age-matched control.

The reconstitution is fairly convincing, and also the functionary constitution, this is pre-gene therapy, and this is after gene therapy, and in relation to the age-matched control, and that is the last --

And you see that this is true for RPHA candidates, and so on and so forth, and this is also for the correction of overall immunity, and you see here a different type, and the level of IgG here, of course, the patient was using IgG and so forth.

I refer to pre-clonal, and the complete correction that we get with the hematopoietic stem cell also relates to the metabolic defect. This is actually more relevant than if it was normal, and it was perceived before gene therapy, and actually although we never reached the level of normal individuals in the different cell lineage, this more than sufficient to detoxify the system.

Again, not through normal levels, but to levels that are very similar to the heterozygous parents, and therefore, influential for the health of patient.

Now, let's take a quick look to the engraftment, because this is probably in relation to the total number of cells and the integration event. It is probably relevant to have a sense of how the different cell doses are working for the patient.

This is Patient 1, and these are the days in the range for these numbers. And T-cells are a hundred percent, and K, and so on, and so forth. But also the other lineages are pretty good even at a relatively long period after the procedure.

This is the second patient, and do you remember the patient who got the 0.9 dose, the lowest dose, despite this low dose, selective pressure was sufficient to get nearly a hundred percent of transduced T-cells.

But in K-cells and B-cells, and all the other non-lymphoid lineages, are very, very low. This patient, without the serious adverse events reported probably was going to receive a second dose, but since he is clinically well and doing fine, we decided to postpone a second procedure.

Patient 3 is similar to Patient 1, and Patient 4 was actually receiving gene therapy during the banned suspension that is ongoing in Italy, with the specific permission to treat this patient on a single patient basis, is now in the phase of reconstitution, and the numbers for the time look pretty good.

This should not worry because we usually see the cells taking off after 3 months, and the patient is in Saudi Arabia, and so we cannot do the controls all the time.

But the level of lymphocytes are climbing, and all the other patients are predicted with the climbing of the level of lymphocytes to the proportion of new cells should also go up.

Conclusion. A restoration of lymphocyte counts and corrections of humoral and cellular response, and correction of the ADA and metabolic defect, and this is what makes together with the full immune reconstitution, and this was perceived better than the lymphocytes.

Multi-lineage stable engraftment of the new cells, no adverse effects. Clinical benefit. PEG-ADA. How good is the clinical benefit. How does this compare with allogenic bone marrow transportation. This data are due to the courtesy of Shimon Slavin, and collaborators Memet Aker, and Shoshana Morecki.

The have family that has obviously the same genetic background, a Palestinian family, in which you can compare in that sense a normal individual, and an ideal transplant, and gene therapy, because the family has all three cases.

So this is Patient 1, and our Patient 1 from gene therapy, and the second sibling is a healthy one, the blue one, and the middle is a patient four years after umbilical cord blood with a full HLA match.

And you will see that I am not going to argue that this is a bit better than this or this, but in essence, with the difference of K-cells, all the different parameters, the gene therapy is either the same or better than the transplant, and often very close to the normal individual.

And I am flipping very quickly for the second time through all the different parameters, including the specific cell levels. So let's try to conclude from this comparison and from the rest of the study.

All patients are doing clinically well with normal growth and development. Gene therapy with CD34 cells, combined with mild myeloblation, is safe and efficacious as a single therapy.

You have multilineage stable engraftment, and restoration of the lymphocyte count, correction of humoral and cellular defect, correction of ADA activity and metabolic defects, clearly superior to mismatched transplantation, comparable, or at least comparable to matched transplant.

These are all the people who participated in my group in San Raffaele in Milano, and there are a number of external collaborators without whom all the studies would not be possible.

These are the long term collaborators, with whom we started the vectorology, and the initial studies. Now, let's go and take a look at the preliminary data on the analysis of integration very, very quickly, because these are definitely not complete data.

I am sure that everybody here is familiar with the reverse PCR, and you can essentially cut out, and amplify, and sequence upstream and downstream the regions that are close, adjacent to the integration site.

And you get this type of pattern. In T-cell patients receiving hematopoietic stem cells, and this is Patient 1, and you remember the good cell dose.

This is the T-cells and all the different times, and here and there if you look carefully, you may find an event, and this is the level of progression of transduced cells.

Patient 2, you will remember the very low cell dose, and I apologize, as this is almost impossible to read, but this are from -- two from the same patient, and you see that there is some level of clonality.

Patient 3, again, is similar, and Patient 4, it has some level of clonality, but certainly after reconstitution as we saw in every patient. So you can clone and sequence in a number of different ways.

First of all, you can clone the T-cells, and then you will produce a number of clones, and then they are either single integration or at most a couple of integrations and that would be related to the fact that they are not really clone.

But it is a clone with two integrations, but you can cut out and sequence the whole thing. This is the study on integration in peripheral blood granulocytes, and so this is the myeloid compartment.

Here in all patients you see a limited number of clones, and that makes it easy because you can cut out any one of these and sequence them all. So for this patient, I will show you that they have the analysis on all the clones.

And these are again in peripheral blood, and in bone marrow in different lineages. And again you see fully cloned, oligoclonal integration. This is the patient who received -- and this is the only one I am going to show, the patient who received peripheral blood gene therapy.

And as you see, this is 25 months after the last infusion, and 90 percent of the cells are vector positive, and again you see a big smear with many, many integrational doses, and a few of those are predominant.

If you analyze this at different times, actually some of these switch, and so you don't find the same immediately. How can we do this, and how can we analyze integration. It is relatively simple.

You have two possibilities. If you have these discrete bands as you have seen in the granulacyte, then with the bands you do direct sequencing, and with an algorithm that was prepared by our bioinformatic people, you can go directly and analyze where it is on the genome, if the region is known obviously. But this is the case for the majority as you will see.

If you are not in this situation, you can make a mini-library clone directly in plasmid sequence, pick the positive one sequence and map the same way, and this can be done with rubber, if that's justified, of course.

So this is the preliminary analysis of 25 integration sites that for one reason or another were selected because of the clinical predominance, because they were appearing as more clonal in some patients, or in some phases of life in the clinical follow-up of the patient.

These are the number, the location of flanking regions and so on. So not containing genes are 11, and inside known are predicted genes 70 introns and 20 exons.

The draft genome region one, highly repetitive sequence, one; short non-conformative sequences. We are redoing the analysis of these four.

This is the analysis that I mentioned for the granulocyte. This is in essence the full feature of Patient 1 at a given stage.

These are the number of identical clones identified, and this is the length of the region, and this is the chromosome location, whether it is outside the gene, inside the gene.

And we are continuing with this analysis for all the other patients, and for the different lineages. What can we conclude after this? Not much really. Integration are highly pre-clonal in circulating T-cells of three ADA SCID patients, Patient 1, Patient 3, and Patient 4, and there is a single exception of Patient 2, but you remember Patient 2 really got very, very few cells.

Several different integrations are detected in granulocytes and other subsets, and cloning of integration site and differentation to different subpopulation is in progress. I will come back at the very end of this.

Now, in the last few minutes I will review the suicide gene therapy in hematopoietic stem cell transplantation, and I apologize for those of you who are not directly in the field of gene therapy, because at the speed I will be presenting this data and the clinical protocol will probably not mean too much.

But the purpose is to tell you that this is probably the largest study, both pre-clinical and clinical levels, and therefore, I will show you the data from the safety point of view.

Well, why was this done? For the simple reason that in allogenic bone marrow transplantation, T-cells that are present in the graft really determine the outcome of the transplant in many ways.

If you take the lymphocytes out, you get no graft-versus-host disease, no immune reaction from the graft to the patient. Unfortunately, when this was extensively utilized in the clinical setting, this increased the risk of relapse, demonstrating that actually the T-cells in the transplant to take care of the minimal residual leukemia disease, or lymphoma, or other tumor.

When you were doing the transplant, or we were going the transplant in a conventional way, the idea was to find the balance between graft-versus-host disease and relapse. This was done with a total cell dose, but also with immune-suppressive agents: cyclosporine, methotrexate, and others.

Now, it is so true that the lymphocytes are taking care of the leukemia that from the middle 90s actually, the lymphocytes were extensively used to treat relapse after transplantation in order to obtain maximum anti-leukemia activity, unfortunately this was associated with significant graft-versus-host disease, at least 11.

In essence what you can do and what we did is to insert the gene in here so that you can switch them off if the graft versus host disease runs out of control.

How do you put it in and how do you select the cells through a transgene, truncated form of the deltaLNGFR affinity receptor for the non-growth factor, and so that you have the untransduced cells, and the transduced cells with a mixed population, and these also contains the Swiss Ig, the thymidine kinase.

But you cannot inject this, because you will only be able to kill those cells, and this would still leave graft versus host disease. So thanks to the cell surface marker, you can fish out all the cells and see how the protocol works. The protocol is fairly simple I believe.

If the patient relapses, you bring the donor of the bone marrow back to the clinic, and you do a leukopheresis, and you get all the lymphocytes, and you transfuse those, the transduced cells, and the little flag on the subsurface marker with a monoclonal antibodies, and you select them all out, and then you can give it for transplantation.

If the patient goes into complete remission, great. If the patient goes into complete remission and develops graft-versus-host disease, you can kill out all those cells by grancyclovir. This is the technology that is fairly established, and was published in 1997 for the first time.

And there are a number of groups throughout the world who are working on this. Now, you don't need to read all the names and all the affiliations. Of course, the majority of these as I mentioned is a consortium founded by the European Union to demonstrate that this really actually works at the clinical level.

What is important is that there are some collaborators from outside the consortium, again Shimon Slavin, and Phil Greenberg in this country, and Steve Kornblau from MD Anderson, and a few others.

These groups have been utilizing these tools both for pre-clinical models and for marking studies, and for clinical studies altogether. Now, where are we in relation to numbers.

I will not show you how efficacious this is. This was not my point. The point was not to find the right risk balance in this study, but to show you the size of the study and what we can try to conclude out of this.

These are pre-clinical studies all done in mice, utilizing human cells, human-mouse chimera, so murine cells. And this is the number of mice treated, 140; in the human chimera, 206; a total of 368, and these are the total cell doses that have been reached.

This is the follow-up in weeks, with several secondary and tertiary transplant recipients. This is the total accumulated time. The outcome after or at the end of the study, all the animals were specifically controlled to check whether there was anything abnormal in the hematopoiesis, and all the animals appear to have a normal hematopoiesis.

So the cumulative data on over 300 animals transplanted with the hematopoietic stem cell, and transduced with delta-LNGFR-coding vector, revealed no adverse event, with normal engraftment, persistence of differentiation of transduced cells, secondary and tertiary successful transplants.

More than a hundred of these mice were maintained for more than 20 weeks, and more than 70 animals, including 16 recipients of secondary, were sacrificed over 28 weeks.

Risk of oncogenic transgenic transformation by transduction with this vector system is less than 1 in 10 to the 9th events is what we can say so far. What is happening in pre-clinical studies supporting the safety of marking of T-lymphocytes with deltaLNGFR.

So this summary was on hematopoietic and I hope I didn't confuse you. Hematopoietic, first, and lymphocytes after. These are all done with lymphocytes. This again is human into mouse, 20 animals, mouse in mouse, rat in rat, and three long term dogs.

This again are the total number of cells infused with the total cumulative dose, and this is the follow-up in weeks, with a significant range for the dog. Of course, these are very precious experiments, with a total cumulative follow-up.

Again, the same condition. All animals were analyzed for lymphopoiesis, and in this case it is written hematopoietic and it should be lymphoiesis and no abnormality was observed.

So no other toxic events related to the gene transfer procedure of the transgene expression or differences in the ability to induce donor chimerism associated with graft versus host disease were observed with this transduction system in T-cells in 356 mice, 200 rats, and 3 dogs.

So, let's try to again accumulate everything. Studies supporting the safety of gene marking with a delta-LNGFR, a total of 905 animals, 7,169 10 to the 6th for the number of transduced positive cells, and the accumulative follow-up in weeks is 12,000 or so weeks.

And the number of animals that were followed for over 20 weeks is 151; and for over 28 weeks, 74. In every single animal, whether an hematopoietic stem cell was transduced for marking, or lymphocytes were transduced for the graft versus leukemia approach, the hematopoiesis and leukopheresis was normal.

The analysis of 93 independent studies of human T-cells revealed no change in the expression of markers of lineage activation of additional proliferative capacity.

The analysis of 102 independent consecutive transduction of human T-cells with two different vectors, and a variation of the same construct, revealed that all the cells remained strictly dependent on IL-2 for growth and survival.

These are the clinical data. Twenty-three patients treated with HLA identical sibling, so this is the graft versus leukemia, these are lymphocytes only. Eight are identical family donors, and these are the total number of cells that have been injected, again you see here that we are in the 10th to the 9th range, 105 times 10 to the 9, and the dose here is lower obviously because the risk of graft-versus-host disease is much higher for this kind of transplant.

Survival of cells in patient months calculation. The longest survival of the cells, the peak frequency that you see is significant for the immediately post-transplantation period. And here are the cumulative numbers.

Again, in the clinical study, not a single adverse toxic event, acute and chronic related to the gene transfer procedure, or through transgene expression that was observed during this trial, which involved the infusion of more than a hundred-million cells generated by more than 15 independent in vitro transductions.

What do these data suggest? The data are available to suggest that PBL gene therapy does not carry an increased risk of tumor transformation.

Of course, these are done, these studies are done with a limited number of transgenes, and we were discussing the issue of gamma-chain before, but this could be true for any gene.

Here I am arguing on the issue of insertion on mutagenesis if you can call it this way. So in that regard, for the ADA gene, the TK gene, and the truncated form of the LNGFR apparently there is no reason today to say we should limit the total dose of lymphocytes.

The available data on hematopoietic stem cell gene therapy for ADA-deficient SCID do not show an expansion of dangerous clonal integration. We have monitored the change over time there is no expansion.

We are sequencing, and we have not found anything. The analysis will continue. The analysis will continue in my mind doesn't mean that this is really useful. I am just giving you my opinion. We are doing it.

And whether or not we will get out anything I don't know, but I think that some of the groups that have the possibility because of the size of the study of the patient should do it, but whether this should really be mandatory for each study at this stage for the data that we have in hand I am not really sure.

But of course I am in a different position related to the one of this committee. Instead of transgene-specific risk, the discussion that has already started, it is not my position to fuel the discussion also because as the Chairman already said, this is fueling it enough.

How extensive should be the monitoring of the integration sites, I don't know. We will definitely go on to analyze the different technology, probably a few thousand sites, and hopefully it will make for a nice paper, but if it will make for anything more, I don't know. Thank you very much.

(Applause.)

CHAIRMAN SALOMON: We will return to this for questions, Claudio, after Dr. Thrasher has given his presentation. So the second speaker is Dr. Adrian Thrasher, from the Center for Gene Therapy and Childhood Diseases in the U.K. Is it the Imperial Cancer Centre in London?

DR. THRASHER: I am actually an immunologist at the Greater Ormond Street Hospital in London. We have one of the biggest immune deficiency services in Europe.

Okay. So we have several Phase-I clinical gene therapy trials which were open and I am going to talk to you today a little bit about our X-SCID study. Marina actually introduced X-SCID and has given you most of the background excellently earlier, and so I can be relatively brief.

I will say that the other two studies that we have is for X-linked chronic granulomatous disease, and in this study we have treated one patient, and the ADA study is approved by our G-type, but we have not treated any patients yet in that study.

I just want to reinforce what Marina first said and Claudio said with the importance of reference points, because when we are trying to devise a new therapy, we need to compare that with the existing therapies.

Now, the European study has been extensively collected since 1968 and has been recently published in The Lancet as Claudio has said, and I just want to show you a couple of slides. The first one is the probability of survival after a mismatched transplant in SCID patients according to conditioning regime.

And many of you will know that the preference treatment of SCID in Europe and in most centers in the U.S., I believe, is to use conditioning in the mismatch setting. And since 1968 in the transplants the survival rate has been measured at about 50 percent.

Now, of course, we have gotten a lot better at transplanting these patients, and this is not accurate data for today's transplants. Now, we can also break down these transplants into subcategories of SCID, and we can definitely say that the outcome for the positive forms of SCID, such as JAK-3 deficiency or gamma-c SCID, is a lot better than, for example, ADA-SCID.

And currently as I said the survival rate across Europe and in our center for mismatched transplanting ADA SCID remains at about 30 center at one year. So it is very poor.

So this slide Marina showed earlier, and this indicates the improval in outcome for transplants in SCID and B-positive SCID for mismatched grafting.

And we counsel our patients now about an 80 percent one-year survival rate. Now, this does not take into account the patients who don't make it through the transplant center. We know that several of those or many of those occurring over time.

The other thing that we have to remember is that to achieve efficient or improve the B cell engraftment and T-cell engraftment that we certainly believe in Europe that you need to use some form of preferative chemotherapy.

And when you are considering risks and benefits, you have to build in the effects of the chemotherapy into the equation, and there are certainly long term effects related to chemotherapy employed in infancy, and the chemotherapy usually or has been until very recently been using alkylating agents, such a busulfan.

And there are clear effects documented on growth, fertility, possible developments of secondary malignancies later in life, and also neuropsychological effects, which may in-part at least be related to the use of chemotherapy.

And, of course, in many of these cases there is incomplete immunity reconstitution. And I would just show this, because I think this is a graphic example of how brutal using chemotherapy is in children, and so these are dental x-rays of a normal seven year old, and a seven year old treated in infancy for SCID by condition transplants.

And the normal seven year old is showing secondary teeth, and this child has no secondary teeth. Now, for a seven year old child, this is a big deal both very psychologically and physically, and we see dental abnormalities in most of our conditioned transplants.

So really these are the drivers to try and develop less toxic both in the short term, and in the long term treatments, and more effective treatments in terms of immuno-reconstitution.

Now, Marina set the case for this type of study and our gene therapy study for X-SCID is very, very similar to the study conducted in Paris. So what I will do is highlight some differences.

The entry criteria are the same and so there is no matched sibling donor. The vector we use is also an MFG based vector, and it doesn't have the BT mutation, and whether that is relevant or not I doubt.

There is a single gamma change, cDNA, and so this is virtually identical to the vector used in the Paris study. One significant difference perhaps is that all vectors are produced in PG13 cells, and so it has a different envelope. It has the gibbon ape leukemia virus envelope.

The transduction cycle is the same, and we started them on Monday morning and we finished on Friday evening when the cells are reinfused back into the patients.

The CD34 positive cells are purified by the CliniMACS system, so it is magnetic bead selection system. The cells are preactivated for 40 hours in the presence of high dose cytokines, stem cell factor, FLT-3 ligands, thrombopoietin, and a slightly lower dose of IL-3 than used in the Paris study.

And another difference is that we actually use serum-free conditions to culture our cells, as opposed to using some fetal calf serum. So there are three cycles of transduction over three days on Retronectin-coated surfaces. So this is in a closed system, and the back to the bag type system, which many of you would be familiar with.

And as I said, these cells are reinfused without any preconditioning for the patients. So the only procedure that the patient goes through is a bone marrow harvest, where we take probably around a hundred mls of bone marrow under general anesthetic for about an hour.

And so these are the details of the patients who we treated to date, and as I said, a follow-up is -- the maximum follow-up is 18 months for Patient 1, and they all present with classical features of SCID.

The age at therapy is 10 months, and 10 months, and 4 months, which is our youngest patients, and this patient was diagnosed early because of a previously affected family member.

And then the final two patients are slightly atypical, and this is a 20 year old patient with a failed graft, and this is a patient who has an unconditioned haploidentical transplant in infancy, and who has had waning immunological function, and consequently has developed end-organ damage in the form of bronchiectasis, liver disease, and gut disease.

And the fifth patient is a rather unusual patient, also from Denmark, who has classical SCID and expresses no gamma-chain, and presented with PCP at the age of 10 to 12 months, but survived at home with prophylactic antibiotics and immunoglobin cover until the age three years, and at the time of treatment he in fact was completely well and free from infections, and that is unusual.

So this is total cells infused times 10 to the 6th, and 180 million, 180 million, 78 million, 150 million, and 115 million. This is our transduction procedure, and you look at CD34 along here, and gamma-chain here, and we start off with virtually pure CD34 populations, and through the week we lose CD34 positivity on our cells, which is a side effect if you like of the culture conditions that we are using.

And so at the end of the culture we would say that approximately 60 percent of our cells are transduced for gamma-chain, and that 30 percent of the population retains CD34 positivity.

This is the data on the first three patients, and I will show you a bit more on Patient 1 because this is the longest follow-up. The first thing that we noticed, and the same as in the Paris study, is that we see many NK cells emerging very consistently between or after four weeks of infusion of cells.

And these rise and then tend to fall a little bit, and then stabilize at or just below the normal range for an age match control. At 10 to 12 weeks, we begin to see new lymphocytes appearing, and these gradually rise over time, and then stabilize, and at 18 months in Patient 1, we see normal numbers of lymphocytes, normal CD34 and CD8 ratios and slightly low numbers of NK cells.

Patient 3 is a Saudi patient, and again the follow-up is shorter on him, and he is in Saudi at the moment, and from the information that we have his immune reconstitution continues to be very good.

Patient 2 had an arrest of immune reconstitution of about 12 weeks. Now, this coincided with the time when he developed severe gastrointestinal bleeding, and we believe that this is a consequence of immune reconstitution. It is not a side effect of gene therapy, per se.

He then underwent radical surgery to remove part of his bowel, and developed nutritional problems afterwards, and so we believe that these factors in combination caused this delay or this plateau of immune reconstitution.

Now, around about here we managed to correct his nutritional problems, and we corrected the stoma that was produced at the time of surgery here, and so his gut was linked back up again.

And his immune reconstitution appears to be running back on track. So I think that is interesting, because it may be saying that at this point in time that we are reconstituting his T-cells from cells that have been engrafted in the bone marrow, rather than cells that are in the initial graft going into the patients, but that is speculation at this point.

And this is an interesting point. This is Patient 3, and at the time of T-cell emergence, he also developed this rash on his hands and feet, and in fact all over his body, and which really is indistinguishable from a mild cutaneous graft versus host disease. Now, this responded very well to topical steroids, and has not reoccurred subsequently.

Now, this is Patient 1, and this just shows you in a different way the emergence of naive T-cells, and so if we look at CD27 and CD45RO, and the 27 positive RO population are naive cells, and so you begin to see them at 10 weeks, and these gradually increase in number over time here.

And we know that if we take these cells out and we stimulate them with mitogens, or antibodies to CD03, that they will proliferate entirely normally.

We can also use the V-D-J recombination process to measure the diversity of the immune reconstitution as you see earlier, and so both by immunofluorescence studies we can show that the TCRVbeta families are represented entirely normally, and with a normal distribution.

And we can also measure by a spectral-type analysis by that diversity of CD03 links, and so again this is in Patient 1, and look at the total peripheral blood mononuclear cells between weeks 12 and 16, and you can see that this is week 12 and this is week 16, and you can see at the beginning that there is an oligoclonal pattern, and this is very typical, whether it be in the gene therapy setting or in the transplant setting.

The reconstitution is always oligoclonal to start with, and in fact some of these clones you can see from quite a distance after or for quite a time after engraftment, and they do persist.

But because there is a clone there doesn't mean that it is a pathological thing. So by week 16, you can see some normalization of these distributions, and then by week 24, when he has enough cells, you can separate the CD34s and CD8s, and see again a highly complex Gaussian distribution in those representative families. And again at week 52, the same.

So this is the follow-up on these patients, and the first patient was followed up at 18 months, 13 months, and 9 months, 6 months, and 2 months. The first three patients are alive and well, and at home, and this patient is off all therapy and off antibiotics and immunoglobulin.

The third patient is about to discontinue the immunoglobulin therapy and has normal levels of IgA and M. And Patient 3 at the moment remains on immunoglobulin therapy and prophylactic antibodies, but again is well and thriving.

Patient 4, which is the 20 year old patient, we can see no evidence for change in his immunological function or diversity by an immunoscope or spectral-type analysis, although we do see a positive signal for transgene CD34 positive cells.

And again Patient 5 is a three year old patient that was treated just before Christmas, and the follow-up is 2 months, and he has NK cells, and he also has small numbers, just detectable now of naive T cells, and he remains alive and well at home.

The copy number in our cells is very similar to that seen in Marina's study. We have seen approximately one transgene copy number per T-cell, and we also see markings of B cells between 1 and 5 percent, and in myeloid cells between .1 and 1 percent.

This is a LAM-PCR analysis, which we are all familiar with, and this was very kindly done for us by Christof von Kalle, and Manfred Schmidt, and again you see a very familiar pattern in these two patients of polyclonal smears in the CD3 populations in these two patients.

We were also interested in perhaps more bands, more polyclonality in the B cell lineage, and monocyte and granulocyte image than seen in the Paris study, and that may be a reflection of the different envelope that we have used for this study.

And we believe that also may be important for the robustness of the immune reconstitution, because if you can keep applying gamma-c transduced cells that contribute to the thymopoiesis, then the longevity of reconstitution may be more robust.

Okay. So the monitoring that we are undergoing or undertaking at the moment is conventional, in terms of spectral-type analysis. Obviously lymphocyte counts and mitogen responses.

Christof was very kind to do the integration profiles for us. We have performed RT-PCR for LMO-2, and have found in sorted T-cell populations, and have found low levels of LMO-2 transcription in some of the patients, but then we have also seen low levels of LMO-2 transcription in some of our normal controls.

So I am not sure that is a useful test to do, except that if you can clearly see hyperexpression of LMO-2, and then it may be useful. We are performing microarray analysis on all sorts of populations at the moment to look at global gene expression of cells, but I don't have any data on that.

And finally we have also looked for the presence of VL-30 sequences in our patients, and I can say that in one of the patients by PCR we can see the very low level in CD-3 cells. But it has been under textbook in 4 out of the 5 patients.

So as you know, our current trial status, which is this is a restriction imposed by our GTAC and our Medicines Control Agency, is that the X-SCID study at least is reviewed on a case by case basis. Okay. I will stop there and take any questions that are necessary. Thank you.

(Applause.)

CHAIRMAN SALOMON: Thank you very much, Adrian. If I can have Claudio join you, then what I would like to do is turn these two speakers to the questions.

To start, I wanted to make sure that I was really clear on how exactly similar, Adrian, your studies in the X-SCID children were to Marina's studies at the Hospital Necker.

So I guess the key question, and you didn't give us kilograms of the children, but I was trying to figure out that even if they were 10 kilos or more, they were getting more like 10 to -- somewhere around 10 million per kilo.

DR. THRASHER: Yes.

CHAIRMAN SALOMON: So, the number one question was can you give us some range of CD34 cells per kilo. The preconditioning, there was no preconditioning in your studies. And I apologize to Marina, because I don't remember whether there was any preconditioning in yours?

DR. CAVAZZANA-CALVO: No.

CHAIRMAN SALOMON: So that was similar.

DR. CAVAZZANA-CALVO: Yes, it was identical.

CHAIRMAN SALOMON: Identical. Anyway, if you can help me with any other --

DR. THRASHER: So the differences are our vector is serotyped with the gibbon ape leukemia virus envelope, okay? One difference. And the second difference, which may or may not be important, is that we don't use fetal calf serum in our culture process, and that may alter the distributions. So we don't know.

I mean, we have no data on that. And the expansion of total cells we see is probably a little bit less than what Marina sees. We see about a 4 to 5-fold expansion in total cell number, but we also see a big loss in CD34 positivity.

So our expansion in CD34 positive cells is probably only two-fold or three-fold maybe, or something like that. I can't actually give you the exact numbers as I don't happen to have them.

CHAIRMAN SALOMON: Yes. If you will just forgive the Chair's prerogative for a moment. I just wanted to track that down, and so the one thing is that when you -- you got a 60 percent transduction if I remember right, but about 10 percent of the cells were CD34 at the end of the --

DR. THRASHER: No, 30 percent.

CHAIRMAN SALOMON: Thirty percent. Okay. Marina, can you --

DR. CAVAZZANA-CALVO: Similar. With the total cells transduced.

CHAIRMAN SALOMON: Just so it is on the transcribed record, if you could come to the mike.

DR. CAVAZZANA-CALVO: In terms of the total number of transduced cells, I think Adrian is highest than me, because he mentioned 60 percent of the total cells transduced, and we never reached this level. It is around 40 percent for us.

And in terms of CD34, that positive cell for the gamma-chain is the same, around 30 percent from 25, up to 40 percent. This is the range limit we had.

CHAIRMAN SALOMON: So you gave 40 million -- 40 to 44 million in the two patients, and CD34 cells per kilo.

DR. CAVAZZANA-CALVO: No. For P4/P5, the question is related to the two severe adverse events, they received, at maximum, 20 million per kilogram of CD34 positive cells, gamma-c positive.

CHAIRMAN SALOMON: Okay. Right. About 50 percent of the total, they got. Okay. Now, that was CD34 --

DR. CAVAZZANA-CALVO: Positive.

CHAIRMAN SALOMON: -- positive. So that would be 30 percent of what you infused, right?

DR. THRASHER: Right. Well, before finding those, we would have to sit down and do it accurately.

DR. MULLIGAN: Can I make a suggestion? This is going to haunt us unless we -- are we always going to do per kilogram, total cells CD-34? I think the simplest thing is the total number of cells that you put in a patient.

DR. THRASHER: Yes, I agree.

DR. MULLIGAN: Okay. So what is the difference between that? Which is going to be comparable.

DR. THRASHER: Yes.

CHAIRMAN SALOMON: That's fine, and thank you for that. So how many --

DR. THRASHER: Well, that is the total number of cells that we are putting into the patient is between 80 and 200 million.

DR. MULLIGAN: And how many per kilo?

DR. THRASHER: No, no, that is total cells.

DR. CAVAZZANA-CALVO: Yes, if you can permit a comment. In terms of a bone marrow transplantation, the cells given to the cell must be expressed per kilogram and not in total number, and there is no sense, because all comparison with the hematopoietic stem transplantation is based on the kilogram basis.

DR. THRASHER: Yes, I would agree with that, but I think that in terms of measuring or trying to determine the number of integration events you are putting into the cell, I should change that slide to put in the total number of cells, okay?

The other thing that I think you may want to consider, I know you were talking about cord blood earlier, is that the CD34 positive cells you are putting into the patient here are different than the CD34 positive cells you get out from fresh cord blood.

So that way you can directly compare your numbers if it comes to that.

CHAIRMAN SALOMON: Right. Now, just one last question. So this is to Claudio. So when you did the ADA-SCID children, how many -- can you give us a sense of how many transduced hematopoietic stem cells you were putting in?

DR. BORDIGNON: It is about one log less, and the reason -- and also there is a significant range, from 0.9 to 20 or 15, and so even inside the study there is some difference. The reason for that is -- and in that regard ADA is different and would probably not be so informative.

So depending upon the level of toxicity that the patient gets there, you get or you don't get any expansion of the CD34. You start with the cell dose, and you finish with the same cell dose.

You don't get any expansion, and sometimes even the harvest of the CD34 is worse. One of the patients, the patient from 0.9 dose, was actually going through some myelotoxicity due to a previous viral infection.

So I think ADA is a very different story in this regard, and it would be difficult to compare, except for what is concerning the total dose of transduced cells, because that would give you an indication of the transduction events in vivo, and therefore, a comparison on whether or not that carrier is associated.

But this study is definitely at the low range compared to the X-linked studies.

CHAIRMAN SALOMON: Okay. John.

DR. COFFIN: Okay. For the sake of the back of the envelope calculations, which is by far the best we could possibly hope to achieve here, could we agree that the total number of transduced cells was somewhere -- and in the case of the X-linked SCID, was somewhere up to around 10 to the 8th per patient in total dose of integration events in these cells, and probably that level, and down to about 10-fold less, in that range.

Would that be sort of a fair range that we are thinking of? And that the two patients with the adverse outcomes were the two that were at the high end of that range. And I am not talking about per kilogram.

(Discussion off the record.)

DR. COFFIN: Well, they were still at the 10 to the 8th range.

DR. CAVAZZANA-CALVO: Yes.

CHAIRMAN SALOMON: Just as -- just a minute, John.

DR. BORDIGNON: I think --

CHAIRMAN SALOMON: One of the things that you have to do here is keep the comments on the mic, which is hard. I screwed up earlier on that, and so if you have a comment, I will do my best to recognize everybody, but you have to be on the mic, or otherwise the transcript will have holes in it and it won't work. I'm sorry.

DR. BORDIGNON: I think the three of us can sit down and give you a table and exact the numbers, per kilos, per total, and everything else, probably in 10 minutes.

DR. COFFIN: I think that would be great.

CHAIRMAN SALOMON: Then what I would like to do is to continue the discussions, and move towards when the first wave of the discussion is over, take a 10 minute break, and then let them do that, because I think that is really important to me, and I am getting some nods. Is that a good idea?

Because I am trying to square all these numbers and am having trouble with it, and so let's continue with the discussion that doesn't involve the numbers.

If there is any question that someone has about the numbers, they will hold it until we move up to a 10 minute break, okay? Mahendra and then Rich.

DR. RAO: Well, this is sort of a general question to anybody who can answer it, but it seemed to me that when everybody talked about having no adverse events, the time period was too short given what was reported.

So that if you had reported at that time on that study, you would have also said that there were no adverse events. Is that a fair statement?

DR. BORDIGNON: Definitely not for the ADA. I may have confused you there. The initial study, the first bone marrow studies, were done in 1992. So this is true. However, if you look --

DR. RAO: For the stem cells?

DR. BORDIGNON: Well, that was bone marrow stem cells transduction. It is true if you compare similar studies, and so if you compare the last study for CD34 purified, with the same cytokine combination. So the study that is comparative, and the ADA that is comparable to the gamma chain, then we are in that range.

With one relevant difference, I believe, that having looked at the analysis of clonal expansion, I think that at least the first two patients would have been already informative.

CHAIRMAN SALOMON: Let's do this. Richard, Barbara, Crystal, and Tom.

DR. MULLIGAN: Well, just on Mahendra's point, I want to echo that I think that this is key. I think when we go off and do our calculations that we should also do a calculation just on simply how many, what number of patients does Claudio have that are on the 3 year period.

Let's just say that is some magic period, and how many patients are in your collection of patients that have actually gone past 3 years and have had some demonstrable gene transfer.

CHAIRMAN SALOMON: We will put you on the numbers subcommittee.

DR. MULLIGAN: I am not finished, and so the second thing is comparison of Adrian's protocol with Marina's. During the in vitro pre-stimulation cocktail period, what was the total of A of growth of your cells in culture, versus -- and were they comparable factors? It looks like they were comparable factors, except you had calf serum I think, she didn't have calf serum.

Because I think that an issue -- if one hypothesis is that there is some abnormal target cell that is diseased specifically, then those in vitro culture conditions could be very important.

For instance, you may have differentiated out the cells that otherwise could be dangerous cells if they were preserved in the culture. So it would be very important to see if they are comparable.

DR. THRASHER: Yes, I would agree with that. We don't use fetal calf serum, fetal calf serum is not popular in the U.K. for various reasons. But I think that is probably the only significant difference in the culture conditions; that we are serum free.

And I agree with you. I think that made a big difference in the types of cells proliferating, or maybe different regions of the chromosomes that are susceptible to integration, but we have no data on that.

And we do plan to compare very closely the two transduction protocols to see if there are significant differences.

CHAIRMAN SALOMON: Barbara.

MS. BALLARD: Yes. My question is very -- actually kind of an extension of his, but you talked about a different vector envelope.

DR. BORDIGNON: I'm sorry, it may not even matter, but I didn't answer Rich's question. Did you ask how many patients were over the years --

CHAIRMAN SALOMON: Yes, but we will get to that in the 10 minute break. Sorry.

DR. BORDIGNON: Okay.

MS. BALLARD: Somewhat an extension of his, in that you mentioned a different vector envelope was used in your study from the Paris study, and I was curious if there is any way to know how much difference that made in how the insertions occurred?

DR. THRASHER: Well, Christof may want to comment on this, but just looking superficially at the LAM-PCR data, the number of integrations in CD3 are probably similar, but I have not looked at the comparative data. I mean, I think Christof should maybe comment on that.

DR. Von KALLE: The thing that we think is a little different

CHAIRMAN SALOMON: I'm sorry, but this is Christof von Kalle.

DR. VON KALLE: I'm sorry. The thing that may be a little different between the Paris and the London trial is that the amount of insertion sites we tend to find per given number of myeloid cells is higher in the trial from London.

I don't think that we can say anything firm about the number of insertion T-cells.

CHAIRMAN SALOMON: Let me just try to clarify. You kind of trailed off. It is higher in the French trial, or higher in the British trial?

DR. VON KALLE: In the British trial. The number of integrations as defined per nanogram DNA of myeloid cells seems to be higher in the British trial.

CHAIRMAN SALOMON: Is it still an average of one copy per cell?

DR. VON KALLE: Oh, yes. Oh, yes. Well, actually in myeloid cells it is a lot lower on average than one copy per cell.

DR. MULLIGAN: Can I just address her question technically? The GALV versus the ampho, shouldn't infect the types of integrations that occur, I think you might have been asking would it go into different locations, and there is no sense that would occur.

But in fact the kinds of cells that could be infected could be different. And so there was a weird target cell population and it might be more infectable by ampho than by GALV, or less infectable?

CHAIRMAN SALOMON: And remember that by the premise that we have on the table here, which we will get to later, if integration is not completely random, or if integration is affected by which genes' chromatins are open at a given time, then the possibility that different packaging envelopes would target different populations of cells could also affect the spectrum of the genes that might see the integration event.

And so it is not totally -- it is an interesting question, but I don't have any answers for you. Crystal, and then Tom.

DR. MACKALL: The first question is for Dr. Bordignon. It would seem that the combination of LMO-2 integration or around there, and gamma-c is a bad combo. I mean, a very simplistic interpretation. And you have listed for us thousands of integration events that have been accomplished in mouse and human models without adverse events.

But do you know if any of those were in or around LMO-2?

DR. BORDIGNON: Yes. That is a very good point. We have not found any so far, and we are trying to do -- to have a quicker way to answer specifically only that question, rather than analyzing zillions, how many go to the LMO-2, and I don't have an answer at this stage.

My impression is that from what we have analyzed so far, is that it is not a hot spot. It is not something that happens because it is more susceptible in the same transaction conditions, because essentially the transaction conditions are the same.

So from here down it is all speculation, of course. You can imagine that it occurs at exactly the same frequency, but it does not pop up, also that the inverse PCR system is also related to the efficiency.

You know, if you get the right amplfication size and the right primers and so on, they will come out more frequently. I am convinced that you will be able to give a definitive answer to this in a matter of probably a few months, but at this stage it is just my impression.

DR. MACKALL: And I think it is critical with what we are going to be charged with now to try to decide, which is whether the insertion of another gene around LMO-2 is going to be just as dangerous, and until we know whether there is any data to that effect, it is going to be very hard to know how critical the gamm-c component is here.

The second point, just in terms of distinguishing the British from the French studies, I still an struck by the post-gene therapy data that looks different.

In other words, the two patients that got leukemia had this dramatic immune reconstitution that was at an increased rate, and so that within a month and in an increased amount.

I mean, they were up to 10,000 T-cells, and I didn't see any of your patients doing that. Is that true?

DR. THRASHER: Yes, that's true. I mean, those patients are exceptional, and in Marina's study as well, and they stand out as the patients who reconstitute the quickest.

DR. MACKALL: It would be certainly nice if we had an earlier marker for those patients that were going to develop, rather than having to wait 3 years for them, and if you go back and look, in one year they certainly looked different.

DR. THRASHER: Yes, in hindsight, I think that is true, but as Marina pointed out, those parameters remain within the normal range for age-match controls.

DR. MACKALL: We are only setting the normals now and they are standing out now.

DR. THRASHER: Yes.

DR. MURRAY: This is for Adrian. I noticed that you had quite a broad range in age of treatment of your patients, and the youngest was 4 months to 2 years.

And I want to focus on in particular the two youngest patients, at four and 10 months, I believe it was.

DR. THRASHER: Two at 10 months, and one at 4 months.

DR. MURRAY: Okay. There is also the two patients in the French trial were the two youngest patients also who have the immune response. I noted or remembered you saying that Patient 3, who is 4 months old at treatment.

But what I don't remember was which patient did have -- I thought one patient in one of your graphs did have an earlier immune reconstitution than the others.

DR. THRASHER: Actually, the scales are different in total number of lymphocytes, the actual initial reconstitution rates are virtually identical in all of them.

DR. MURRAY: They are, and so you didn't have one with a more significant response before the 3 month period?

DR. THRASHER: No.

DR. MURRAY: Okay. And which patient was it that had -- I can't remember the patient number of the one who needed the bowel surgery.

DR. THRASHER: Two.

DR. MURRAY: That was two, and so how long was that patient in treatment?

DR. THRASHER: Ten months.

DR. MURRAY: Thank you.

CHAIRMAN SALOMON: Phil.

DR. NOGUCHI: It is to follow up on some of those questions, and Patient Number 4, Adrian, was that the patient that appeared to be a very mild GVHD?

DR. THRASHER: That was Patient 3.

DR. NOGUCHI: I'm sorry, Patient 3, and that was the youngest patient; is that correct?

DR. THRASHER: Yes.

DR. NOGUCHI: And do you think that might be any indication that perhaps the cells that are being transduced might be somewhat different than if they are transduced at an older age.

DR. THRASHER: It is impossible to say. I mean, the numbers are so small. GVHD type phenomenon in autologous transplants is well recognized. What the stimulating antigens are we don't know. It could be maternal cells, and it could be infection on board at the time of the graft. We don't know. We believe that it may be a cytokine-driven, CD4-mediated event.

DR. NOGUCHI: Just a couple of other trivial questions. Do you have any information yet on this Patient 3, in terms of integrants? I think you presented, I think, Patients 2 and 5.

DR. THRASHER: They actually all look pretty similar in terms of the distribution of integration.

DR. NOGUCHI: Okay. And was Patient 3 or which patient -- how old was the patient that had this very preliminary evidence of a VL-30 sequence?

DR. THRASHER: How old was he -- well, 10 months at the time of treatment.

DR. NOGUCHI: No, you mentioned one of the five had evidence of PCR positive ?-

DR. THRASHER: Yes, VL-30 sequences. We can detect that in Patient 2 a year after treatment, and at a very low level, you know, between 1 and a thousand, and 1 and a hundred T-Cells, compared to one transgene of T-cells.

DR. NOGUCHI: Thank you.

CHAIRMAN SALOMON: Kathy.

DR. HIGH: I know that Marina is looking at this question, but just to the group, are there any data that suggest that CD34 positive cells from infants less than 3 months of age are somehow different from those of older ones?

CHAIRMAN SALOMON: Marina, you might want -- I think she is running away.

DR. CAVAZZANA-CALVO: I think nobody knows, because for a practical reason none of these can harvest the bone marrow sensor from children without any diseases. So the studies were conducted on cord blood, but we don't know when there is a switch of proliferation capacity between cord blood and newborn.

The cord blood proliferates up to 10-fold when you make the same protocol in the transaction rate and is much higher in bone marrow, and the 3 months patient who had very similar to cord blood, and 11 months, and the 8 month patient, are much more similar to bone marrow.

But the figures are so low that you can't make a statistical test.

CHAIRMAN SALOMON: And sort of following on that theme, Claudio, you answered me at one point that the expansion potential, or the expansion seen in the early activation of your stem cells was one or two using Marina's view.

And your statement to me was, oh, because ADA-SCIDs are very different. So can we go back to that? A), are you saying that you used the exact same protocol; and, B), are they really different even if you match for age?

In other words, is a 4 month old ADA-SCID really acting differently than a 4 month old gamma-SCID in terms of their proliferation in these ex vivo things?

DR. BORDIGNON: What I was trying to say that they are less homogeneous in that regard, and therefore it is much more difficult to extract the same type of information.

And they tend to be from a metabolic point of view much more prone to toxicity. Now, the study will probably be different if you would do the same study in PEG-ADA patients when they are detoxified.

But the plans and protocol is designed for patients who do not have access to PEG-ADA, and that was what was requested initially, and that is what we are doing.

So under those conditions, you have patients that have significant metabolic toxicity, and therefore they proliferate relatively little, they harvest poorly, you have a low number of cells and low growth.

On the other hand, you have patients that respond absolutely normally, and they grow 5 to 10 times during that transduction phase as the patient described by Marina. Maybe a comment from Don Kohn on this.

CHAIRMAN SALOMON: All right. Don.

DR. KOHN: Don Kohn, Children's Hospital, Los Angeles. We are doing a trial for ADA gene therapy with Fabio Candotti, Cynthia Dunbar, and myself, and we have harvested marrow from four of those patients on PEG-ADA; a 4 year old, a 5 year old, a 15 year old, and a 20 year old.

It mirrors all the coments that have been made. We get much less cells as they get older, and they tend to transduce less, and so even by four years old, you are over the hill to some extent. And I think it is a big factor that these very young patients just have a lot of cells that transduce very efficiently.

And the disease setting may also be different in ADA, and we have also done children with HIV, and get again lower numbers, like 1 to 2 million per kilo, and not 20 or 30, or 40 per kilo.

CHAIRMAN SALOMON: So I think an interesting thing here is -- and an interesting theme here that we could return to later is that assuming that age is this obviously very powerful biological determinant, there seems also to be the influence of the diseases, and I thought that Claudio's comments are well taken about the toxicity that occurs from the primary gene defect in the ADA kids.

So that may be another factor, that if you take that away, you may level the playing field there to be more of the age. The activation protocols and the level of activation are felt at this point to be necessary to get good transduction efficiency.

But in the context of all of the things that we have begun to talk about, one wonders when we get back to safety whether we should carefully consider how much activation is necessary in a given protocol to achieve a reasonable amount of vector integration.

And certainly one could then go from there to argue for less is better if you can get a sufficient amount of transduction. So, Claudio, lining that up is, what were your transduction efficiencies despite the fact that your proliferative activities were significantly less?

DR. BORDIGNON: Between 10 and 40.

CHAIRMAN SALOMON: Joanne.

DR. KURTZBERG: I think it is also important to notice that the ADA kids are different because if they are detoxified early, they can have their own host cells proliferate without transfection. So their immune regulation may really be different of the cells that are transfected.

And so I think they should be evaluated as a completely different patient subset.

CHAIRMAN SALOMON: I didn't mean to imply that these diseases should be collapsed in any way. I was just trying to talk about the biology of the stem cells. John.

DR. ALLAN: This is probably a naive question, but in your Patient Number 2, Adrian, that had gastrointestinal problems, it made me start to think that is it possible that you could actually either constituitively express genes that are not directed towards uncontrolled growth, but could actually home to specific tissues.

And so you get -- you may get some sort of diseases arising because of the nature of the cells that have been transduced. So did you do a histopathology on the intestines?

DR. THRASHER: Well, actually, the gastrointestinal problems also pre-dated the gene therapy. So you had some GI hemorrhages from macroscopically inflamed lesions prior to the gene therapy.

The histology was pretty unexciting to be honest, and we never found histological microbiologic reasons for those lesions. But the same lesions bled after gene therapy.

DR. ALLAN: I just sort of wondered if you may get other types of diseases besides cancers that you might want to think about. I don't know if that is even a real possibility, but I would like to throw that around at some point.

CHAIRMAN SALOMON: Ken.

DR. CORNETTA: I am interested to hear what these numbers will be around CD34, but I just sort of give in prospective that we are talking probably an order of magnitude more CD34 in these very young children than we would normally think about per kilogram in adult transplantation.

But when you then factor in that often adults may be two orders of magnitude bigger than these children, if we are looking at actual integration sites, if the Committee is to try to find a number that sort of sets the bar, I think that is going to be very hard to do.

And I think we are ignoring a lot of the other factors that may be here. So I just sort of would have the Committee think about as we do this that this may be a bar that we really are never going to be able to try to set.

CHAIRMAN SALOMON: Mahendra, and then Dave.

DR. RAO: One thing that I wanted to do was just try and get a general comment either from the Committee or one of the speakers about this full integration site issue. I mean, you made the point that perhaps it is gamma-c, along with LMO-2, which might be the important thing.

Maybe it is not gamma-c at all, because when you look at the over-expressing mice, you don't see that kind of procreative response at all, and so it is not the gamma-c.

And when you looked at the integration sites in the inverse PCR experiments, I thought that the number that you saw within genes, you know within the intronic sound boundaries was significant, as a number from the total number that you analyzed, a third, and was that not surprising, or is that not a cause for thinking that you can interrupt genes or do something?

DR. BORDIGNON: Well, I think that will have to be a bit better defined than what you mean around genes. We have been choosing some sort of distance, but whether or not this is meaningful in the sense that the distances are appropriate, and really results in amplification of the gene that is upstream or downstream.

It is an additional part of the work that will need to be done.

DR. RAO: If you look at your table to ask if they were in the intron or exon.

DR. BORDIGNON: No, no, sure. Despite the consideration that it is not really the siting itself there that would matter, but whether or not it is affecting the expression of the gene, and where the gene is and so on.

There is apparently, but I think it is too early to say from this data, but there are data on the HIV integration that a famous science paper by Bushman and co-workers that suggests that there is a skewing towards a transcription active sites for integration, by a factor of two, which is not much.

And which for this type of concern for this type of study will probably not matter at all. I think that I will have to wait for a bigger number.

CHAIRMAN SALOMON: David, and then John.

DR. HARLAN: When the numbers committee meets, I would ask that they put together for us not only total cell number and CD34 positive cell numbers, because the point that I wish to make is that CD34 is not a clear binary definition as Adrian, I think, you showed.

You start off with CD34 bright, and then they get dimmer, and dimmer, and dimmer, and it gets to be an increasingly fuzzy number.

CHAIRMAN SALOMON: John.

DR. COFFIN: I wanted to address two things that came up before, which I probably forget the second part by the time that I get to it. The first thing is that there was discussion about setting a bar for numbers and so on. I don't think we will be able to do that.

But I think it should be in our purview to try to do some back of the envelope calculations to see, sort of, "what if things." And I think at some time in the future somebody is going to have to sit down and try to do the best job they can with that in a format design exclusively for that purpose, which this is not appropriate, and so we might be thinking about that as we go forward.

The second point regards both the distribution of integration sites and their relationship with integration sites to turn on a gene expression. Our analyses have found that integration sites are very, very widely distributed in cells as far as regions are concerned, with very, very strong, and very local preferences.

The Rick Bushman paper that was referred to uses a somewhat different method, but it came to the conclusion that there was a skewing towards regions that had been identified in genes and in HIV integration in human cells.

But the scaling was a factor of two, and about a third of the genome by their definition was genes, and those contained about 60 percent of the integration event.

So that number is small relative to the other levels of uncertainty that we are discussing, and that is actually sort of a small error, or a small correction, if that is the number that holds up.

And finally the experience with MLV systems in particular, and ALV systems, where lots and lots of integration sites activating genes have been looked at, has revealed that the rules for whether an integration site can activate a gene are probably extremely complex, and case-by-case, because in some cases their integration is more than a hundred KB away from a gene that are known to activate it.

But that doesn't mean that the target for that gene is a hundred KB. It may well be that anything closer to a hundred KB won't work. We just don't know. It all depends on how the chromosomes fall and everything else.

So we really can't know that. We can only make real guesses as to what in any given case might be the sensitive target region. I usually use the figure of somewhere between 1 and 10 KB for the sake of the back of the envelope calculations.

But it is certainly at least 1 KB I would say on average per gene would be the sort of region that you could imagine the targeting to occur in.

DR. BORDIGNON: Can I ask a question?

CHAIRMAN SALOMON: Right now? Hang on a bit. What I would like to do is -- Carolyn, you had a question, and then maybe Don, and then I would like to go to the break, where we can get some of these numbers, and then come back if that is okay. And, Claudio, you will be the last question.

DR. WILSON: This is on a completely different topic, but Dr. Bordignon, you summarized very quickly for us a vast amount of preclinical data regarding the use of the LNGFR marker gene, and made a case for the safety of that gene.

And I wanted to just clarify for my own purposes in contrast to the data from Christopher Bounds' group. All of those studies were using transduced lymphocytes, as opposed to hematopoietic stem cells? Is that the difference?

DR. BORDIGNON: No, no, I presented -- I went very fast I realize, and I probably also made some confusion myself. But there are two sets of data. The first half was all related to hematopoietic stem cells. The second half was a lymphocyte study. So the numbers are fairly equal in the two studies.

DR. WILSON: Thank you.

DR. CORNETTA: Again, maybe something else that people can consider as people are working through the numbers. I think this insertional mutagenesis, we sort of fixate on, that these vectors will just go in and knock out genes.

But I think something that is very telling here is that in all of these patients there has only been two occurrences, and it is both in the same locus.

So I think again focusing rather from how many CD34 cells are at risk or whatever, and also taking into big consideration that presumably some interaction, whether it is the LPR enhancer or something else that is really playing a major role here, I think that needs to be considered in these.

And the second point is I think that not only do we need to think about how many CD-34 cells have been transduced, but there have been many other T-cell studies that we have not talked about, and these are mature T-cells that have been transduced and had many more integrations, if you tried to put in how many have occurred over these trials that have not seen this event.

So again I think that this is a pretty complex issue and just to make sure that we are thinking about all of those.

CHAIRMAN SALOMON: The last question.

DR. BORDIGNON: Yes. There was a very interesting piece of information in what Christof von Kalle and Marina said about the existence of another couple of -- I probably understood another couple of LMO-02 related integration observed in the analysis.

Do you know anything about whether or not this is affecting the expression of the LMO-2 in those circumstances? Because probably if there was an increased expression or not an increased expression, and will analysis in the other one. I hope that I didn't ask the wrong question.

DR. VON KALLE: No, I think you didn't ask the wrong question at all. One of the clones that we looked at that have occurred in the patients earlier, and we have tried to track it, and that was the one that was about 40 KB away, and apparently it has not led to any lympho-proliferation.

The other one that we are looking at that is closer, we also don't have any evidence yet that is proliferating, but we have not done as much tracking on this clone.