Confirmation Number:370957 Event Started: 6/15/2005

Good morning, everybody, I have the privilege of welcoming everyone to this 100ct meeting of the Recombinant Advisory Committee. Hundred meetings. Any rate, hopefully this one will be as prodic tb as those that have gone in the past. Again, I welcome everyone to participate actively including those here from the public in our deliberations and discussions we will begin the minutes of the from March 16th, 2005.

Hello, the read the minutes from the last RAC meeting and there were issues that were clarified and I think they are fine now.

Dr. Barkley, any comments?

I read them carefully as well and again it -- I was impressed with nare quality and clarity and it's education to read them again. I adopted

Always an education to reread what we discussed. Any other points from other members of the RAC? Would someone like to move them to adopt the minutes. I move to adopt the minutes.

And I second it. All have to go through -- so our first vote.

Dr.

-- pours? --

I'm just looking for where I am. -- thank you all very much we will move forward to our next discussion while waiting for Dr. -- he is here? Dr. Kington, welcome and we were glad you can join us this morning. We have a place for you here at the front table. Dr. Kington is the deputy director of the national institutes of health and is here to present the certificate for appreciation and say a few words.

Thank you. Pleasure to be here. As deputy I get to do a lot of things. of the things I get to do a lot of is saying a few words and that can be on my epithet, he said a few words. Really pleasure to represent doctors and the Singhtary this morning. -- secretary this morning. This committee actually is an extraordinarily important committee and we wanted to take this special occasion to thank those of you who served this committee as sort of advisory role that actually has extraordinary impact. Actually, that -- don't have an impact but you definitely have an impact. This was a critically important advisory role and played different parts of providing perspectives to help the agency and the department deal with the significant safety dimensions and the clinical applications related to recomeinant DNA research. I learned by the American association of association of advancement of science for this committee for its main correct.

The freedom and responsibility award which is remarkable given it was selected for individuals and to give this a committee in all of you who served on committees I think is an indication of how important your role has been. I hear a lot of good things from doctor Patterson about the various advice that you give and it's a pleasure to get -- for those who will be -- I can assure you that your impetus has been valued and we will still continue to call upon you. You are not out of the loop yet. We have a remarkable tendency to call people over and over and over again to in our committee. We have over 20,000 committee assigned, scientists from all parts of the scientific community who advise us on various commities and advisories committee and study sections and it's a strength of agency and the department to allow -- and clinical perspectives and patient's perspective and advocacy perspectives to do the right thing and this is one part of our recognition of those services. You have to attorney give me if I mispronounce your names p. There are no nonics here. Dr. DeLuca. Dr. DeMint. Thank you. This is going to give me trouble.

Doctor Kington again thank you very much for taking the time to join us this morning.

Pleasure our next agenda item is to re-- our next agenda item is to have Dr. Patterson discuss the review with us the rules for conflict of interest.

Thank you. As you know this is a standing part of our agenda but to read into the record our rules for conduct -- conflict of interest review. Be a member of this committee makes you a special government employee and be there subject to rules of conduct that apply to government employees. The rules and regulations are explained in the report titled standards of ethical conduct for employees of the executive branch. You may remember each of you received a copy of this document when you a were appointed to committee. In addition to following the rules, we like to review the steps that we take and ask you to take to ensure any conflicts of interest are both identified and addressed. As you know, the before meeting you provide us with information about your personal, professional and financial interest and we use this information for basing assessing if you have real, potential or apparent conflicts of interest that could compromise your ability to being objective to give advice during committee meetings. While we waive conflicts of interest for general matters because we weave your ability to be objective that won't affect your interest in such matters we rely to a great degree on you to be attentive in our meetings to the possibility that an issue will arise that could affect your interest in a specific way. So just want to underscore that point that during the discussion of both issues as well as specific protocols it's important to be is continually mindful of the dynamic nature of the review and that issues might come up that you weren't aware of during the initial conflict of the screening and be mindful to alert us to any conflict even in the course of the meeting. If this happens we ask you to recuse yourself from the discussion and leave the room and not to leave the table or sit in the audience. And you are required to recuse yourself from the preliminary protocol review process when you have a real or apparent conflict of interest. Involving a particular protocol. If you have any questions about the rules of conduct or conflict of interest, our committee management officers will be happy to address them. Thank you.

The next agenda item is to rediscuss the RAC recommendations for conduct of ongoing and future gene transfer studies in X link severe combined -- deficiency. We began this discussion shortly after the initial two children in the French X-SCID study developed a proliferative disorder subsequently found to be a form of acute limb foesetic leukemia. Does anyone know a third infant, child was diagnosed with that same disorder post gene transfer early this winter. And one of the initial children died with their leukemia. The other back drop to this discussion is that 11 children received gene transfer in Paris. Ten had some of the 11 had some immediate, meaning within months, and continued evidence of increased well being and reconstitution of their underlying SCIDs. So we and the rest of the scientific community are left a dilemma which not a disorder X-SCID which is uniformly fatal if not treated. When gene transfer is found to be relatively open conscious as it was and is in this cohort of well defined children, how much risk should one assume in the conduct of these studies? We met as a group several times to examine the scientific information, which has emerged and has been very on a personal note, very carefully done since the initial reports of leukemia in this cohort. And I believe all of us would agree that the science which has been carry out by multiple investigators really rests on the careful accumulation of samples from this cohort of children, both before their gene transfer and subsequently. And if it weren't for the careful sample collection the studies that had been done to examine why the leukemia has occurred would not have been possible. We most recently examined the information in March of this year, and readvised our recommendations, although -- reviced our representation -- revised our recommendations although rapidly. And -- it's the slides. We revised those recommendations between March and today. New information has emerged, although nothing directly related to the gene transfer. And we are presenting for you this morning a reconsideration of the recommendations in March. And that's what we will discuss over the next ten minutes or 20 minutes or so. And these are the questions for and the review again -- I'm moving. And to give credit to the appropriate people when I arrived this morning I wasn't certain that there were any overheads or slides prepared. That was exactly 30 minutes ago. And within the last 30 minutes our wonderful staff has put together overheads to guide us through this discussion. So the questions for consideration are the same as those that we worked with in March. And I think everyone can read them. I already commented on the first. I suppose I should read it into the record. Should the assessment of the balance of the potential benefits and risks in X-SCID protocols be modified once again? In 2003, we made a series of recommendations, the most important one was, quote, retro viral gene transfer for ex-link skids for patients who have failed -- stem cell transferation or no suitable stem cell donor can be identified. End quote. And rather than discussing that at this juncture, I think we will move forward fairly rapidly to the statements that we put together in March. Then ask simply is this still how we wish to proceed? Point four, extends that potential point three, extends the potential qualifications to other SCID protocols beyond X-SCID and includes the ADA deficient SCID protocols and possibly chronic grand uleo TOU -- disease at least reaffirmed for everyone is not a form of X-SCID. It's not a deficiency or an abnormality in lymphocytes but rather in luterofalls Mono sites. It's a different form of immuno+ deficiency affecting a different group of cells but still monogenic in nature. Four, even moving beyond immuno+ deficiency, are there other disease indications using retro viral vectors that we should rethink in light of the X-SCID information. And finally, what new information should we ask be included in the informed consent document to reflect the X-SCID experience. And should appendix MM be modified? We asked earlier how might the risk of leukemia be reduced in gene transfer studies using retro viral vectors. We discussed that at length and that's really not the purpose of this morning's discussion. And then next something that is an underpinning for all of our discussions, what is the ethical status of a research intervention that provides affective therapy as a does in X-SCID but carries a severe and recognizable risk. And it looks like those are the slides that the staff put up for me. So if you will give me a moment, I will get my handout. In your pacts, you -- packet, you all received a copy entitled follow up from March '05 safety X-SCID symposium, if you all please open that and go to the last section which is the draft of the conclusions and recommendations that we made in March. We reviewed a substantial portion of what is summarized here for us, and probably can agree to all of the bullet points that are above the formal recommendation. And those bullet points include first the majority of children in the X-SCID gene transfer study had major clinical improvement to date. Are there any comments or changes in that sentence? All right, second, of the nine children in this study who had successful engraphment of their gam eas transduce cells, three developed leukemia three years after treatment and required chemotherapy. One participant subsequently died. The overall frequents of this adverse -- frequency in this trial cannot be -- that's straight presentation of fact. Next, the gene transfer was a clause of the leukemias. Next, the occurrence of leukemia in this protocol is not a random event and constitutes a series of at risk in this study. Comments on wording. All right, next, some subjects in gene transfer studies for nonX-link skid, AGA skid and chronic grand lmitous disease experienced mild to moderate clinical improvement. Now, apparently at the recent ASGT meeting and I did not attend that meeting, there was a report of significant clinical improvement in two adult subjects in contrast to the infants who were enrolled in the gene transfer study and X-SCID. So two adult subjects. After gene transfer the retro virus expressing GT of 91 fox which is form of chronic grand lmitous disease. My question have a personal question would like the group to assist me with, is that I have not personally reviewed that data nor have I seen it presented or written and I am a little bit concerned about including that specific a scripter in a -- des scripter in a RAC recommendation and I'm wondering if we should just go back to the X-SCID paragraph that we have and not include chronic granuleo us to disease but I would -- granuleomate us it disease.

Regardless if it has been published.

I think -- I agree. I think it's very specific because it actually mentions the precise gene and as everybody here probably knows there are multiple causes of chronic grand lmitous disease in terms of genetic abnormalities and I think this is specific date thaw we haven't reviewed. Provocative, interesting. All right.

I don't recall us discussing the integration of the participants with -- inaudible --

I'm sorry where are you?

I'm sorry, I missed that in the discussion. Should we go back to that paragraph and then move forward to the end? All right. There is new language in the intro ducktory paragraph that -- introductory that is open, the entire paragraph is open to discussion. Online seven -- on line seven, three children inserted two children has been inserted and is that certainly correct. On line 10 the description of the integration sites is expanded and is that, Dr. DeLuka, that's a section you would like to go over?

I wondering where that came from, the published information. The last meeting we didn't know where -- inaudible.

I haven't seen that information published either.

I tend to ask the investigator from the French of the X-SCID link report recent finding of that integration -- participant. But this data will present it as a meeting but I don't believe it is published.

That's correct, it's not published.

It's not published? So we are in the same I would say predicament or position with this data as we are with the chronic granulomitous data provocative but not published or reviewed by any of our members.

Diane, I make a suggestion because it is clearly data that's been made publicly available, although not yet peer reviewed and published, perhaps, and furthermore it was not presented at the March meeting and so I wonder if it would be reasonable to go back to the initial language but add a footnote to that language stating that additional information has been recently presented at the ASGT and describe what was presented there. It might be a way to handle it. Well, if we did that, I would propose actually referencing it to a specific abstract or presentation of the ASGT so that people could get back to the primary information. I'm a little uncomfortable moving in to RAC language what most of us would agree is unpublished information. But I would like some further discussion. Yes?

I actually think in some ways that this particular point is different to the point you raised earlier. The point that you raised earlier relates to a whole area of information that we haven't had an opportunity to review. While I look at this information as more of additional data that has come up since we were able to have a full discussion. So I actually like the footnoting or some option such as that so that it's an information point as opposed to part of the recommendation. In the other area, the chronic granulomitous disease I think it should be left out all together.

There is a qualitative difference between the kinds of information we are discussing.

I think you have to take into consideration this data was released really before the ASGT meeting and is new information. It's not clear what this information means and I'm not saying isn't Val vau led because it's been sequenced but you have to take this into consideration, this was just released a few weeks ago.

It's an interesting discussion issue because the question before us is, what do we do with information that is very new-- was anybody of the RAC members, was anybody in the session at which this was presented at ASGT? None of us have had the opportunity to even interact with the investigators and we haven't seen the data?

You could be a little less specific if you wanted to and you say use the word to appear involve LMN2 and other genes and I think it's important that LNO2 was involved because it links it to the other two.

I believe at the last RAC meeting when we were discussing this -- in March I thought I heard one of the discussants remark that at the time they didn't think LNO2 was involved, okay? So now it is involved and I like the doctor's comment, appears to involve LNO2 and possibly other ONCO genes and they probably couldn't tell and there may not a way to tell which of these integration events is causative of the leukemia. So I don't know -- that's relevant to list them like that.

Dr.

Although it's certainly anecdotal, I think people in the retro virus community began talking that there was an LNO2 insertion, apparently, although I haven't seen the data myself as early as early April so actually quite soon after the meeting. But I do like the suggestion appears to involve and perhaps others.

And if the wording were that, which I really like your suggestion, then we could even list the others and others including "appears to involve" and "others including." Other thoughts? MissKwan.

The issue here is RAC acts as an information disbursle to the public because we were the only group that all of these discussions in public all the time. To let people know that additional information is around but that they should go looking for it themselves because we can't stand behind the validity of it.

have a concern, though, if indeed these data are correct and if it were to be determined that these other integration sites were the cause of the leukemia itself, it would suggest to me that this is information that we should look out very carefully because it might in fact increase the level of risk we see using this vector and it may augment some of the -- of our recommendations. I think it requires very serious conversation with those who generated those data and perhaps reassessment of some of the statements we made already.

Certainly agree with that. But in terms of finalizing a statement about X-SCID, the -- and gene transfer, the issue is how much of this information should we include. And I am comfortable including "appears to involve." I'm comfortable loosening the language and then leaving the information in the body rather than resorting to footnotes. Dr. Lo.

This is clearly an important and moving target. This will be the data or from the French are published and future events dictate.

Dr. Patterson.

Just hear about what we are hearing right now the phrase is "integration sites in the cell third participant appears to involve BMI1LY1LMO2 and" and I think a we have heard here is this information would be put as a footnote with a specific reference to the abstract and presentation that ASGT and would while it's currently says it appears to involve, we further saw in that that a third participant appears to involve LMO2 and possibly other onco genes and list those. And I think to Bernie's point I think that it's true that each day goes by hopefully we will have more information and we could sit here constantly and revise these at some point we have to fish or cut bait and go out with the conclusion based on the information as it is today but it's important to revisit and I might propose that we invite the investigators to come and present this data at an upcoming RAC meeting so you have a chance to hear the data first hand as questions and discuss it and revise your recommendations as need be. Does that accurately reflect sentiments?

Dr.

Might I suggest that in that -- where you convey that information we make the statement that the RAC will be inviting the scientists to discuss this at the next RAC meeting so that they know it's a continuum.

Let's move on back then to the issues under the clinical bullets. And the last bullet currently states some subjects of gene transfer studies for nonX-SCID, eg, chronic granulomitous disease experience mild to moderate clinical improvement. And the question in my mind about this bullet is that the nonX-linkSCID and chronic granulomitous disease, and chronic granulomitous disease is not a form of nonX-SCID first. Second, I don't recall our group reviewing at all since the initial discussion in the fall immediately following the first two X-SCID SAEs, I don't recall our group sighing the chronic granulomitous disease data. So I actually would suggest removing any comments about chronic granulomitous disease from this document and narrowing the document when we mentioned disease to just X-SCID and nonX-SCID. Is there any discussion of that approach? And then we come to the recommendations. The first sentence of the recommendation has been modified. Pending further data or extend with theeing circumstances reviewed on a case by case basis, retro viral gene transfer studies for X-link SCID limited to -- identical stem cell -- period. So so exclude or for whom know suitable stem cell donor can be identified. And the reason for doing that at least that we discussed in March, is that it is I suppose remotely conceivable that we would not be able to identify a stem cell donor but it's very remote that we couldn't find either a you have a comment?

I would say there is a chance that the -- may not be -- and may not -- criteria. Inaudible.

You're correct. Without going through all of the specific criteria, for HAPLO donors that they do exist

Particularly -- inaudible.

So are you proposing that we leave in the statement for whom no suitable stem cell donor can be identified? To make certain that every child is potentially eligible. Dr.

I think that will take out the rest of that paragraph and it adds nothing to what is already known about standard practices and procedures and informed consent and monitoring. It makes it clear anyway in that it's case by case. I prefer us not characterize experimental investigations. I wouldn't use standard of care language anyway as compared to standard treatment. At least cases extinguish those two -- Standard of care has liability for some people, lexicon not all -- I would just rid and I don't this it adds to our vision or any special advice we have to offer the world. Then would you leave that comment about the stem cell donor -- inaudible. I had a similar memory from that discussion but what it changed.

The group has been struggling and that's what the emphasis is from the doctor attending the ASGT meeting, the group has been struggling with the continued debate amongst those who provide HAPLs to babies with X-SCID and nonX-SCID. The continued debate about the relative efficacy of HAPLOs in that study, identical marrow transplants in that setting. You will recall we had an extension discussion during the March meeting about the need to ablate and prepare the marrow in babies with X-SCID versus not doing that and we had two very senior established investigators present to us, Dr. Buckley and Dr. O'Reilly who came to us with very different opinions, both with longstanding and extremely successful transplant programs. One using ablateo therapy and the other not. We were left with a dilemma. We can handle that by not matching it at all, which I think Dr. Powers is a you are saying, there is no established -- the discussion between those two experts and the different but related discussion at the recent meetings about of HAPLOs means that there is no standard approach. That this is -- and I wouldn't call it experimental. It lies somewhere in between. It's a judgment about how to manage and care for a baby with X-SCID.

I think my main point is we ought not be the business of characterizing which of these are experimental and we shouldn't be experimental or standard of care. These are all case by case gum of unproven things. What I don't know is whether as Dr. Says we should perhaps reinclude all for whom those suitable stem cell donor can be identified because it did sound as if the last meetings -- that there are instances which that seems to be an appropriate avenue and I wouldn't want us to foreclose that possibility by leaving that part of the sentence out nor would I want us to foreclose haul these continuing debates by adding any further material. I mean we probably shouldn't rush to some premature closure.

I think we are all in agreement. As I look at the wording this morning, the clause we were talking about or from no suitable stem cell donor could be identified really relates to inclusion and exclusion criteria and we were narrowing them too much if we delete that clause and the remainder of the paragraph deals with how we should treat babies with X-SCID and not our purview. I like your proposal very much. Let's clean this up, simplify it, reinclude that the clause that the doctor commented on and not use the rest of the paragraph. Which returns us, I must remind people almost exactly to our earlier statement, which was very clean in my mind that the final bullet is straightforward. We are not recommending any modifications to it at all. Any further discussion? Dr. Patterson, do we need a vote?

Sure do.

Okay.

I thank you all for thinking about this over the last two years and it will continue to -- the statement will continue to be modified as others have commented on as new information becomes available to us. And we on staff will make certain that as important information moves into the field we invite the investigators here in a timely manner to present that information. So thank you. We will -- yes?

I would like to raise a question related to the discussion we had last time and that was the presentation I thought rather compelling evidence that half identical transplants worked quite well in early infancy and not infected children and there it was a iffy proposition done later in patients that were already infected. And at that time we discussed appropriately then any issues relates to screening for X-link SCID to be part of our recommendations but was going to think about ways to weigh into that discussion going on in various -- inaudible. I can send you all out by E-mail to the entire committee we did actually look into that and the first of all there is a separate committee, a second father-in-law -- secretarile advisory committee to add to the screening that are currently in place for newborn screening. And the test for SCID has been on the docket or on the table and under discussion until recently how a reliable easily reproducible highly sensitive test with good reliability has not been available. Primarily bio-chemical. The more recent test developed by including the team here in NIH for doing PCR based diagnostics has recently come forward and under evaluation. This issue is back on the table again and under consideration. It is recognizing the efficacy in many instances of early transplant in these infants. The issue is one that has come up frequently but it's been the quality and the reliability of the test that is at issue.

would add to that that we are all struggling with including newborn diagnostic testing for X-SCID in our newborn panels. And certainly I can speak for California, we were looking carefully at using the PCR based methodology which does come from NIH primarily and including testing for X-SCID in our newborn panel. Other comments or questions? I would like to move forward to our first protocol for this morning. The discussion of human gene transfer protocol 6 13RBGS Phase 1 dose escalation study of intratermeral herpes simplex virus mutant rRp450 in patients with retracktory sarcoma or neuroblastoma. I been asked after I introduced the protocol to announce the recuseles of Dr. Hess lap has already left the room. She is a member of DSMB for this study being conducted at Cincinnati children's hospital. I understand that Dr. Timothy Cripe, M.D., Ph.D. will lead the discussion this morning. Thank you, Dr. Cripe.

Thank you Dr. Wara, members of the RAC. Pleasure to be here today. I appreciate your thoughtful review of our protocol today. Look forward to in-depth discussion today. I will be discussing our proposed proceedical which is a face I dose -- rRp450 in patients with retracktory sarcoma or neuroblastoma. I would like to begin by talking about the tumor targets we are looking at for this trial. Proceed then to discuss briefly on collatic and use for potential cancer therapy and describe vector we are planning to use rRp450, the structure, efficacy data and safety data and describe and outline the clinical trial. Beginning with the first topic, the tumor targets we are looking at for this clinical study are relapse soft Tish knew sarcoma in pediatric is ram dough -- sarcoma which occurs in children as young as adolescent and young adults particularly those relapse. -- which is the most common soft tissue sarcoma if adults and variety of other sarcomas that we tested in preclinical studies that maybe eligible for this trial including for instance malignant tumors which are common in patients with neuro fibromaitoussis. In addition this clinical study would include patients with relapsed bone sarcomas and osteo sarcoma or one of the members of the sarcoma sarcoma of tumor's the final target is relapse neuro belstima. What they have common is their dismal prognosis when they relapse as illustrated by the neck several slides. This is a survival curve of patients after their diagnosed of the sarcoma and in this particular graph it's for those that are in the -- area. And head and neck that as you can see if most patients do not survive this, very few do and it depends a little bit on the characteristics of their relapse regional versus CNS, et cetera. Similar data are found for the other tumor targets that we included in this study. Here are data for relapsed osteo sarcoma where most patients die of their disease the first year, year a half following relapse. Survival after relapse Ewing sarcoma is no better with the curve dropping close to zero the first year for those who have received paltive therapy with mild prolongation of survival with therapeutic treatment meaning somewhat usually experimental agents on clinical studies. And finally survival after diagnosis of neuroblastoma and in this case it's not even relapse but high risk neuroblastoma and those with certain age groups the survival again is dismal. So these collectively these diseases represent significant unmathematical need moving on to the discussion of -- attractiveness as potential cancer therapeutic are many fold including their ability to kill cells directly. Often turned okayo liesis. And bypassing therapy that often mediated by p a 3 mutations and ability to self-propagate within a tumor and overcoming the gene transfer inefficiency that is commonly found with recommendation incompetent vectors and enhanced delivery of therapeutic genes. The -- are illustrated on this slide with the genomic structure across the top and consists of a long region and unique short region both of which are flanked by repeats. The NV1020 vector that is reviewed in the first 100 RAC in the past, contains a deletion or replacement of genes in the internal repeat region with replacement of likeo protein sequences from agency which was to using it as a vaccine for both types 1 and 2. This deletion and replacement results in deletion of the UL55 and 6 genes. There is a deletion of the UL24 gene. Mechanism of tumor cell Shrektivity for this vector is not characterized. The ability of the virus to replicate because it is impaired partly it's thought because of the deletion of the -- repeats in the middle allowing the virus to not undergo its proper recommendation but it does have intact the gamma 131- a gene which is touted as a neuro veer license gene which is important to be deleted for safety. One of those copies in the gene and other repeat on the geno. And TK deletion. But it has been replaced under control of the early promoter alalpha 4. Complete absence of the and coded gene gamma 131.5 demonstrated by the deletions of the long repeats. And it also has insertionle mute genesis in the ICP6 gene for the gene and coding ICP6 which encode the large sub unit private -- re-- and in that vector there is the beta Gary Locketic gene. Final vector that's tested in humans fairly extensively is 17-16 most of this work has been done in Europe. It contains simple simply deletions of the gamma 131.5 genes. These vectors have been used in I'ms on doses used in this slide. The 1020 gene is used in published in abstract form by human -- with intraPatic arterial injection for patiences with metsetic colon cancer to deliver up to the plaque forming units without significant adverse events. has been used in intracerebral or tumoral or for brain cancer. Up to three times ten the ninth by forming units and 1716 has been used in patients with brain cancer with tumor injection and head and neck squamous cell carcinoma with tumor injection and doses up to one time tens to the sixth. Vector we would like to use in this trial is rRp450. This contains insertionle deletion mutation in the UL39 gene coding ICP36. In place of that gene appears the CDNA encoding -- which is p450 enzyme pro drug enzyme that activates -- it's under control of the early ICP6 promoter from the virus. This gene retains or this virus retains the native -- and it's sensitivity to a -- The HSV1, 139 or I-36 protein, coding the large -- converts -- as illustrated on the slide. This critical for virus recommendation to generate sufficient pools of -- for virus recommendation and also critical for reactivation of the lateen virus. Tumor cells are known to express high levels. As illustrated in this from our laboratory with this panel of human malignant tumor using also the monkey kidney as a control and these are compared on the farther left in norm human cells which do not express high levels. So this differential expression of this enzyme is the major basis we believe for the selective recommendation rRp450 in actively dividing tumor cells. We published the number of papers showing the efficacy of the vectors for the target many of the target tumor times we like to include in this clinical study both cell lines and human xeno graph data to show the cell line date u a is ile strighted -- illustrated on this -- indicates a ten fold increase in susceptibility or sensitivity to the virus in an MTT assay. We looked at random sarcoma and Ewing sarcoma one cell item and itema and large panel of cells derived from neuroblastoma and new from it was not included in our appendix and maaing ininant tumor cells. With the vectors in clinical trials we found a broad range of sensitivity to these vectors both. G207 was weaker in general and the Ewing sarcoma cell line was not sensitive to either of these. Where as the neuroblastoma and malignant tumors were quite sensitive to infection and -- by the vectors. RRp450 showed somewhat increased sensitivity or siteo toxicity for many of the cell lines particular note was the Ewing sarcoma cell lines which appeared sensitive to infection with rRp450 increased above that of those vectors that are used in clinical trials. In addition to even those cell lines that showed quite a bit of sensitivity, neuroblastoma and the cell line they also showed increased killing with the rRp450 vector. This data suggest that rRp450 may be more ef conscious than those vectors in clinical trial and parentthetcally some the criticism of these vectors has been that although they have been shown to be quite safe in human study, the efficacy has not been robust. Using xeno graph tumor models with human -- from human tumors placed in mice we shown remarkable antitumor effect with a single injection of some the different viruses as shown here with NV1020. When tumors were fairly large between 250 and 750 millimeters cubed and in this study that single injection gave us a 50% complete response rate four out of eight tumors shrunk in the eight weeks disappeared and remained absent in the animals until the end of the study. The other in another two we had a partial response meeting that the tumors to less than 50% of their original size although they grew back to progressive disease. And one of them appeared to be have stable diseaser to awhile and then grew back as well. We had an overall response rate of 75% and we have similar data that we now published in models of neuroblastoma as well. Using the rRp450 vector and in this study at a lower dose 6 times to the fifth, we also had a significant antitumor effect with a 50% complete response rate, five of these ten tumors shrunk in the first three to four weeks and disappeared. of them show prod gr'sive disease and another several had stable disease at the time of the end of the study. Considering a complete response and a partial response to the response rate here was 90% or nine out of ten and base ten. Most of the concern with this vector has been its potential toxicity and therefore we embarked on a number of safety studies recently I would like to review. began those with looking at the attain wings of our rRp450 recommendation in normal human cell lines. This data repeated several times from normal human hepet site showing hepetsets showing ROP4350 or as a control. The apparent strain from rRp450 was derived and the cost strain. Here examples are shown for multiplicity for infections. 01 .1 and the culture was measured at one hour, 24 hours, 48 hours and 72 hours. In the rRp450 cultures, infected cultures we did not detech virus recommendation over time the amount of recoverable replicating virus as determined by plaque assays fell off with half is that seen when add virus to immediate alone without cells. In contrast the cost vector replicated robustly increasing three to four orders of magnitude over those 72 hours. So these data suggests the rRp450 is quite attain with theed and normal human pepetsets. We know in cells that divide this culture andup hepetsets are not healthy and do not take well in terms of dividing in culture but cells that divide are predicted to support virus recommendations than those that are were live are theed cells to have enough regulation reducatis. In cultures of prolive are theeing which came from donors we do detect robust virus recommendations both of rRp450 as well as -- and in this experiment we included the NV1020 vector in human trials and showed it replicated to a similar degree in these cultures. Interestingly and has predicted when the cells were differentiated with calcium we no longer saw probust recommendation of the rRp450 vector. We did see continued recommendation of the wild cost cost vector on the far right and interestingly enough we saw fair amount of recommendation of the NB1020 vector. So the data actually suggest rRp450 was more attain with theed and attain with theed if not more in NV1020. Terms of our safety studies we performed a variety of studies in mice. And we took basically what was our highest available or given volumes and tighters and injected mice with to the 8th plaque forming units of virus. We use virus alone plus Foss md. Fosso md was added 24 hours after the virus injection to allow for transgene expression from the virus. The clinical studies after both intraVenus and tumoral virus injection were performed on day three and late time based orecommendations by the FDA and we found no clinically significant effects on animal waits. Complete blood counts electrolytes renal function or liver function in the detail graphs of those are in appendix M. We tested virus with the additional psycho Foss and did not see added toxicity due to virus that we in addition to that toxicity with -- as predicted it did drop blood counts those recovered. We also performed survival studies with both intravees if and intracerebral virus injections. The data on this slide with the well type virus are new but used as a comparison to the data with the rRp450. So when we gave animals this virus intravenously or intracerebraly it was toxic and had tended fifth dose of plaque forming units nine of ten animals were alive but all of them had neurologic toxicity, paralysis and nine of those recovered. So there did appear to be morbidity at that level and doses higher than that, none of the animals survived and these animals in these groups died within three and seven days following intraVenus injection. For the intracerebral injection, two of the ten animals died. Ten to the fourth and all of the animals at ten to the fifth. These deaths occurred within predominantly in the first week following virus injection. These data suggest that this model is suitable for assessing toxicity. At least to this degree. In contrast when we used rRp450 at the highest dose tested with well type 10 to the eight -- all of the 20 animals injected with intraVenus virus have survived and the experiment is out to 150 days. Intracerebral injections two of the mice die shortly after injection and was thought to be due to the procedure and or the anesthesia but 18 of them have survived and now out approximately a month or so. So the data suggests there is quite several orders of magnitude difference in VIVO toxicity of rRp450 compared to HSV. Similarly in those studies we have added extra -- Foss md 24 hours after injection of the virus. Here are the data from the previous slide recording rRp450 in animals given psycho Foss md alone we routinely lose about 10% of those animals due to toxicity from psycho Foss md presumed low blood counts susceptibility to infection and when we add in the virus to psycho Foss md we also lose about 10% no different from psycho Foss md alone. These data actually suggests that the combination is no more toxic than psycho Foss md alone. So finally I like to discuss our proposal for the clinical trial. Our aims are to establish the maximum tolerated dose of rRp450 administered as a intertumoral injection within the con fineses of the dose escalation scheme we won't continue to escalate until we get maximum tolerated dose but rather escalate into the highest dose that we are able to give based on the tighter and rather try to push it even further. Secondly to determine the dose limbing toxicity of using intratumoral of rRp450 in humans. Our secondary aims are to measure antiviral immune response to measure the systemic Byrumia and cee fine the antitumor activity of rRp450 within the confines of a Phase 1 study. Pairthen catted as well this study will serve as platform for future studies with the combination of psycho Foss md order to enhance efficacy. Our criteria are patients that are older than age two and younger than 30 so they may be treated at a single institution. Cincinnati children ps hospital. Hiss logic diagnosis of the sarcoma neuroblastoma that is relapse to therapy where there is accessible measurable disease accessible meaning it has been determined by the interventionle radiologist and the life expectancy of the subjects will be required to estimate at greater than eight weeks. Our exclews criteria are those that we use in standard pediatrics face I driveteria no acute toxic affects are active and ongoing from their prior therapy and sufficient time from prior therapy to recover from those such as mild suppressive chemotherapy and othery ino plastic biologic therapy and radiation autologous bone marrow transplant and no patients would be allowed if they received the transplant prior to gene therapy, pregnant or breast-feeding. Our treatment plan is to administer the democratic CT guided intratumoral injection. Virus we propose to enroll three young adults first in the age range of 22 to 30 and subsconly for each dose stratta is then enrolled pediatric patients. We will give up tower injections three weeks apart and initially plan for six dose stratta beginning at one time down -- and escalating out to three times ten to the ninth plaque forming units. Our dose determination was done because there is no maximum tolerated dose determined in humans yet. Each of the studies I cited gone up to the previously determined maximum dose. Again, the maximum dose with NV1020 which appears to be similar in terms of its attain wings if not less attain with theed with rRp450 was 1.3 times ten to the sendth. NCI and -- has allowed us and pediatrics in general to begin phase I dosing at pediatrics at 81% of dote es which is one times ten to the 7 pght. Procure a gram basis the exposure will be at this initial dose at 10,000 foal less than the maximum dose tested in mice which was ten to the eighth|1 th |0. Regarding good manufacturing production of the vector, I learned on Monday the national gene vector laboratory has provisionally agreed to support this project and need administrative issues though I not received the official notification and if that's the case, then we would plan po produce this vector in the human gene therapy -- university of Pittsburgh. In fact, through Dave we had a number of successful premanufacturing runs with rRp450. The method that he uses are purification methods though we have not been privy to the details of those methods. He has achieved 8.5 plaque forming units per mill. And with these premanufacturing runs which feed quite sufficient for the kinds of dose pros posed in our study. So just to conclude my formal remarks, the intended participants have incurrable cancers and therefore there is quite unmathematical need. Cancer model of these diseases are susceptible to -- the efficacy of ROP450 is attractive because of the increased siteo toxicity and use it as a platform for future combination with and rRp450 from our study appears safe in that it's attain with theed by fold versus wild type -- cells. It is at least attain with theed as MV1020 which is in human trials. High doses tolerated by mice. And there is no added when combined with psycho Foss md. Therefore it is our belief that rRp450 warrants safety testing in subjects with retracktory potential -- cancers. I look forward to our lively discussion about this proposal. Thank you very much. We will begin this morning by asking the individual reviewers to review their initial concerns publicly for the record and then to state whether that concern is resolved or continuous. Those concerns that continue will generate some discussion with you. If you like to stay there, that's probably -- you may either choose to stay there

I'm happy to main.

Why don't you stay right there. Dr. DeLuca, would you like to begin?

Sure. I would like to thank Dr. Cripe for the presentation today. His responses he has answered some of my questions, quite a few things. So what I would like to do is to go through each point and then perhaps -- a point we didn't have a discussion. I think we already been through what the -- protocol. The stated objectives I would like to add something here, the state of the object Science Centertives of the protocol were to opts objectives of the protocol. I didn't so there is one thing I need clarification now pending your talk is the -- do one of the objectives of the protocol depend on the expression of the rRp450 gene?

Do one of the objectives? This first study do not depend on the expression.

That's clear. So most of my concerns had to do with safety and toxicity of the virus and I will go through those. I had some questions about the slant put on the tropism of the virus. This virus and there is actually a very good paper out there from 1988 from Goldstein and Weller for people who first made this virus who showed that this virus -- is defective for growth in dividing cells. Normal human di employed di ployed. The other important thing which we will get to in a minute is it's temperature sensitive for growth no one knowize. It elevated temperatures the virus will not grow even in dividing cells. So given that and the results of another paper from Don Cohen's lab in 199 where he first looked at the pathogensis of this model and showed that the virus was attain with theed in the growth in the -- and for the establishment and reakes have of latiness that his conclusions were that because the mouse has an elevated temperature, one has to use -- these are -- I could quote this thing but I would have to get online to do that and we referent allowed to, apparently. -- we aren't allowed to, apparently. The main conclusion is that you could not extrapolate from mouse to humans with respect to the pathogeninousty of this virus because of the elevated temperature of theThis to something that the and I had was that a discussion about the need for some other animal model or for primate data which was done with the other two herpes, the other three herpes trials, the one with the gamma 4.5 mutant and one with the big deletion and the one with the ICP6 gamma 34.5. So I would like to have some discussion on why -- I'm not sure I understand your response why you think you don't need a primate model or some other animal model to test for safety.

Sure. I think in my discussion with members of the FDA their feeling was that the need for various animal models is dependent what they take into consideration is risk benefit analysis of the proposed clinical trial and proposed route of the administration, et cetera, and their feeling was that even primate studies would not necessarily accurately reflect the situation in humans and that ultimately if one needs to test safety of this vector it needs to be done in humans and therefore their criteria and level of concern would be predicated or dictated by in part by the risk category of the potential enRollies of the clinical trial and they felt that the showing safety in mouse models might quite well be sufficient, obviously they couldn't give a final verdict until they reviewed all the data. And the fact that the data from the 1980s showed elevation of the temperature attain with thees to some degree the virus -- we have taken the temperature of our mice in our studies. I know that the average mouse is slightly warmer than the average human. Approximately 38 degrees. And it's true there may be some attain wings of the virus at that temperature. I don't know our data about continuation of -- herpes or any other type of mutants as well. I guess our bottom line was the -- it wasn't clear that it seemed clear that this virus is with theed compared to wild type in these mouse studies and indications from the FDA is that may be sufficient. So in the interest of being able to move the field forward, in a relatively timely fashion, and actually to make it doable from resources standpoint, we have not undertaken those studies.

Well then let me skip forward to my sixth point. Tan little bit on that. If we look at the other viruses that are in clinical trials they have all been tested in in primate models, not human primate models. The two I want to talk about are the G-207 and 1716. Those are good ones to think about because G-207 has the same mutation that you have here. That has been up to three times ten to the ninth and very attain with theed. Now taking the single mutation, 1716, the one that was done over in Europe, they only gone up as high as ten to the ten to the six in virus and that's because that virus is -- that virus is easy to hurt monkeys with. And you have a starting dose of 10 to the sendth here. Due to the lack of testing -- ten to the seventh here. Due to the lack of testing shouldn't you consider lowering the dose especially since you are proposing to inoculate children.

I think it is reasonable question and my written response we had basically been worried about losing efficacy by going to -- I guess we will be happy to revisit that question.

What's the primary objective of the protocol and the primary objective of the protocol is safety.

Correct. However, it's against the code of federal regulation to do clinical trials on children without the possibility of some benefit and therefore we felt that this was a dose that did well in our mouse studies in terms of efficacy and therefore had the likelihood of potentially having benefit in humans. But again, we are happy to consider a lower dose if the lack of primate data dictated that consideration.

My third question is, has a marker rescue version of this virus been generated to assure the attain wings is solely due to the coding sequence? The reason for doing that is because this protocol will be precedent setting and that's why it's reviewed here. The a virus with a single deletion in it has never been used in people before. We want to know whether the attain wings is due to that deletion just in case other similar protocols come forward in the future. If this does have additional mutations in it which are likely that often happens with herpes, then it may be attain with theed for another reason. Now I know you said that in the 1020, that rescue weren't generated for that but they were for the G-207. So something it to consider. And the other thing would be give an the novelty of this vector for use in humans and the fact that the p450 gene is not employed in this application, it be better if you just had an ICP6 mutant that didn't, prex p450 and -- p450 and the virus will be established latiness and it can replicate and have expressing this p450 gene and it adds another variable that really isn't used in this trial. It doesn't -- it seems like it's an add-own. Almost asks for expression data which you don't include in this. There is a transgene here and there is not expression data given for this transgene in a clinical -- in a kind of model. It would be clear and better science if it was just ICP6 deletion unit.

Now, in my view it speaks somewhat to the practicality of the matter in formers of what we like to do is use this as a platform to use fuseture studies and as you know if we were that want to expand to a p450 transgene expressing vector it would require enormous amount of resources, time, effort, money to then redo all the studies that we are proposing and all of the preclinical data with the separate vector. of that data will have to be reduplicated again with vector now that doesn't contain that virus where as if we show safety with this vector that we will look forward to look at transgene expression and potentially expand that array of studies that could be conducted with this vector in the future. I think it's somewhat of an investment in the future and include that now.

In appendix M in this trial we usually call for or requires an expression data transgene for this particular trial in some kind of model duration of expression of what cells are the transgene expressed. Are data like that available?

Data like that are available. Most of that is done by Kent as general who has looked at transgene expression, functionally. Mainly there is not great assays for functionle transgene expression other than taking superor conditioned media from virus infected cells and using that to show toxicity of other cells from it. Akin to one might see with the addition of the chemotherapy to those cells with appropriate controls, of course. I'm not aware of any inviiveo data.

How about the western block? Flies the IV6 promoter?

Yeah, data there. We haven't been focussing on the use of that transgene, I don't have those data at my laboratory.

Do you think this will listen to the rat -- than listen to the immune response?

It may well but there are a lot of foreign genes expressed because all the virus genes are foreign. Usually with at least in mouse models of the infection there are immuno+ dominant peptide that most of the antitumor activities again in many cases depending on the mouse train is from lack of protein B.

Most of these people have antibodies to herpes.

Many will and if we go down to the lower age rates are talking some of them may not. But I don't think that the contribution of the rat gene to the immuno+logic reaction will be significant relative or within the contact of all the other viral genes that are being expressed.

And my last point is to what extent does the virus express lay-in see used in a modwell cancer? I know in the standard herpes lateensy model the establishment is very low. Now you are putting in the cells where presumably theoretically this virus will replicate well and generate a large pool of virus and then can go and feed peripheral neurons. So the question is in these mice that get these, that's your treat, cure, are there ganglia loaded up with the herpes rRp450 gene.

We have not looked at that. I worked closely with Nancy who is a expert in herpes lateensy and it's not known and it's clear administratoring high loads of virus by other routes than what I acquired may result in cells containing lateen genomes that otherwise would not have and that's an unknown and something that its yet to be looked at. don't have any data about lateensy in our animals but we have tissue sections in our animals that could be looking for a lateen virus. We aren't concerns because lateensy is a welcome risk in this high risk population in patients that will succumb to their disease.

And lastly, did you actually test for -- sensitivity.

We did. I can show you a graph if you like to see it. That was tested in the contact of psycho Foss md and against ant cyclever with a concept the fear that the psycho Foss md might inhunt virus recommendation and so actually in that experiment was used as a control and that it inhibited rRp450 recommendation four or five orders of magnitude where as the psycho Foss md did not.

You had minor points and I think were were -- they were taken care of and that's all have I.

Dr. DeLuca can you review for us your remaining concerns?

My remaining concern, the reliance on the mouse safety data. The marker rescue virus and the use of the rRp450 gene. And the high starting dose rate.

Thank you. Dr., let's move to you.

Dr. Cripe, first of all an apology in order to talk to the microphone have to turn my back it to you. But thank you for your responses to the concerns that I raised. My major concern again is the safety issue and the lack of not human primate data that Dr. DeLuca has reviewed. And that has been discussed. So I still concerned about that, but I really would defer my concerns as well to Dr. DeLuca who is far more expert on herpes. A second concern was that of about immune compromise in patients who have already undergon significant chemotherapy. That has been addrested, although I must say I still am concerned by the comment that immuno+ deficiency related to chemotherapy is not expected to increase virus related toxicity. My third concern was about the insertion of the siteo chrome p50 gene when it's not going to be used the same as Dr. DeLuca has raised and I guess I'm satisfied with the response in part that this is a platform for future study. I raise the question that about comparing the biologies of the tumors to be treated in adults and children. I think the idea of treating adults first before the children in each given doses is highly appropriate. However, with that result in problems related to the different biology of the tumors this has been addressed satisfactorily. In the original protocol that three different potential sources of the transfer vector were discussed and been selected and that has been addressed by the presentation today. And my final concern then was about the informed consent and the expectation of benefit that was raised that also has been addressed by the regulations regarding studies in children. I have a couple of questions related to the presentation. And one just by the way since this virus the attain with theed virus can grow divided cells which include the get, does herpes infect gut cells and do you see shedding of herpes in the stool if so is that something that should be looked for?

Most of the time -- I'm assuming you are asking in certain native certain infections one see shedding of the stools. Most of the time the infections are local, mucosal and -- checked by the immune system. I actually had patients with severe immuno+ deficiency which acquired herpes simplex virus die of HSV disease which disseminated openthal infection in those one rare case one be able to virus in the stool. I'm not aware of the study. Usually it's a local disease.

And then I wondered if you could go back to the two slights you -- slides you showed about the response to the intratumoral injection into the mice. It looks to me as if three of the mice if I'm looking at the numbers right had a complete response. Three others that had a partial response that not sustained and then one started to see further recommendation or further growth of the tumor and one that clearly had an early response but then failed. So where does the 90%

Five of them ticked down in the slide here.

Five, six -- there is ten in the study. Five have -- each of these is a different mouse and they are different symbols but it gets a little difficult to discern from one another as they are clustered in here. There are five.

So there is a 50% -- this from a single injection?

Single injection. Yes. 50% complete response rate and another four that had a partial response rate. In terms of categorizing their best response, these four or these nine out of ten is what I would say has the response rate. Two or three of these that had partial responses eventually regrew at least to the point where we would categorize them a stable disease which means they are 50% of their original size and one had progressive disease which means it grown beyond 50% of its original size. And the other one remained down here as a portion response less than 50% so that's -- it's right there.

And the next slide please.

This is within the 50. Similar data are here. Four of these had complete response and remained on this line at the end of the experiment. Two of those the trunks were less than 50% of their size and therefore categorize as partial response and went on to regrow. One of those that its best response was within 50% of its size. And stable disease went on to have progressive disease and one had earlier progressive disease. So again the initial response would be six out of the eight or 75%.

And in these mouse study dozen you have data on repeated injections as you are planning to do with --

We have data on repeated injections. I didn't show that graph. We have given mice published in our paper courier now. There were -- we gave mice ten different injections, three times weekly, Monday, Wednesday, Friday for ten doses and that was an attempt to be able to shrink larger tumors and I might parentthetcally add that 90% if not more of the study if not more out there tumors there are 50 to 100 millimeters cubed. Very few had large tumors. We decided we wanted to have efficacy in large tumors in mice. So these are quite large in mice. We taken it up to 1100 millimeters cubed and those president experiments where we did ten repeated injections and did get some efficacy in the larger tumors as well.

Are there any studies where the injections let's say three or four weeks apart were done? Something more similar to the clinical protocol?

No, we have not done those because I guess the closest thing we have done is published in our paper which is on neuroblastoma where we have injected tumors as they have grown here. Most of the time they are shrinking and no need to inject them. And those that have regrown we have taketen upon ourselves to do reinjection at points where it's clear they have progress fen disease and in those cases we seen complete responses in many cases actually. There is no data -- no evidence that these cells are actually acquire resistance to HSV. believe it's a matter of dosage and the cells are dividing faster than the virus is by giving it a boost we overcome that. It would be helpful to have seen that. Thank you.

Doctor.

I believe that when I left the room it was stated that federal regulations for research involving children require that you can't involve children unless there is prospective of direct benefit and that's not true. I wanted to point out that there are four categories of research that are allowed for children, one is research not involving greater than minimal risk. The second one is research involving greater than minimal risk but presenting the prospic of direct benefit to the individual subjects. The one that requires that there is direct benefit to the subjects. The other one, third one is research involving greater than minimal risk and no prospect to den -- generalizable knowledge about the subject's disorder or condition and the fourth one which is only approvable by a second father-in-law panel -- secretarile pattern that presents an opportunity to alleviate a serious problem affecting the health and welfare of children.

Clarification, the first or four of those don't apply in this case. Clearly the first one which is can't be greater than minimal risk in healthy children applies. The second one is the one I cited and the third one there is generalizable knowledge about their disease which I don't see this study will yield. I still think that it's not fitting with the code of regulations as written that we would conduct this trial if there weren't any potential benefits. It's not to say that our consent form stating there is a benefit in fact we say specifically that you may not derive any benefit which is standard language for our phase I studies.

So, there is ongoing confusion in the pediatric community and among IRBs about this precise issue that has just been discussed. In other words, in what groups of phase I studies should we move forward in children without prospect of direct benefit to the participants. I think that's a subject that's beyond resolution in this forum, but agree with its having been raised in the contact of this particular study.

Yeah, I haven't really looked closely enough at the protocol to be able to say whether or not this fits into any of these individual categories. However, just from what I know it seems to me that this research would be likely to yield generalizable knowledge about this disease. The times when you get into trouble with that category is when you involving normal subjects, normal children and I don't believe you are planning on doing that now. Those normal children don't have a disease or a disorder or condition so that's when people usually get hung up on that third category. Or that you could argue perhaps that in this case it has to be just a minor increase over minimal risk and it may not fit that part of it. I can see that possibility.

Yes, the experience we had with IRB suggest that category three wouldn't have gone very well.

So the conflict here is what we have been whispering back and forth about and that is the application of the guidance to identification of the starting dose. In other words, the balance between risk benefit to a particular child subject. And Ms. Kwan on that note I will turn to you.

Actually, I would like to before I begin my independent comments to sort of piggyback on the conversation because in listening to doctor-recite the categories, I'm sensing that it might be a kind of a bizarre application of the wording that is being used. I think to me what that says is you are not supposed to expose kids to a heightened risk if there is no potential for benefit. And that makes sense. But I don't think the answer or the response to that is to therefore up the dosage. I think the intention is to decrease the risk rather than to up the ante, so to speak. So that's a lay person's interpretation of what those guidelines probably should be meaning. And I think this isn't first time that this has come up. I think even last time something similar came up. On to my specific review. I was particularly concerned because the population of potential participants are children who have almost no alternatives. And for children as young as two years old I think it's very emotional issue for the parents to be making these decisions. And that makes it all the more important that the explanation given to the parent is absolutely clear because I think the emotions is do anything to save my child. But I think what we really need to convey is if there is not a reasonable chance of helping the child, is it reasonable to expect the child to be exposed to higher risks of greater discomfort in the process of dying essentially. And I think that really needs to be scrupulously handled. So I was very concerned that the language both in the nonscientific abstract and the informed consent seem to me to be very complex, much too technical, and not really addressing some of the issues that I think that the parents of these children should really be confronted with in making these decisions. A number of the specific requests I made have been addressed. The lay abstract has been modified and so have a number of the statements in the informed consent document. But I would just forewarn people I think a lot of the times people are using some of these automated readability indexes to determine whether or not they fit into this eighth grade or below understanding and those are mainly based on statistics such as how many syllables are used in the words you and that often doesn't come really recognize what's practically true of people Reading the documents. So even though there have been modifications, I find that it's still really quite complicated and that may be a reflection of the fact that the issues involved in this protocol are complicated. And I think we did have a symposium at a previous RAC meeting where our outside experts had really strongly said it's not so much what's in the written documents, it's how you handle the discussion. So I would add to my comments that a lot of this can't really be assessed at a RAC meeting but that the investigators really need to be alert to the fact that they have to be extremely sensitive to how this is explained to very emotional parents at this point. Some specifics there was a difference between the age groups that were allowed and I believe that the investigators have now decided that the age range will be from two years old to 30 years old. There were some documents that showed children as young as one year would be enrolled and that's been corrected. I asked that in parts of the document that talked about herpes simplex virus that some recognizable examples be given in and that has been inserted. I asked in general that terms such as patient therapy and treatment be replaced. And that has been done. I also asked that some language would -- which said if you have an adverse result and there are unanticipated infections that you would be treated and my comment was that the language suggested that you would be treated successfully so that there a little change in that so that they recognize that even though there would be treatment, it's possible that the treatment would not be affective. And finally just some again language in the informed consent that said there are other risks and this will be explained and the response was that the explanation would be on a case by case basis in discussion. And that seems to be satisfactory. Just in terms of the conversation that has gone on in this review, I just really think that overall the risks of causing tremendous discomfort or prolongation of discomfort for the participants should really be carefully discussed with parents. If in fact so much is still unknown and those unknowns may get very marginal that there will be any kind of positive effect.

Thank you very much. Any response to?

I appreciate Ms. Kwan's comments I think they are appropriate and thank you.

Dr.?

Thank you.

All right.

Any further discussions from RAC members?

Hi, I'm sorry. I reviewed this protocol in detail so you may have the date -- if you ever done these injections into mouse tumors in fully immuno+ competent mice and also in the situation where those mice have already got some sort of latent herpes infection?

We have done the former and not the latter. We injected in our random -- mouse model in a genetically susceptible mouse for a knockout hepetset growth factor these mice get early onset -- sarcomas and we are working with that model to do exactly what you said. That model actually is quite aggressive. These tumors grow quite rapidly and we have not seen any adverse events from those injections but we haven't got very good efficacy and I think the tumors are just too aggressive for the virus to do its thing. We have not preimmunized or exposed animals to HSV prior to these.

So the injections where you don't cure -- where you can't cure the mouse tumor, how long does that normally take to kill the mouse?

It usually the mice need to be euthanized or the rule at our institution is when the tumors reach 10% of the animal body weight and those models it takes seven to ten days, at the outset maybe 14. That's as far out we were following those mice following.

And you haven't been able to do --

In the repeated injections we can get a little delayed growth and get them out to three weeks or so. And we have done that and again not seen adverse events but not many watch -- I guess my question really relates partly to efficacy obviously but also partly to the issues of toxicity. With these studies which you shown I would predict there would be complex toxicity which might be associated with immune response both to the virus and tumors and any other latent virus existed in the patients. I raise that as another sort of toxicity issue.

I think those issues have certainly been discussed in the contact of the other clinical -- con text in other clinical trials regardle of which mutant interestingly in the G-207 model the antitumor effect is only seen when there is a immuno+ competent animal and they lose that effect in the -- and been shown that the antitumor effect is actually more cytotoxic T-limpset to the tumor specific antigen rather than the virus recommendation.

Is it not possible to use some other model not necessarily -- sarcoma to get some of those data with your virus in immuno+ competent model?

We could test other -- in fact that has been done at Mass general he has used mouse colon cancer models. But colon cancer models are immuno+ competent animal. His model injection the tumor sells and follows the virus injection and either in the spleens they get metsetic lesions in the liver and it's delivered by injecting it into the spleen and they have done extensive studies in that record with rRp450 and haven't seen any adverse Evans and has pretty good efficacy in the mouse colon cancer models.

Thank you.

I would like to make some comments coming from the perspective of someone who has done and is doing human gene therapy for cancer patients. There is always a huge battle within everyone's mind between practicality and, quote, good science. The kind of thing you can do in mice. And I think it's important to discuss that and I was Dr. DeLuca's comments I think very important in that and I have suggestions where you may try to balance those two. Obviously if you are forced too far to the perfect science, the field will never advance and never be able to get into human trials. On the other hand we need some level of safety. It seems to me from just a practical point I accept your explanation about keeping the transsheet in. To rye try to do all the ting in phase I and have to repeat the whole thing with the second virus kill the program which is impossible to do that. That's another problem. We are using old virus by the time we get to humans. It's important to think ahead as you did. My personal opinion is that I don't think it would add significant risk. The immediate response you generate might be good in terms of also being something that would help get rid of the tumor. On the other hand, you aren't or you are not proposing doing primate data which I think is risky since this is a different virus. So I agree that I think the start -- if you aren't going to do primate data that's very expensive to do. The starting dose needs to be lower. Especially based on the data from the European studies where they saw toxicity. If you worried about this no possible benefit in children which I'm not sure is really an issue, then start with the with the lower doses until you feel more comfortable. So my suggestion then would be to try to go down it a low dose, 10 to the six or maybe lower for the first few patients. If you feel you must do that in your adult patients and move forward especially see if you aren't doing with the primate studies. I think these patients are in desperate need. There is nothing to offer. There are three trials with herpes. This is less risky Not injecting it into people's brains. The virus does seem to be more active. May be more toxicity and will have to find that out. I believe it's an important trial to move forward with. If you are not going to do the primate data and the FDA will guide you on that, then I would suggest starting at lower doses.

Can I add?

Yes.

I think the lower dose is a good idea and really does address the issue raised by Dr. DeLuca. And you are not going to get in trouble clearly with the provision that says you need efficacy because you aren't throwing away your high dose which is where you expect to see efficacy. That's not an issue.

I would be happy to do that.

I was going to ask you to given the clarification of the federal R regs that we heard from Dr. Borrow to review the rational from the ten to the searchth starting dose the concern of the group are clear. How would you respond so that we can structure our formal response to you?

Sure. I guess we would be happy to start at a lower dose and actually reviewing and looking at the data on the slide in the experiment when we did use rRp450 which we take more efficacyious and at the time we had limited and only given ten to the fifth or six times ten to the fifth and got significant efficacy. I think in retrospect there is quite good reason to believe that even a lower dose may have potential for benefits.

Thank you. I think we -- doctor?

I would agree with Dr -- about development of a program that involves virus that has components that would be necessary for the first study. Are there preclinical data perhaps in the model that Dr -- asked you about where the virus alone wasn't sufficient with getting psycho Foss md over the top in the immuno+ compromised.

Yes, at the end I'm not sure if I deleted -- inaudible.

I think everything after the black box. In reserve. I will describe it to you but pictures are really better. I will show you some cell data and these data from a group. Doleen cancer model. A panel of five cell lines. In the first lane you can see that's -- the lights there. These are the control survival of the cells by definition at 100%. Psycho Foss md alone the dose used had a minimal effect on the survival of the cells. Psycho veer loan had a minimal effect on the cells. Virus alone reduced cell survival down to the 30 to 40% range and psycho Foss md plus virus reduced it further. Again psycho veer plus virus inhibited the virus effect on the cell line. In addition, they did -- this is the modal that I alluded to. The hepatic metastasis colon cancer model with -- and this is looking at theism of metastasize in the liver, creating 100. Psychos if md, virus alone with PBS control and metastasize were down to 10 to 15 and virus -- virus with psycho Foss md was down to five. The data suggests that the additional of psycho md even with the setting will add to the efficacy.

Do those results depend on the expression of the rRp450 gene? -- p450 gene.

I don't believe they measured that in that paper. But I would have to defer that to data.

Did those results depend on intact immune system so had affected the immune system and took the immune system away did you get the same?

Usually see the same results in the human Ken graph models as well as -- xeno graph models. I would like to see the literature on that.

Am the met tesic model?

Uh-huh. I know they also did pass imings in in administration of antibodies which did not inhibit the result.

Dr.?

One last thing. I do think the latenty questions are important and those are easy experiments to do I would recommend that you do them model like this or your nude mouse model and go back and answer those questions what type of lateensy in the ganglia of things.

Discussion from the public? Yes, sir, could you please give us your name.

Luis -- university of California in Irvin. I want to introduce a concept here which RAC doesn't seem to be thinking of as you move forward with your approval of applied human viruses to people, you need to start worrying about some especially rare but potentially significant events of creating a virus from recome byings in which us unlike we start off. We had experiences with trying to eliminate polio of continue with theed strains picking up sequences from unexpected sources. I think RAC needs to worry about that a little bit with respect to humans harboring a lot of other potential sources of recombination -- recombination.

Thank very much. Thank you very much for your clear presentation and interaction with the RAC.

thank you for the opportunity.

I have taken a stab at this. So let me read what appear to be the remaining concerns. Number one, concerns remain that the mouse model utilized for the preclinical data indicating safety may not be fully valid. I'm trying to make the language appropriately loose. May not be fully fully valid because of the elevated temperature of the mouse attain with theeing recommendation. The immuno+ virus and thus potentially underreporting toxicity in man. Preclinical studies in a nonclinical primate model should be -- I put under take and should be considered is probably more appropriate. So should be considered. Number two, collusion of the p450 gene may not be fully justified in this viral transgene as the virus will life likely establish long lateensy and expression of rRp450 is not -- p450 is not an integral performance of this protocol. We had a lot of of that during the last half hour the discussion really has I think tried to balance using a particular study as a platform for future studies versus the construct of the exact protocol we were reviewing. I am in favor of deleting what I have just read from our concerns and want to know if there is anyone who objects to that.

If you want to, you can say however these concerns are balanced by issues related to practicality and using this as a platform in the future. Both sides of the story there.

Inaudible.

We could focus on the p4 50RBGS we have how many genes in herpes, why are we focussing on the p450 specifically. I think this is really a -- it should be a focus of the study where we actually induce or use it possibly but not necessarily -- just another gene that we have in there.

Dr. DeLuca, want to comment on that?

Sure. I was concerned because it's not normally expressed. Recomeinant transgene. The RAC gene -- there will be a response to herpes before the injection -- inaudible.

Again, concepts of the herpes genes producing immune response. I'm having trouble trying to find a safety issue in that. That's my point.

Although I raised that question, too, I guess I would go along with the issue. We tend to focus on not having genes in being transferred that aren't essential. I think that's been recently justified here.

I just deleted it. Thank you. Now the new number two, given the ongoing concerns with regarding the relevance of the mouse model and observation that ICP6 mutant replicate in the brains of young mice consideration should be given to decreasing the starting dose to less than 10 to the 7th and decrease risk of this vulnerable patient population. Dr.?

Just one thing that we can discuss and I guess mostly I'm wondering what everybody also, what about the notion of selecting -- positives, will that decrease safety because with the young kids you have a potentially --

That's a possibility. Actually, that's what I guess the RAC decided when they reviewed the NV1020 protocol. Was to go forward with seera positives first. That was with young people. And this much then go forward with seera negatives. And I don't -- I don't think that it would have been -- the outcome would have been different if they just when at the same time. The way things worked out. It is a possibility.

But selecting for zero positives when the subject population is so young is a strategy that is likely to exclude almost all of the desired subjects.

It would certainly potentially exclude children below the age of five, possibly.

Yes. And the tumors that are mentioned in this protocol are those which frequently affect very young children. So I think we are getting into a piece of the protocol that really might end up totally reducing it for marginal benefit.

Perhaps reducing the starting dose takes care of it.

I think so.

That's fine. I want to make sure that --

Thank you. Other -- all right and number three, given the on going concerns regarding this about -- begs to Ms. Kwan's comments given the ongoing concerns given potential risk to children the informed consent document should be reviewed for language clarity in order to assure the consent for this vulnerable possibling will be obtained following the complete understanding of potential risks including discomfort. I would like motion for approval.

Couple of others.

Please.

People the animal model -- to examine the extent which the virus established laytensy where the virus established lateensy purposes in okay.

And what animal model, Dr. DeLuca, other --

Think they only used one.

You mean species or particular tumor models? Only one it works.

Well, you used a mouse model where you put -- not model but tumors. Right.

So it's important that that be -- that lateensy be established and I have that as part of point two and I deleted it. What impact will that information have because it's like lie that latency does occur. What impact will that have on the design of this study?

Actually, one of the rationales for doing this is the virus establishes latency very poorly in a mouse model. Now there is a situation where they are providing a whole bowl of susceptible cells.

It would affect risk benefit analysis if you were establishing latency or not. I think it's important to know that.

Okay.

I have writ in number four, determining extent to which the virus established latency in the peripheral nervous system in the mouse --

In the model cancer. If you are not providing those susceptible cells you are not darn.

In the cancer model. To fully assess or fully develop the risk benefit analysis.

Just to clarify that that means if we find that there is 1% or five percent or 100% latency we will move forward but put in the consent document for the data.

That's my understanding. That is why I asked the question. It's important to understand the degree of latency established but not a protocol stopper is the way I --

Is that fair?

Correct other issues for discussion?

In the statement of starting with the lower dose in the first discussion of ten to the fifth and ten to the sixth will we recommend a specific dose.

I did and my second point consideration should be given to decreasing the starting dose to less than ten to the seventh. We could recommend a specific dose. Unless someone can help me with this, I don't personally have the information from the protocol as written to make that kind of recommendation.

And should that recommendation be done in the contact of lack of primate studies or --

Remember our first recommendation states preclinical studies in a nonhuman primate model should be considered and our second recommendation then is, given the ongoing concerns regarding the relevance of the mouse model, consideration should be given to decreasing the dose to less than. So we were nowhere saying that a nonhuman primate preclinical study needs to be done. We are saying consideration should be given to the information which we have available suggests that the starting dose should be less than and we are not prescribing a dose.

One other question about that. One of the reasons we avoided primate studies we don't have a tumor model to emulate the clinical trial for the primate model. What kinds of studies and what would be sufficient.

We can't define that for you. You have to to the fda. Anybody doing cancer therapy doesn't have a tumor model for the primate. There are preclinical studies you could do. Usually say worst case scenario direct intracranial injection, iv injection if you fled do that and that will be determined between you and the fda primarily you can discuss that. But that's the usual -- there are no tumor models in primates. We usually go with the worst case scenario.

All right, may have I -- may I have a motion for approval, please. Dr. DeLuca.

I want to see what people think about the other recommendation that a rescue -- to assure the attain wings solely due to the mutation since this will be precedent setting. And people do and they don't do that and I know if you wanted to publish it in the journal of -- you wouldn't be able to simply for simple esoteric study there if you didn't do that experiment. The reason I think that it could be important is other mutation dozen occur especially when you generate viruses by transvex. And if the attain wings of this virus is in part to another mutation it doesn't give a feel to a clear picture of what the toxicity of the virus is.

I agree. I think that should be done and it's very important issue for a brand-new virus.

It's a fairly simple experiment you can do. You don't have to do a lot of go through and repeat everything.

So given the likelihood that the gene money fieed virus will be utilized in future studies, it is important to assure that the attain wings is solely due to the deletion of L.A.-de-da inserted?

They are going to use this particular virus regardless what the outcome of the study is, the virus going in with and so that would put a couch in the language of please give consideration of this because this will obviously help the failure points and --

Other people may come put in protocols utilizing this or similar approach and use this as precedent. They will have to do this work and find out the virus replicates better.

I guess the question would be to what extent does it rescue it if it doesn't totally rescue it and to what extent does it inhibit moving forward with this protocol even if it's shown to be a safe virus?

I would say none at all.

We don't know all the mechanism of how preed Nissan works -- preedson works.

Then I guess the RAC should consider publicly reviewing every protocol and that's what we would be heading if we don't make sure that the precedence that are establishing are based on sound finds.

We aren't asking to define the mechanism of how the virus works. Only to prove that's the gene that is responsible for the attain wings.

Okay. Just give me a second. People want me to reread what I written? Dr. Powers says no. I will take no. I'm going to go over this. We have changed it so morn much that I think it's important before we vote. One, concerns remain that the mouse model utilized for the preclinical data indicating safety not be fully valid because of the elevated temperature of the mouse. Atten with theeing recommendation of the mutant virus and underreporting toxicity of man and nonhuman primate model should be considered. Two, given the ongoing concerns regarding the relevance of the mouse model and the observation that icp6 mutants replicate in the brains of young mice consideration should be given to decreasing the starting dose to less than 10 to the seventh in order to decrease risk to less vulnerable patient population. Three, given the likelihood the gene modified virus will be used in the future crrtion should be given to the generation of rescue version of rRp450 to be given that is generated to assure that the attain wings is solely due to the deletion of the icp6 coding sequence. Four, given the ongoing concerns regarding potential risk to children and that's the informed consent piece we haven't changed that. And five, to determine the it which the virus establishes latency in the peripheral nervous system in a mouse in a cancer model to fully develop the risk benefit analysis. I will have to beginning wording of that and I will do that. Given those five comments, could I have a motion for apofl?

Second.

Thank you, both. Let's take a vote then. Thank you all. We will take a break. And we will return in 30 minutes. RP RP

Our next -- wait, conflict of interestwe have one investigator, Dr. He is lip in conflictwith the next study we're reviewing, 699. again, he is a member of a DSMB.First study being conducted at Cincinnati Children'sHospital.She is also a consultant to St. Jude children's researchhospital and therefore is excused for the discussion ofthis study.Protocol is entitled a pilot study of Temozolomide andBenzylguanine for the treatment of high rate gliomagenetically modified for chemoprotection.And two investigators will be presenting.We're going to start with Dr. Lars Wagner. Good morning.I wanted to thank Dr. WARA and members of the RAC forallowing us to participate and for your reviews.This will be a two-part study.We'll be discussing some of the clinical aspects of thetrial and then the study's sponsor Dr. Von Kalle will bediscussing some of the safety issues such as insertionalmutagenesis.I wanted to start by providing some clinical backgroundregarding the disease we'll be treating.Brain tumors are the leading cause of death in childrenand high gried gliomas are clearly associated with apoor prognosis such that the five year progression forsurvival for kids with this disease is about 5% and themedian overall survival ranges from -- it's more he thatwill in adults where the five year survival is about 5%.So this is clearly a disease type and a patientpopulation that is in need of more effective therapies.Historically, patients with high grade glioma have beentreated with a multimodality approach that involves thesurgical removal of the tumor.This is important but not expected to be curativebecause of extensive microscopic infiltration of theFocal radiation is then provided in this maneuverprolongs disease free survival but doesn't really impacttoo much on overall outcome and so clinicians have oftenoffered additional chemo therapy following radiationprimarily in the past using nitro agents.This may modestly improve survival perhaps 5 to 10percentage points but in general is not curative andtreatment with myotro surrrhea -- requiring delays indose reductions and nonlife threatening pulmoto logicbut really the main problem is tumor cell resistance tothe chemo therapy agents that we've used.One of the big developments in the field ofneuro-oncology the past decades has been the use of has a nonto profile and soin a large phase 3 trial that was recently reported theuse of Temozolomide used after surgery concurrent withand then following radiation significantly improvedsurvival for adults glioblastoma compared withtreatment of radiation alone.And so this type of treatment with Temozolomide duringand after radiation is now kind of evolving as the newstandard of care at least for adults with glioblastomabased on the limited to bes sisty and the use of oralThis approach is currently being evaluated in pediatricpatients through an oncology group trial but similarresults are expected and while this is an encouragingdevelopment in the field, still with a two-year overallsurvival rate of 26% in adults indicates Temozolomide isreally going to expected to be curative for themajority of newly diagnosed patients.And the primary problem of course is tumor cellresistance.We know that MGM T as the DNA repair protein that's theprimary repair system and MGMT is set to remove addictsby either Temozolomide or conventional agents.And this is a scheme attic representation in which thechemo therapy agent here nitrous yourrrhea deposits thegroup on the O 6 position of DNA and then that can beremoved in a one to one reaction by MGMT.And MGMT really appears as a relevant therapeutic target it's expressed in about 90% of Peted pediatricgliomas and about 90% of adult gliomas and theimpression to Temozolomide in patients withhigh gried gliomas and so the more the MGMT expressionthe less response is scene.And now there's Benzylguanine which effectivelyincreases the sensitivity of tumor cells toTemozolomide.So that preclinical been used to lead onto clinical trials, phase one trials of Temozolomide and 6 Benzylguanine both in kids and adults.And unfortunately, we've seen some exacerbation ofhematologic to bes sisty and this is thought to be due inactivation of low but still protective levels ofthe hematic skrem cells and this is required a dosereduction of over 50% for Temozolomide in phase 1 trialssuch that the usual dose has to be reduced to a maximumtolerated dose of 75 milligrams per meters squared perday.In a similar in adults showed a requirementofMaybe the optization of this Temozolomide BG approach require a method to help circumvebt the hematologicto bes sisty.And one of the developments that's helped push thisforward has been the creation of the mutant form aproline has been substituted for LYSONEcreate add form of the enzyme which is resistant to bothTemozolomide and O 6 BG and so this kre creates a way toof to overcome toxicty and if you get repeateddoses of this combination after transduction then youcan kill the nontransduced hematic transduced cells andthere by increasing the relative percentage of protectedcells and the so called enrichment effect actuallycausing less myolow suppression with further therapy andthis is in sharp contrast with what we see withconventional where the myolow suppression gets worsewith each course.So this overcomes initial transduction efficiency bytransducing out the selected cell population withselected treatment.This may allow for increasing doses of Temozolomidewhich could lead to superior antitumor effects at leastin a clinical model.So to briefly summarize about what we know about thepreclinical evidence of this type of approach using theMGMT P 140 K mutant, we've seen that protection frrchemo therapy to beicty is seen both with mouse andhuman stem cells and this leads to improvement in the counts with continued therapy and durablechemoprotection is observed even when doing serialtransplants into another mouse that the mice are able totolerate higher dose intensity and this translates intoa signe graft model into antitu mar activity so nicelyproving the principle at least preclinically.And this leads to multiprotection enrichment in a modelin which the percent of marked cells went from 15% allthe way up to 90%.This also in the Al yes nays setting increased from 95%and highlights the potential approach of this Al geneticsetting.You take the transgene and put it into the stem cells,up matly administering it to the patient and here wehave just a sampling of what's in the stem cellcompartment with a few transduced cells in thebackground of many nontransduced cells but withtreatment of Temozolomide and BG then you can kill offthe nonprotected cell tells and then causing enrichmentand expacks of the chemo protected cells and this canlead to multilineage enhancement here.So our clinical trial is set up in the following way.We chose to do a pilot study of ten patients and wechose ten patients because this seemed to be the minimumnumber of patients that we hoped would accomplish allthe study objectives here and this number of patients issimilar to what has been seen in other pilot studies.And so the main study objectives is really to the feasibility and the safety of genetra*bs fer and the proposed chemo therapy regimen andthen also to assess the efficiency of gene transfer andthe durability of the MGMT expression to ashes thedegree of chemo resistance and ability to dointrapatient dosing of Temozolomide and so this will beone of the hallmarks of success in this clinical trialis the ability to -- whether this can meaningfullytranslate to a dose increase in the Temozolomide.And so we'll be carefully dose escalating patients in anintrapatient fashion using extent of the transunitexpression as one of the criteria for dose escalationand then responses in the confines of a very small pilotstudy.I'm just noting these responses, but of course thisnot going to be a demonstration of the efficacy of thisapproach.The population we tried to be very careful aboutwho we would include and exclude.We chose children five years of age and older becauseyounger children with brain tumors are often treatedwith different radio therapy approaches and we wanted auniform population with which to provide treatment to.We chose new patients instead of relapse patientsbecause we wanted patients to have an expected survivalat least for the length of the entire study and then type patients adequate organ function. excluded patients that may have either too good of aprognosis or too poor of a prognosis.So may not be appropriate for this experimental type ofapproach.Similarly shs patients with this type of histology eventhough they're high grade gliomas do better withconventional therapy and would not be appropriate forenrollment on this trial.Patients with extracranial disease may not have theexpected life expectancy to last for the duration ofthis trial.The study is designed into three blocks, the first ofwhich we'll term the standard of care block and again,this experimental trial is really piggybacked on to whatis currently available treatment for patients with newlydiagnosed high grade glioma so this will includeresection followed by radiation and kind of a standarddose of concurrent Temozolomide.The second part will be the gene transfer part wherepatients receive two days' wo*t of low dose bu sul feignto create space and allow for transfer.This is not a myolow aplaytive regimen.This is to create space and this is a drug and thesedoses that has been used in other gene therapy trialsfor this purpose.And the third part will be the more experimentalpart where we'll try to achieve dose escalation and so will receive chemo therapy with six monthlycourses with intrapatient dose escalation as toleratedif they meet certain criteria.The study scheme looks like this.Patients undergo surgery.We collect stem cells by aforree sis.In the lab these undergo cell collection andtransduction and administration to the patient at theappropriate time.Meanwhile after stem cell collection patients undergoradiation.We have MRIs KIND OF BUILT IN to make sure thesepatients are not clinically progressing and of course not receive further therapy.Block 2 is low dose bul sul fan.We expect recount recovery to be about tloo to fourweeks and proceed on to block three which is the sixcourses with attempts at dose escalation in courses 2through 6.So this intrapatient dose escalation will be dependenton the absence of dose limiting toxicty as defined inthe protocol in previous courses as well as some evidence of presence of the transgene andperipheral neutrophils and so we describe the definitionof a marked A and C or absolute neutrophil count and wecan get some estimate that these patients may indeedhave at least some degree of chemoprotection.And then if they meet the criteria we would increase thedose.This was the was reached in a mediate tricktrial without any additional stem cell support so we think based on the previous phase 1 trial thatthis dose combined with the O 6 would be a safestarting point and we'll try to increase by 30 to 35%with subsequent dose levels.If a patient was able to make it all the way through bycourse 6 they would receive almost four times as much as they would have been able to get justbased on the conventional phase 1 trial.So the study end points involve safety and of coursewe'll be assessing toxicty involving standard NIHcriteria and then throughout the protocol and thetreatment will be monitoring blood and bone marrow fromsoxicty and we'll describe that a little more later. be looking at the extent of gene transfer of realtime PCR.The efficacy of gene transfer is assessed by MGMT and then from a functional standpoint theability to actually dose escalate and give higher dosesof Temozolomide in block three and then looking at theantitumor activity with physical exams and MRIs so thistherapy is not without potential toxicty and Dr. Kallewill be reviewing this in a little more detail but thereare efforts in protocol to help address theseconcerns, so looking for secondary leukemia frominsertional mutagenesis and built in are plannedevaluations of the blood and bone marrow forconventionalmorph logical and then clialty and the useof the land PCR for determination of monoclonalinsertion sites will be also evaluating for developmentof replication retrovirus and severe acute infusionrelated toxicty from the product.And then another potential concern is delayedengraftment.Patients will have in the freezer a backup of stem cells that can be given in thissituation and for the first three of theseconsiderations there is a stock in place if any of thesefirst three toxicties occur and they'll be evaluatedbefore the study can continue.And just to highlight some potential clinicalapplications.What we're trying to do in this trial and this could bealso applicable for a disease like metastatic melanoma,modifying the Al yes nate -- leukemia graft is alsoanother avenue of integration and then transfer bysingle disorders and this had been demonstrated bythe preclinical model.Soy eel turn it over to Dr. Kalle who will be talkingabout the laboratory monitoring and some of the othersafety issues.

Can I ask one question on the trial?Are giving one dose of the stem cells?Are you escalating that and how did you choose that dosein your clinical protocol?

So there will be just a single administration of thestem cells.It is somewhat dependent on the amount of cells that arecollected.There will be -- for every patient a backup kept as anunmanipulated frozen Al kwaun in the case of delayedtransfer.We're setting some limitations on the range of stemcells that will be transduced and administered into the but going to be hard to prescribe a -- acertain dose for patients because that's going to besomewhat dependent on the number of stem cells

I just quickly wanted to go and summarize some of thepreclinical as well as followup experience that we havewith regards to the gene transfer of this transgene.This is very quickly a summary of some of the mouse data with MGMT as one of the examples where youcan see primary recipient after two cycles of selectionand you can see the gene marking actually stays present.Our group in Seattle as has already been mentioned hasdone the same in the canine model where you can see thatcourse rs of selection raised the number of marked cellsin the frifly all the way up to almost 100% in thismodel and some information that you can't see on thisslide is also that selections at intensities thatproduced neutropenia down here have no longer producedneutropenia or some cytopenia at the later stages ofselections.Here's the date that Lars already mentioned with thedonor timerism that adult stem cell behavior could bemodified in that the recovery was returned to greaterthan 95% donor just with the selection process alonewithout any T-cells which would be a majortherapeutic accomplishment in hall yes nays Joantransfer.I wanted to use the opportunity as I've been asked tosummarize some of the insertion sites that have been inthis trial.Of course it's in principle a dose finding to these genetherapy vectors and gives information about stem cellinsertion site distribution.The cell type dependence of such insertions.Genomic side effects if there are any and behavior ofparticular vector types and I've already talked to thisgroup in March about this so I will quickly go over thisas I've shown before.We have developed method dolling that can identifydifferent insertion sites even in a complex group andallows us to have sequencing information about thesequence and have a more moleccue lar mark for eachinvolved clone.You're also aware of the clinical trials both fromDr. Fishers and the group in Paris as well as fromDr. Thrash's group in London where again the C functionis replaced by a retrovirus vector and I had shown dataearlier in that trial we can see a very pollicclone recovery in the CD 3 component.But we can basically identify by sequencing these thatsome the cells that have been corrected are actuallyproduced both granule sites.Macrophages and lymphosites meaning that it couldqualifyChronology in the cases the first to leukemia sideeffects in the fr*efrng trial and you have heard and hada discussion this morning about the third case, thatfrom the moleccue lar mechanism is actually fairlysimilar and we have gone on to do a more in-depth of post transplantation home porsis and looked at the insertion sites.It has been published and we have also found evidence inthe clinical model and again, I already talked aboutthat in March that the distribution isn't completelyhome general yous and the cross of chromosomes isto the genes than the genetic nfks as such.There's also a bias after retrovirus insertion so theinsertions actually sit closer to the start site.If we look by gene ontology analysis of where theseinsertions occur you can basically see the genes withkinase activities transforrace activities and someothers in these trials have a statisticallysignificantly increased likelihood of becoming hits,probably mainly because such genes are expressed in thetarget cell population at the time of retrovirus entry.So summary of what we that 60 -- about 60%of insertions occur within or close to genes and ofthese 20, 50% of the first 20 sit around the front ofthe gene.This clustering around some common insertion sites and about that in a minute and there are before and after and some of theseinsertions have some influence on the biology of the cell clones and then there's also differencesbetween trials in that severe adverse events have onlybe observed in one of the trials far and it will beinteresting see what -- what happens with the othertrials.But also we have to keep in mind that in summary, allthe immune know dep efficiency trials, the ADA trial andthe German trial, so far have treated 27 patients.These are the French, British, Italian combined.26 of the 27 are alive and the vast majority of themThe successful correction has been described in theliterature in 24 out of 27 from these patients. so far had three SAEs reported and two of the threepatients affects by this have survived the SAEs and arestill using the immunity of the successfully correctedcells basically for their day-to-day immunity, andagain, there seem to be some very significantdifferences between the different types of trials, theADA and the SCID wanted to close with summarizingsome of the more recent experience that we have had withthe -- our colleagues who have conducted these KTDtrials.These are the trials of the German and Swiss groups.The principal investigator in frank further, we havefollowup of over one year and close to one year in thefirst two patients.This trial is remarkable because more than 20% of myolowpor ree sis has been corrected in both patients byretrovirus vector so this not -- basically anisolation to mile porosis.You also see that there is a more regular pattern ofthe LAM analysis toward the second half of theobservation period.We have looked into this more closely.You can see here as a red line the marking level or thepresence of transgene in these patients that hasincreased in long-term he ma toe paresis that it has upto about 50% peripheral blood cells and we couldidentify that certain insertion sites basically -- wenow think selected by growth are doing the course ofthis process before this whole process plateaus out andwe have followup of more than a half a year with and it seems to be a stable process.The second patient shows a very similar picture ofstarting out around 15% and plateauing at around 50% ofcorrection in sites.One of the most frequent hits, we basically found threehits, two of which we're working on the moll cueevidence.You see the insertions on the first and the second also.This is a trial where very many corrected cells havebeen reinfused and 1.5 to 2 insertions had occurred percorrected cell and you can see that there is very strongclustering of insertions around certain locations evenwithin the gene that allows the activation of this geneand then facilitates the expansion of these affectedclones.We can then show by long-term followup that thatbasically -- the plateauing is also happening at thelevel of each single clone.You can see that there's basically, if you so wish anexmro ra*r ration of the -- and that then stabilizes outover the course of time.You have to keep in mind that none of these patientshave a serious event or anything similar by thedefinition of the trial and that the leukosite numbersin these numbers not elevated and show no signs ofincreasing.I wanted to put in as a reminder also here that the bothanalogous and ADA trials that we've been following from on has a contribution pattern that is biologicallystable over the course of time.So in summary, on random insertion had some veryobviously positive results in these patients in thetrial we've seen unprecedented levels of correction evenprior to the expansion of the cell -- of the cells and Ididn't have the time to go into this, but both patientshave substantially cleared their preexisting infectionso there seems to be a strong therapeutic benefit fromthe trial.These are likely single and double copy insertions oftheThe cells seem to be continued to be regulated in growthand some of the clones tend to disappear over time so wethink that many of these may actually be burning outover time.We do not know what the expression pattern of thesegenes in long-term he ma to forree sis is so we do notknow what the normal physiology of these genes or somechanges that have to do more with regulation of apoptosis and regulation of operation.We have likely identified regulators and none of thesecell types also not in animal models where this findingcan be closely reproduced.This also shows that prospective screening by the methoddolling we apply is possible and so if we look at therisk benefits of these clinical applications as I saidearlier, number of modified stem cells over thecourse of time that these trials have managed to put infor a single patient have reached very high numbers.We have started to see therapeutic efficacy, not --mostly based on the increased efficacy and of coursethis is the therapeutic window that will also definewhen we will see potential side effects of thesemeasures.So I would like to conclude that stem cells can begenetically modified and effectively selected by what weare proposing.The side effects are clearly based on gene activationsand so far there's one major mechanism that has beenidentified.Determination of stem cell contributions at the clonelevel offers insights and gene transfer continues tooffer unique theer putic opportunities.With that would like to close and thank thecollaborators in Cincinnati who have made the definitionof this possible and we'll be happy to take yourquestions.

Thank you very much not only for presenting for thisparticular study but for updating us once again.We should have had you speak before our first discussionthis morning.I'd like you both, Dr. Wagner, perhaps you could go tothe podium as well, and we're going to open discussion.We'll do that by moving from one to the other of theformal reviewers for the protocol and asking you tointeract directly with them and then we'll open yourprotocol for broader discussion.So our first -- I have owlOur first discusser, Dr. Rosenberg.Is that all right?Great.

Thank you very much for your presentations and alsofor your thoughtful responses to our reviews.Most of my particular questions, many of which have beenanswered, so many of my questions have been answered andmost of my questions know kused on the retroviralaspects of the protocol.So it will be those that I'll cover.And I'll go through them in part for the record, I dohave a couple of remaining questions.Several of my questions related to the efficacy withwhich transduction was going to be improved inyour -- in your protocol, and you certainly haveresponded that you estimate 20 to 40% of the cells to becarrying an insertion and probably about 1 insertionwere cell.Although I didn't directly ask it in my questions, I'mwondering if you could give some feeling for how manycells or the range, I realize you don't knowspecifically will be infused into each patient and myreason for asking is what's the chance that eachindividual will receive at least one cell that has anintegration more or less gene?

The amount that we're -- that we're -- isbased on -- and then based on the outcome of thosecultures that usually are between 50% and 150% of thatnumber be reinfused.That, of course, means that there are basically millionsof cells per patient that can potentially carryinsertions and so as we have described earlier, thestatistical likelihood of integrations in the vicinityof active genes is basically given in these patientslike for all of the other trials.

Right. just wanted to make sure that was -- that was clear.One thing that you included, but if you couldjust comment on that now, you are using cyber neck tincomponent to enhance your infection and I realize you'veadded references to that, but since that didn't come upin the discussion, if you could briefly comment on howthat actually works.

Yes.This actually a recum bant and fiber neck tincomponent that allows to coat the plastic of the culturevessel and basically contains binding sites both for theretrovirus envelopes as well as for the target cellpopulation.Now current understanding produces a vicinity betweenthe two.It also seems to be the case that long-term survival ofrepopulating cell is better. is something that has been described as early as1995 and '96 and has been a standard for transductionprotocols that in itself is not a test question forthis trial.

My next series of questions relate really to thenature of the CD 34 positive cells that are to be usedand I wonder if you could comment a bit more clearly, Iknow we -- everyone knows perhaps less than might beideal concerning the nature of the populations that areused and how those populations may vary from patient topatient or trial to trial, and you've certainly pointthat out in your response, but I'm wondering if youcould comment a bit more on steps that you will betaking to try to increase information to the field as tothe specific types of cells that are infused into thepatient beyond what we Okay.This is an interesting question.The validity of the -- especially CD 34 positivesubpopulations for long-term recovery has not beenclearly established.Of course lineage negative cells are supposed to be theones that carry short and long-term repopulationalthough some lineage markers may be on short termrepopulating cells so our analysis on which cells haveactually been genetically modified will focus more onthe post transplantation and on the analysis of thedifferent markings that we have seen or demonstrated inthe SCID trial we can identify the relationship betweenIt will be very difficult because basically the analysisof a given cell in the pretransplantation material willuse up that cell for the purpose of transplantation.It will be very difficult to do studies that try topredict behavior from the pretransplantation immunotype.

I think several of us asked questions concerning --oh, wait.I left one thing out and I -- we did ask about thepotential autogenic potential of the transgenes to beused and you comment that -- that there's no knownpotential for that and there were also questions raisedwhich I believe you've addressed concerning theconcerning the length of time it would take to obtaindata from your patients and how that data might be usedto the trial.And I believe you've answered those questions.My question related to the use of the particularenvelope gene that you had chosen if your virus and Irealize it certainly is a highly efficacious one whichyou use to point out.I'm wondering if there's anything that you can saybeyond the efficacy that might lead us to wonder ifdifferent populations or tell us anything about whetherdifferent populations of progenitors may be infectedwhen this envelope is used as opposed to other envelopesthat are used with other retroviral vectors,particularlily I'm thinking of the difference betweenthe trial -- the fisher trial and the Thrasher trialwhich I believe use different envelopes.

Yes.And that is an interesting point, of course, and a very question.The differences between the slasher and the fisher trialwere that the Paris trial was used in an tlo toepicenvelope where as Thrasher's trial was using thewhich the same envelope that we're using.I think the agreement is that the efficacy especially inthe immature long-term repopulating -- so we wouldexpect to see a more efficient gene transfer into, ifyou so wish, he ma to poetic cells.

Thank you very much.

Dr. Rosenberg, do you have any remaining concerns?

I do not think so. Because I heard a discussion suggesting that all of carefully written concerns have been addressed.

I believe that that is true.I think from all of this, the one thing that -- and itdoesn't speak specifically to what we've heard aboutthis, really relates to ways to understand better whichinfected cells really contribute or have the potentialto contribute to problems in particular patients.However, believe those concerns are addressed withrespect to this specific trial.

Thank you.Dr. Nem ro?

So I -- I really was impressoed by the thorareness ofthe trial and the multiple stopping points that had beenbuilt into protect patients and I thought it was a verynovel but a complex trial that merited furtherdiscussion and I had many of the same questions thatDr. Rosenberg raised.And I'd like to maybe follow up on a couple of those anda couple ones as well.So it's mainly clarifications rather than concerns fortheSo I wondered during the collection of the CD 34positive cells, there's also likely to contaminationwith CD 34 negative cell populations.Will you be looking at which, you know, the -- thepresence of those cells and also is there a concern thatthere may be contaminating glee mall sales or cancercells that would also be potentially transduced andcreate a problem for treatment?

I think Lars Wagner can answer that question becausehe has just undertaken a review of the literature withrespect to the presence of such cells in the bonemarrow.

Yeah.That is a theoretical concern but should be exceedinglyrare.There have been multiple studies, the use of autologousstem cells before and this has not been found to be aclinically significant problem.There's never been a reported case of brain tumor cellsbeing contaminating the stem cell product and thenhaving complications from that.The incidence of extracranial me ta*s cease at the timeof diagnosis is exceedingly rare with high grade gliomaso it's probably less than 1% of patients will haveextracranial metastasis so it's probably more uncommon the bone marrow would be involved.In a review of the literature, there have been 13 casesof bone marrow involvement with glee owe blastoma, onlyone of which was at the time of diagnosis and about 85%of these cases would not have been eligible for thestudy because of hematologic problems so hopefully thiswill be an exceedingly rare problem that we will notencounter.

Thank you much.I -- one of the other questions -- clarification was toto indicate the decision of using bio safety level 1versus bio safety level 2 containment in theinvestigator was -- in his written response was able toclarify the decision to use bio safety level 1 and I'm that.One of the things itches interested in was thatwe've been discussing the -- the problem with MLVvectors and the SCID trials in the last several monthsand because of the experience of the investigator, I'mwondering was any thought given to different retroviralvectors in designing these kinds of trials that perhapsmight have less or either built in regulatory sequencesor perhaps less propensity to insert in the five primeregion or near the five prime regions of -- of indodgenows

That is of course an ongoing thought process and aswe go forward with these trials.However, I think we're still at the stage where we'retrying to understand what the actual distribution is andwhat the influence on the biological fate of these cellsexactly is.So in terms of the feasibility of putting these thingsand also the safety of putting these things forward intoclinical trials, I think a lot of research, preclinicalresearch has still to be conducted.We have to find out whether the newly envisioned vectorsare actually as effective and as safe even as the olderones.And so I think what we're -- we're intending to use hereis if you so wish, a state of the art retrovirus vectorand incorporating our current knowledge.

Thank you.So another question that I had really was more of aclarification.In the written response, you indicated that 20 to 40% ofthe CD 34 population is transduced.But in another look, the appendix M in the proposal,there's an indication that there's 35 to 64% that areactually transduced with a vector having the MGMT gene.So I'm wondering which of those is the more accurateassessment of the transduction?

Okay.The that we stated with regards to what we'reexpecting from the trial is incorporating our experiencewith the scaleout to the clinical dimension of the data is what has been done on a smaller and as you be aware, we have an active genecorrection trial for anemia and have some experiencewith the upscaling now of similar types of vectors tothe procedure and so those are the numbers that we'reexpecting to see in CD 34 cells.

Okay.So one of the other questions came up also in thewritten response and you indicated that there a --one of the significant differences between the Britishand French trials in the transduction was using serumfree transduction.I wondered if you could clarify what that meant.

Okay.

What that's related to.

Yes.The media that are actually used to culture the cellsexVIVO contains animal ser ra in one case and contain noanimal products in the case of serum pregene transfer.The serum containing media are tlauth to induce moreproliferation and also probably more differentiation inthe cell populations and the -- the thinking is that theserum free conditions maintain a more immature state anda less proliferative state of the target cellpopulation.

And that's related to the fiber knack tin that's inthe serum probably?

No, these are probably growth factor come poebts andother components.The fiber neck tin I wouldn't think because the fiberneck tin is also a component of the serum free cultures.

Okay.Let's see.So I had several other questions that were addressed inthe written response, and one of which was to look forrearrangement of the vector in CD 34 positive cells andthe invest sga torr indicated that was technicallydifficult to do, but it has been looked at in the healcells.The other question I had was whether they have looked atgerm line transmission in their mouse studies andalthough they haven't -- the written response indicatedthat that hadn't been done, that other studies supportedthe -- the likelihood that that's not a major -- goingto be a major problem.I also had a question on the fiber neck tin questionthat was already addressed and then a minor comment onthe -- what the investigator meant by normal variant ofthe MGMT gene, which is the -- I gather is while time.So that's all the questions I have.

So Dr. Nemerow, as with Dr. Rosenberg, I amlistening.I hear that your written questions have all beenanswered?

Yes.

Is there anything remaining?

The only remaining suggestion would be that a carefulcharacterization of the CD 34 population that's going tobe used, you know, whether there -- what -- to thedegree that those are contaminated with CD 4 negativecells and what are they, using the state of the arts itreally those

Thank you very much.Why don't we go on Dr.

Well, I also wanted to thank the investigators for avery thoughtful and carefully designed protocol and verythorough responses to our questions.I raised a number of concerns which really has beenaddressed but I'm going to read them as a formality toget them in the record and I have one issue I'd like tohear a little more discussion of.My first question had to do with the relationshipbetween the analysis for cloalty of vector insertionsites and how that might affect subsequent enrollment onthe trials and findings were made and that wasvery well discussed by investigators.I raised a question about potential survival advantageof clones of transduced cells because of the repeatedcycles to chemo therapy and they have a very response for that.I asked about the potential situation of engraftment notoccurring with the subjects be able to tolerate the --the basic doses of TM, the 75 milligrams meters squaredand that was qualified as well tolerated in otherstudies.I had a question about this theoretical adverseconsequences of fiber neck tin and developing fiber necktin antibodies from the transvection process and thatwas clarified.I had a minor concern about the consent form, which wasaddressed, asked for clarification of the assent processfor the child participants which was also nicelyaddressed and asked the question about financialarrangements with the manufacturers which was addressed.My remaining question is really a question fordiscussion and it follows up -- it's not -- it's both aquestion but for your protocol but also relates back toan ongoing discussion on the RAC about bio statisticalmethod dolling of gene transfer.My question really had to do with the selection of theten subjects for this trial and sort of what theinvestigators hoped to gain from the study.And the investigators very rightly pointed out that thisis a novel design in separate ways.They're introducing the gene transfer and then seeingthat that enables the subjects to tolerate higher than doses of chemo therapy.And I -- the investigators very nicely laid out how theten subjects enabled them to detect serious add verconsequences at very true toxicty rates and then theyhad very interesting statement in the next section, thisis on page 35, where they say that this is not a typicaldose escalation study, and they intend to -- at one they say they want to demonstrate prove ofprinciple and then the primary goals to demonstratesafety and efficacy but they then use the typical samplesize of ten and in a sense, this is a comment thatDr. Demets has made at numerous occasions at thesemeetings and perhaps I'm sort of giving a preamblehoping he will comment, but I think the real issue issort of what -- you know, why ten subjects for the proof principle rather than ap different number, and whatinformation -- sort of what results will be to whatconclusions and inferences for the design presumably ofphase 2 trial.For example, one might hypothetically say that there isa question of how rapidly or how far you could use thedose escalation for the combined chemo therapy afterengraftment and so is that escalation scheme is thenumber ten, you know, sufficient to address any you might want to have for planning your nextstudies but as I said, this is really an issue that's amuch broader issue that we have tried to deal with andso I'm really using your protocols and examples for usto try to think through these issues.

Dr. Wagner, would you like to comment beforehopefully Dr. Dements will weigh in?

Sure.I appreciate your comments, Dr. Lo.You know, as I mentioned earlier, we had chosenarbitrarily somewhat, the number of ten because wethought that that may be the minimum number of parentsin which we could get at least a fair estimate of thesafety of this drug combination.One of the issues that we not discus but potentiallycould be seen is nonheme to logic toxicty as we go tohigher levels of Temozolomide and BG.This is a unexpected thing that could develop and so wewanted to keep the numbers down to a reasonable degreebut still enough to adequately kind of test the -- orpilot this type of approach.And these were numbers that have been done although ourtrial is novel in many respects these are kind of thenumbers that had been done in similar gene transfertrials as well.

So the question for Dr. Dements is in a study whichis not truly a phase one study but rather as a proof ofprinciple study, how does one determine the number ofsubjects to be enrolled?

Well, I guess the simple answer is I don't know forsure.But I think there's -- in -- in your protocol in section11.1 you lay out a rational argument for if -- you know,we don't know what the toxicty is but if it were in theneighborhood of 10% -- if that were true by ten patientsyou would have an 8% chance of seeing a problem of thekind that you identified as the three or four.So that's a good rational for ten.I Dr. Lo's trouble is with the next section thatsays everybody else does it.That's not what I would want to read in a section thatsays why -- why ten patients.I think you have part of the answer there in yourprevious section.You just don't describe it.But then -- you don't describe it in that section butthen you don't go on to address why the ten patientswill address or not address some of the other questions have raised.Pilot may mean feasibility as a concept and o I supposemaybe one is good enough.I don't know but that argument is not laid outin here but do have one very good argument about thetoxicty you don't take advantage of it in the nextsection and that's style and presentation but there is agood argument there.I don't think it's acceptable to say everybody else doesit.I don't know what the references are but that's not agood argument.I guess we've been struggling with generally given thequestion you're trying to achieve, what's the rationalfor the number of patients you need to meet thatobjective and we've been struggling with that for thepast few years and -- and I think we're -- the answersare not easy.We know that, but never the less, we're putting subjectsinto studies and we should have a reason for why ten notfive.Why ten, not 20.We don't know.

So restating it, then, and we'll put it differentlyin the summary comments, but restating it, what we'vereally been struggling with, I think, is how todetermine appropriate sample sizes for early and proofof principle may not be the right term, but early firstnontraditional phase 1 studies and how do we do that.So that we don't unintentionally increase risk to alarger than necessary group of subjects.Is that -- and that's what I understand the study groupwill be struggling with over the last -- next six monthsor so and that you, Dr. Dements are going to take a leadrole in bringing that answer back to us.Is that true?

Well, I -- I will work on it.I don't know about bringing you answers, but I will -- Imean, just to amplify that a little bit, I don't want totake away from this protocol's time but these arequestions that the answers are not easy to come up withbecause we're in a new era and all the things that we'velearned to do traditionally may or may not apply andwe've seen sometimes the application of appropriatedesigns and sometimes they work but it really is goingto require status tigss my colleagues of mine andworking with clinical researchers to figure this out butit's not going to be a simple answer that six monthsfrom now I'll have a five-page solution.

No, no.

Yeah.

I wanted to add some comments to what they just saidthat there are actually several questions that are oftenbeing asked in a phase one trial and safety is certainlyone the reason that your section on safety is a verystrong argument for how ten subjects enables you to havea very strong power to detect the kinds of adverseaffects you're at.But then phase one are also in a sense preparatory totraditional studies traditionally the question iswhat's the maximum tolerated dose.That's not necessarily your concern here but then theancillary question is assuming that this study, a phaseone study demonstrates safety, has it also addressed theother questions that need to be answered in order toplan the next study that would be a phase 2 or phase 3?And that's the -- in a sense, the harder question toanswer that we struggled with as we have investigatorsdoing these kind of innovative promising studies.

Ms. Kwan?

I just wonder if there are specific bio status tigssworking with your team that might want to contribute toDr. Dements' group.

There is.A doctor at our institution is a study bio statisticianand we could encourage his participation to furthercomment on this.

Thank you very much.That's really what we were driving at, Ms. Kwan.

Are there any other questions for the investigatorsfrom members of the RAC?Yes.Dr. DeLuca?

I just have one question for Dr. Von Kalle.You an interesting bit of data on there that Ididn't fully appreciate before.And that is the two subjects that where the leukemia wasresolved, they remain corrected.Are they -- are they -- did they shift to ap differentclone or are they now poly cloneAL or --

Yes.It has somewhat lesser amount of clones than it hasbefore.The patient who was transplanted was unfortunately 2 onethat relapsed and died and both of the other ones stillhave their own T-cells and the number of clones has gonedown but it's still poly cloned.

Other questions or comments? -- Dr. Paterson has a question.

This is a question that was raised by one of ourreviewers but just for the completion of the record andalso because I think that your solution to the issue isinteresting and would be helpful to the could youaddress your variant of the WPRE vector element?

Yes.Basically there has been some discussion although finaldata isn't out yet that in a model in -- of feebleinjection model WPRE variant may be involved ingenerating liver tumors and there is a question ofwhether the exprotein that is part of the WPRE elementand that does not need to be expressed is a part of thatmechanism and while the answer is still pending on thatentire issue whether all there is -- what element ofthe WPRE element is contributing to that, we havebasically used a code on the variant of the WPRE elementthat effectively blocks transcription from that part ofthe sequence.

Thank you.Comments from the public?Questions?All right.If not, thank you both very much and could I have -- oh,what have I written?I have two modest comments left.One, recommend a careful characterization of the CD 34positive cell population to be gene modified and infusedin order to identify CD 34 negative cells, contaminatingthe product and potentially contributing to increasedrisk for the patient.Two, please reconsider the stated sample size of ten toassure that this sample size is the most appropriate forthis proof of principle study.Would anyone like to add anything additional or modifymy suggestions?If not, may I hear a motion for approval?

I move.

Thank you.A second?

Second.

Thank you.Dr. Vile?

Yes.Webber

Yes.

Powers?

Yes.

Bark lee?

Yes.

Rosenberg.

Yes.

WarA yes.Dew Hurst?

Yes.

Ms. Kwan?

Yoo*i.

Al bell da and Nemerow?On that note I think we're all very hungry.We will take a break and please return promptly at 5minutes to 2:00.Thank you all very much. TESTING CAPTIONS. TESTING. TESTING CAPTIONS. DEWHURST SIMARI EPSTEIN MOSHE FLUGELMAN BET BET 1 BETHESDA TESTING, TESTING. TESTING CAPTIONS. TESTING, TESTING. 1 RECOMBINANT RECOMBINANT DNA DNA 1 testing captions on Wednesday, June 15th, 2,005.

We're going to begin this afternoon with our data management report, and Dr. . Albelda will start off.

A number of protocols were reviewed between the march meeting and the June meeting, 16 submissions in total were evaluated.

That 136 these were not selected for in-depth review and public discussion. of the 16 submissions will be discussed at this meeting. We already heard about the protocol. This afternoon we'll hear about retro viruses expressioning angipotent 1 for peripheral artery disease. And tomorrow we'll hear a protocol about polio virus. In addition, there were submissions previously selected for public review, but the review was deferred by the submitter. of those is the protocol we heard this morning about 4. Erpes simplex virus. And we'll hear about adeno associated virus, expressing rpe-65 for retinal diseases due to mutations in that gene. For a general overview of the submissions not selected for the public review, 11 of these were for cancer; 1 was for peripheral artery disease, used retral virus, was a the falpox investigators, and was an rna transfer protocol. The staff reviewed a number of adverse event, and these were discussed in teleconference. And this time, none of these was thought to be ready for public discussion.

Thank very much. . Hesl-op?

I have 126 amendments. 16 of these involve sight and/or data changes. 9 involve protocol design modification, of which were single-patient exemption, and involved protocol status changes. of these was initiation of the study, and was closure or discontinuation. There was a total of 65 annual reports submitted, and 3m1c1 responses. of these was a full response to protocol 658, which is now open for a cool. And were partial responses for protocol that was entered at march meeting, in which they were unrespondent to issues raised during discussions. For those protocol, full responses will be submitted at appropriate time. There were 26 other notifications and amendment, presented to our committee meeting. And we decided that none were ready for discussion at the current meeting. And I'd like to thank the staff for all their efforts in putting this information together. Thank you both very much. That is the shortest data management report we've had in the last 4 years.

You and I also would like to thank the very much for carefully putting all the information together for us to review. We're going to begin with our and final protocol this afternoon entitled "a phase 1 safety dose escalating study of multigeneangio with patients with peripheral arterial disease. Dr. . Michael grossman from the university of Michigan will begin the discussion for us. Oh, thank you, Dr. . , see you at dinner.

Members of the committee, and gentlemen, good afternoon. I'm representing mgvs, sponsoring this study. I'm a cardiologist fromist rail, and I had post-doc training in vascular and gene stability in 1,990 through 1,992. -- Israel. 11 and with me here are Dr. . Michael grossman from the university of Michigan, Ann arrest bore, who is the principal investigator of this protocol, and Dr. . Mary price, who is responsible for the animal studies you're going to see soon. I would like to thank you for your previews. We've done a of work that I think improves the of the data we're going to present you today. In our presentation, I will start with a brief introduction. We have an overview of mgvs in order to define peripheral artery disease, talking about anti again since and tell you what the multigeneangio product is composed of. Then I will talk about the pre-clinical studies of the multigeneangio product, with a special focus on the issues that were raised by the reviewers related to it. Oxicity and biodistribution. I will express some of the important issues raised by the reveers. -- reviewers. Finally, Dr. . Grossman will talk about the clinical protocol. Mgvs was started in August 2,002, based here, that's a photograph of the hospital in Israel. It has 22 full-time employees. We focus on gene therapy for heart and believe disorders. And the prime focus was initiated in November 2,002, and the pre-clinical studies that I'm going present were completed in June 2,005. We plan on conducting phase 1 study with the university of Michigan Ann arrest bore. Products for growing new arteries and tissue are operative in the natural development and maintenance. I would like now to show you some of our scientific advisory board member, starting with professor from Israel, who is a scientist and discovered the gene. Professor neufeld discovered isoforms of vhgf. Robert Louis from haifa, professor Jacob schneiderman, the head of vascular surgery in 26789tel-aviv, hefted institute in Germany, who established the field in cardiovascular medicine 30 years ago; professor from reduce lum, scientist, he identified the premature; profession, gelong, the head of the gene therapy institute in reducelem, Dr. . Frank from lyons, France, and professor, a biochemist, our advisor. He works,and is the bell prize laureate for check industry in 2,004. Peripheral arty disease is defined as narrowing or collision of believes, supplying the lower extremity, most to arterosclerosis. Patients with critical ischemia have a mortality of 25%. Symptoms are pain or discomfort in the muscles of the leg, relieved by rest and aggravated by exercise, and may progress to critical imischemia, manifest by tissue loss and gangrene that eventually may necessitate amputation. To give you a sense of the gravity of this, this is the relative 5-year peripheral artery disease morality rates versus other common pastologies. In 5 year, 28% fortunate patients diagnosed would be dead, compared to 15% with breast cancer 18% with hodgkin's disease. So antiagain since is a very complex cell process. And the keys are depicted in this slide. Vgf binding to its receptors initiates the migration of the cells and eventually lead to a formation of a primitive tube. This tube is leaky to and protein and may be re-absorbed once the still lieu is gone. This complex process is reaching maturation when angi protein 1 is binding to the type 2 receptor and the recruitment of cells that mature the blood vessel and make it unreabsorbable. In the case of capillaries, the surrounding cells are parasite, and with artery, the surrounding cells are the smooth vessel cells. So from the outside you have smooth muscle cells in the artery. When we improved our understanding of the process of anti again since, it ignited the imagination of cardiovascular therapists because we wantd to produce and reduce antiagain since and blocked artery, sometimes of natural methods, surgery. And indeed multiple studies have been performed. a summary of studies to treat phase 2 and phase 3. In yellow you see studies that show vagf and fgf, both in coronary artery disease patients and peripheral artery disease patients. More than 700 patients were treated with preteen, and unfortunately, none of the study reached its primary efficacy . Additional studies for induction of antiagain since were using genes and isoforms, 165 and 121. The gene was used in the form of plasm id, used again for coronary artery and peripheral artery disease patients. More than 300 patients were studied. And if you put together the agent 3 study, more than 600 patients were studied. And again, the primary every chargecy endpoints were not received. So with this disappointment of failure of single-gene antiagain since therapy, we looked and tried to define the goal. In order to demonstrate should be the goal, I bring here example of a patient ma. In panel a you see the coronary artery supplying the dr.ircumflex artery of his heart. But you see on the stent of the left descending artery, a major artery that's supposedo supply blood to the whole anterior wall of the left ventrical, it's plugged. The right corner artery gives rise to the missing artery, right here. And this occurs due to the development of a very large artery. It's large enough to accommodate enough blood to supply the territory of the left ventrical. The of the production of this large collateral artery was described and studied for the last 30 years. In order to make a more die nationalcally significant collateral artery, you need multiple factors and the relevant cells to be there, and you need flow shift to direct the development of this new artery. When you have all of those, operating in coordination, you may end up with successful therapeutic antiagain since. So the rationelle for our process, the use of cells, was that the use of a single gene -- this is probably the fact that antiagain since is a complex process, involving multiple cells and protein, all operating in coordination. Many investigators now focus on stem cell therapy, but the cells that they use lack caught risation and require factors for differentiation. And some reports showing that when you use stem cell, you end up with increased calcification in the or the development of seratomas. We therefore focus on the smooth muscle cell, activated by growth factors, and injected intraterially to take of the sheer stress to produce a more die nationalcally significant collateral arteries. We have conducted this in vitro study to Illinoiste the importance of cells and gene types. In this study, we mix the smooth smous l cells and tag them with yellow dye and the smooth muscle cell was a green dye. We see them on the membrane. 24 hours later, they produced the ferrate, which is transferd to lick wud collagen that solidified at 24 degrees. 22 hours later, cells start to sprout. This the equivalent of production of new believes. This is a summary of many experiments in which we tried several combinations of cells and genes. In the you see that cells were expressing just the green protein, and you see that there is some sprouting, mostly of smooth muscle cell, green cells. When the smooth muscle cells were modifyd to express g -- vagf, we see more robust sprouting, again mostly of smooth muscle cells. When reintroducing the predeep 1, see a sprouting of yellow cell, mostly the equivalent of producing new capillaries sprouting. But only when we introduced vgf to smooth muscle cells we got this robust sprouting. Even if you see this for the first time, you see how robust it is, when you compare it to the other combinations. more importantly, if you look close, you see that in every sprout, there is a combination of a yellow and a green cell. So we have a sprouting of both types of cells. And this is as close as you get in vitro to the production of a new artery. We tested this concept in the proof of concept experiments, which you have in your report. And we've been able to show that flow was increased -fold with this cells and gene combination. Also, there was a 40% increase of arteries. So what is the multigeneangio that is the subject of protocol 703? It is a product that starts with a forearm vein, which we grade to smooth muscle cells. The cells are expunged in the cell processing center, then using investigators are expressed. After expulsion, the cell, in a liquid suspension, are returnd to the hospital, where this suspension is injected at sites of occlusion to enhance existing collaterals, and to produce from very small collaterals and more significant arteries that will provide the distal organ, the leg next, with enough mood to alleviate symptoms or even prevent amputation. In order to bring this concept to the clinical arena, we have established a model of ischemia, shown here in city and geography. You see that the super fish femal artery is disruptd to large side branches that keep collateral to the distal artery. Using this model, we focussed on safety. And the major issue, the issue that was brought up during our discussions with the fda was the cells, big cell, even when there are chip, and they may block capillaries, which are 7-microns. So we tested, and you have the data in front of you. And I'm not going show it, because there is no finding that supports this. Then we tested for toxicity of the trans genes, retinopathy and finally the production. When you overexpress growth factors you may end up with inefficient blood vessels in the form of actual oh,. Then we looked at efficacy, measured slow in the femal artery, and correlated the measurements with angi-ography. We measured biodistribution to test for tissue and temporal kinetics. Finally, we used the model, as we studied very many rabbits to set up the production and quality control processes. We studied human primary cells. We focussed here on production. We wantd to make sure that we can isolate and expunge primary vascular and smooth muscle cells in patient was cardiovascular disorders. Then we tested for gene drfer efficiency. And you have the data in front of you. We also tested for the safety of transferring cells to -- thereforing genes to the human cells. And the major issue was whether we can transfer so. Cells to cancer cells. So we used cultures to exclude the presence of stem cells in our primary cells. We tested for tumorous activity, a landmark of cells. And then we tested the number of transgene copies in our cells that compose the multigeneangio product. , safety. And here is statements, for gross pathology, so no gross leagues related to the treatment were observed in any organ or treatment any of the rabbits at any time point. Then a thorough study, no evidence of local or systemic toxicity of the multigeneangio to the rabbit. And finally, summary. No evidence of blood sample vessels shows dilating, noticed before sacrifice in all study animals tall points. What about trans gene expression? We see here on the left-hand side, we see expression of vgf in the plasma of the rabbits at different time points. There was a slight increase, 7 days after deduction of vgf levels in the plasma. However, all measurements are within the sensitivity range of this. Angioprotein levels were not detected in the plasma of the rabbits any time point. What -- we have done 72 city geographies. We looked for efficacy and safety, and it's shown here in the animals that were studied 6 months after cell injection. There is no for the development of angioma so. We have geoparagraphic evidence for the absence of angioma. Also we studied 15 massive biopsies, and could not find any evidence. Now we move to the safety aspects of the human cells. So the cultures are cultures that support bothed a s adherent and non-adherent stem cells. There are no adherent cells after 14 day, but there are multiple call news circulating. This is in contrast to primary human vein cells. And you see multiple call news of adherant cell, but no call misof non-adherent cells. We repeated this enter empty with primary human smooth muscle cells. Just for the sake of brevity, I don't show you the data, but it's in your documents. We see here positive control. We also used 293 cells, immortalized neo-natal human kidney cells. We also took human umbilical vein cell, and again relatively high activity. But in all the cells that we tested, forcing the gene, or after being transduced to overexpress angiopreteen, none this significant activity, indicating that these cells cannot reach a high number of divisions and they don't have a potential to be cancer cells. Again, we repeated this experiment with human smooth muscle cells and found the same observation. Last but not least again, the number of copies in human cell, this is a test of several samples of indoterial cells, expressing angiopreteen 1, that's the blue cells, and then smoothness expressing vgt. And the average number of copies was 3 for angioter-ial cells and 3. 7 for smooth cells. the y-axis is the number of cells transduced without selection. So we gate fairly high transduction rate. Based on the literature we know that leukemic complications of mice were observinged when typically more than copies were found in cells. Another recent pub ligs -- publication showed when the transgene copy number was more than 5 -- it is of note that all the cells studied in the papers were cells that originated from bone marrow. And the cells that we study are fully differentiated systemic cells so. This may not get you to the cell west study, but any event, the averagings are 3 and 3. 7. We now move to efficacy and we show you the mid-dosing. We refer to it also as therapeutic in our documents. We have many control, as you can see here. We measured the flow at day of sacrifice in the artery where it was surgerically occluded and later on treatd to the normal artery on the left side. And you can see that in each time point, 3 week, 3 months and 6 month, the flow was higher in the leg that was treated than in the legs that were treated with other combinations, whether it was just a vehicle; whether it was a low dose, or just naive in the arterial cells and smooth muscle cell, not expressing the transgenes. What about the fusion? Here again, we see that the pro fusion in the quadracept muscles were superior when we treated the animals with the multigeneangio when compared to most of their controls. This effect was somewhat weakened in the 6-month period, but this is probably due to the development of collaterals coming from the abdomen and not from the super fish femoral artery. These are examples of an angio, you see the control, the disruption. Femoral artery, and you see a lot of animals developing some collaterals. Another animal was treated with naive cell, and you see very little collaterals. And look you see an animal treated with multigeneangio, and you see several large arteries that supply the distal artery, and the same is true to the high dose, and this is at 3 weeks. But what happened at 6 month s? is this still true at 6 month s? here is another example, 6 months later, the controls, very little collateral circulation, low dose, big artery. component, very few collaterals. And look here; the therapeutic dose produced a very robust high number of visible collaterals that supplied the distal artery, as did the high dose of the therapy. The last issue to discuss is biodistribution. And on the table you have the summary of the biodistribution. You see that all controls were negative at 3 breaks -- weeks. The so-called therapeutic dose was positive at 3 weeks in the ischemic muscle and in the lung. At 6 months -- this is data -- that you don't have because it arrived only this morning -- the muscles turned negative. When look at the high dose, this is 10 times more than what we plan on giving to human, we see that at the lung, out of 4 animals were positive at 3 weeks, reduced to only animal at 6 month, and the ischemic muscle was positive at 3 weeks and 6 months. When we injected as a measure of safety to the worst case scenario, we injected cells to the venous system, the cells ended up lung, and you see that the lodges are very positive at 3 weeks. So in summary, we see that the cells that we inject do stay in the leg in the so-called therapeutic dose, and 6 months later, they are gone, they are not there any more. So I will hand the microphone to Dr. . Mike grossman, and I you for your attention.

Great, thank you, Dr. . Flugelman. I'm a cardiologist. My clinical trrs in coronary and peripheral artery interventions. And my.

Translator: Are in the area of therapeutic multigeneangio. I'm involved in several pre-clinical multigeneangio studies; I've also been the local principle investigator for several clinical multigeneangio study, some of which you may have reviewed in the past here. I'd liked to to discuss briefly the highlights of pro call 0,501-0,703, a phase 1 safety dose escalating study of multigeneangio in patients with peripheral arterial disease. The primary objective of this trial is to evaluate the safety of multigeneangio in the treatment of patients with peripheral arterial disease. We have a secondary objective, to obtain at least preliminary efficacy information in these patients. This is just a table describing in general terms a protocol overview. Initially, patients will be referred and obviously will go through the informed consent process before any other studies are done. Once the informed concept is signed, the patient also undergo screening. And then will return if they meet criteria with screening to visit or for qualification visit, including an x-ray treadmill test and ankle brachial index measurement. If they qualify and meet all inclusion criteria on qualification and screening and have no exclusion criteria, patients will undergo vein harvest, performed at the university of Michigan under local anesthesia by a vascular surgeon, Dr. . Upchurch, a member of our team. After collection of the vein, the vein will be delivered to the university of Michigan, where cell processing will take place. Approximately 21 days after harvest, the patient will return for injection of the study product inter-arrest tierly above the site of occ lusion into the lower extremity, and I will be performing these injections. Patients will then spend the night in the hospital and undergo rigorous follow-up over the next 2 weeks for adverse events. They will then return at day 30, 90, 1 # Ind 365 for exercise treadmill test, and other studies as described in the protocol. As a principal investigator, I will determine enrollment based on inclusion criteria yeah. The study product dose cohorts also predetermined. At the exclusion of every dose cohort, 14-day safety data will be reviewed, and movement from dose cohort to the next higher dose cohort will depend on prive approval of the dsmd. I'd like to briefly go over the inclusion criteria. Patients to be included will be male or female, between the ages of 55 and 80, and able and willing to provide the written informed consent. Patients that are female must be post menopauseall, surgerically sterile or use Kuwait birth control and have a negative pregnancy test within 24 hours. Patients should not be breast feeding. Male patients must use an accepted and effective form of barrier birth control. Patients to be included must have a history of exercise limiting intermittent quantification and privilege all arterial disease, with symptoms in or more lower extremities and the symptoms must be stable in the 2 months prior to screening. In addition, we will make the diagnose of peripheral arterial disease based on the indices of less than or equal to 0. 8 in both lower extremities after 10 minutes of rest. For subjects with non-compressionable vessels and abi's of greater than 1. 3, a toe brachial index will be required in both lower extremities. Subjects must have limitation in walking, with a mean peak walking time of between 1 and 10 minutes on exercise tests there. Can be no more than less than or equal 25% variability between these tests. Just to highlight of the important exclusion criteria, they include the presence of significant in-flow disease, defined as greater than a 50% narrowing or stenosis of the common iliac or the external iliac artery on imaging that must be performed in the prior year before screening. This can include conventional angiography, digital or magnetic resonance angiography. If patients have had prior bypass surgery, they must have documentation. Other exclusion criteria include chronic or acute ischemia, including rest pain, gangrene, a history of neoplasm, renal failure, conditions that you can see here, including those that may preclude retinan photograph, vascular lesion information the anterior section of the eye. Patient, excluded if they have con guestive heart failure or immuno disht conditions. Just a word about the patient population that we have chosen for this phase 1 study. This is a phase 1 study of multigeneangio, and we have chose tone study it in patient was -- this will allow us to follow for at least 1 full year and potentially for a longer of time with regard to safety. It has been that patients with critical inischemia and no option for revascularization may be future targets for the mga product. However, these patients have a high morebility and up to a

That 1-year morality, which would make collection and interpretation of data difficult and may confound the determination of long-term safety. Finally as part of our safety analysis, all adverse results will be analyzed, chemistry, hemotology and vital signs will be measured, electric card grams will be collected. All medications and demographics will be collected and reviewed by the dsmb, as outlined in your protocol. I'd like to thank all of you for your attention, and we'd be happy to answer any questions.

Dr. . Grossman and flugleman, thank you both very much. I'm going ask you to stay up there close to the microphone. And are going to begin by going from reviewer to reviewer. Each reviewer will restate the questions that they originally posed to you by. And whether they feel that that question has been responded to satisfy them, if the answer is they still need more information or we do, then ask you to comment again, question by question. I think as a general comment to everyone, I'd like everyone to sort of sit up after lunch and begin to work hard. This is a very complicated protocol with different cells and different transgenes put together and given to subjects. I'd like, at the the next hour or so, 20 be able to be as helpful as we can everyone in the process. And that's going to require of us, because this is very complicated. And we're going to begin with Dr. . Simari . hank you, diane. And I just want to maybe start off, take my last meeting to editorialize a little bit. I think the protocol is at the cutting edge ever future cardiovascular biologics, a combination of gene products and in combination with cells. And in that regard, I'm very excited about this as a potential. But with the combinations come complexity and with complexity comes the requirements of understanding the necessity of the components. And cells, as diane mentioned, well as gene products, there is a lot of complexity. And obviously in terms of safety, the safest thing is to exclude a component that's not necessary, especially as Dr. . Grossman said, although pad a high-risk population, the lower risk of the population. Because that have of the variety of comments, there are some things that still remain and continue to need to be clarified. So partly the record and partly for clarification, I will go through my review by and look forward to your responses. I have several series of questions raised regard to pre-clinical study, the of which gets at the heart of the mean expression portion of the studies. I asked whether, although the expressions has quantified in vitro, no expression of -- was originally provided. There was data that was provided today that suggested that the ang1 transgene was not detected in the circulation. So I'm left with questions. Is the circulating levels, are those suggested to be negative, or is that suggested to be small increase; and second, is there any data showing tissue expression of either transgene? And if so, from which cell?

If you look at the data, it's very similar to base-line .

Yeah. , so and also

Negative? There's none detected in circulation?

You I think we should stick to the very simple statement, observeative statement, there is a small Francegene increase in 1 week, which may or may not be significant. And afterwards, the levels are very similar to the base-line. So I would say overall, it's negative, but we have to keep an eye on this slight increase. Regarding the expression in the tissues, we looked at it very seriously for obvious reasons. Vgf tissue is extremely difficult, and we work with vhf, so we tried this for a while. And then we went to the biodistribution studies. If you look at the biodistribution studies, this is a very strong support to the presence of the cells, and the cells expressing their transgene. We cannot differentiate whether it's vgf or ang1 stressing cell, but we can definitely say that the ischemic muscles have cells that express the transgene .

o there's no -- you've not detected protein locally?

No, we cannot, because we assume that protein levels that we produce are very low, and based on the biodistribution, we claim that of the product is small apples of cells in a very close neighborhood. And it does not have a large effect or systemic effect.

And I'll to the biodistribution studies. But if I'm reading correctly, that was pcr; that right?

Yes.

So is there any detection of messaging?

No, we could not detect the message, because we didn't look for it. But I think the use of real-time pcr for expression is acceptable method, and it's been used in many protocols, in human protocols for transgene expression, as published in work. So I will take this as a real evidence for the expression of the . o there is messages detected ?

No, we did not look for a message.

So did you do real-time rtpcr?

Dallas no, no .

o there was DNA present?

Right, yes.

So there's really no detection of preseen, locally or

No, right .

So then the next series of questions I asked about had really been addressed by of the data that you presented in terms of the requirement of of the components. And that is, is transduction required and necessary, of the components and each cell type required ? The analysis that you presented here didn't have statistical analysis. Your data that you presented up there did have statistical analysis. There's a reference to usage t-test to compare the groups, multiple time-point, multiple comparisons. Can you walk through that statistical analysis?

You right. It's not only you that has to read 170 pages; it's us that has to write it. So we didn't have the time to do the statistical analysis. We looked into the method that's have to be used in these analogies. of fact, you can use just t-tests, because these are not measurements at different time points. These are different animals. So the data that was provided in the presentation shows that, regarding flow, and each time-point, the therapeutic dose is superior to every component alone, to vehicle alone and to cells with no genes. Regarding flow, the same is true except for several component, mostly at 6 months. But again, using angio, we see that blood to the region is also provided from arteries rising from the abdomen, and therefore, we may not have a very strong relationship between pro and flow. I think if you can ask me which is the most relevant measurement for assessing angiogenesis, it would be flow rather than pro fusion. don't want to harp on this, but since it's fresh and unpublished data, were there adjustments done for multiple comparisons within the multiple group

You well, I think if you divide the number of observations, I think most of this would still be very efficient, yes.

The next series of questions really get to a little of the biodistribution, and I asked whether data are presentd to suggest that liver cells survivors within the ischemic tissue; have you done studies to track the cell, rather than the DNA?

What they have done is biocell biopsies and stains for the smooth selections. There are high-dose animals where we've seen the cells in the muscle, but not in other animals. So we could not track them. And don't think there are other very sensitive methods to track very few cells in a very large muss toll show that they are on top of tissue.

Have you tried labeling a cell, either genetically or with a membrane dye to see whether the cells reside in the tissue? Because there be a possibility that the DNA would stay in the tissue, but the cell may be long .

ell, if you have a ph-ysiological effect and evidence that the cell is DNA or other evidence, you have a physiological effect and the anatomic effect, I think if you put and and , they all sum up that the cells are there, and they secrete the proteins that are and the effect is seen. It's very difficult to look for a single cell in a muscle. And if you inject even million cell, it's very difficult to trace them. We were able to see it in animal, but not in the other animals. And we reviewed many wipecies in .

With all due respect, though, that line of reasoning is lost at the secretion step, though, right? All right. I completely agree. But this was answered by using control cells that don't express the genes. So if you have cells that stick to the muscle and they don't produce the physiological effect and you take cells that have the genes and detect the cells and they produce a physiolodgeal effect, it means there is no direct proof for the of the develops.

Were the cells that were the cells transduced with a investigator that had a control gene?

No, no.

So they were not transduced? We've studied 130 rabbits with atrologous cells . id the animals that received a control get as well?

You yes, of course.

Okay, but they did not receive, so it could be transduction effect? cannot exclude the transduction effect.

Well, you look at each component, it's really carrying too long. You take a combination therapy and it down to all its components, cell alone, each cell with gene, and then you don't see effect. And when you use the whole components together Abe you see effect, the logic tells thaw it does work ands components are all necessary to produce this physiolodgeal and anatomical effect.

But as you've pointed out, the problems with angiogenic studies with human, within the gene of interest and we don't get effect and yet, we don't know why that didn't work. And that's based upon pre-clinical studies that shows robust tissue levels. I'm going to move on. Just let me comment on this. I think that the initial understanding of angioagain since was the sclaition system. You activate protein, and you end up with a clot, because it's amplified. But this a process that needs input all the way along the way. So you can in addition, but it would not be sustained. So you need the cells there to keep secreting the trans preespecially teen and other preteens. If you look at all the studies that were done, animal studies on angioagain since, nobody used different methods for verifying angiogenesis in the animals. We've done flow, geography and pro fusion. And I think that the evidence is there.

I'll move on. Since the complexity of the cell typeactually increased because of the non-homogenate of each of the cell condition, the question was what was the purity of the cells used in the pre-clinical study?

We provided the information. of you. We have used the criteria for smoothness of develops of 70% purity. And we've been able to use higher purities in most of these experiments .

o when you say smooth muscle cells of 730% purity, smooth muscle alfpa and cells of 70%.

Right, right.

What makes up the other 30%?

We don't know. But we know from everybody else that's been going smoothness of cells, you will never get 100% staining in the cells. You know that smoothness of cell lay is heterogeneous. Therefore, 70% staining is very high staining for smooth muscle .

o you think that it's just an insensitive marker?

I do. You think they're all smooth muscle cell s?

Probably most of them are smooth muscle cells. There might be some fibroblasts, but don't think that's a significant safety issue .

then ask whether any further data were presented with the clinical human product from upper extremity vein harvest. Have you ever worked with upper spremity veins?

No, .

I ask then about biodistribution studies how the cells were tracked. It's now clear that cells weren't tracked; the investigator was tracked. I asked the question about the original dose, the response is not clear, the highest the rabbit, arterial flow was not increased, but pro fusion was not increased ?

Well, this is true. And we think that going higher the higher dose probably is not needed we reached a saturation of collateral information. And therefore, usage higher dose is not justified . ould I ask for a clarification? Could you just define the difference when you say we sought the change in flow versus the change in pro fusion, just so that everybody understands? All right. We've been using different methods, ajess fusion, and to assess flow. To assess flow, we have a meter giving you the results in millimeter per second. We compared it to the leg that not operated and was not treated n regard to pro fusion, we used a laser doppler system that sends the laser beam to the muscle and traced the velecity of the blood sample cells in the tissue. And it gives you an arbitrary unit of pro fusion. The faster and more red blood cell, the higher pro fusion of the tissue. So these are different tissues. You can get low flee and still high pro fusion, because you're measuring just source of pro fusion. So if you have arteries supplying a muscle, you can getSo just by measuring flow, you don't get the whole picture. But when you combine pro fusion and flow, you do get the whole point.

I then asked about the role of 1, that was really not discussed, but subsequently discussed. But perhaps for the group in general, you could talk about, are there concerns about that or not?

of course there are concerns in any genes that have never been tried in humans. But overall, our vee of the literature teaches us that angio preseen acts as a maturation factor. Ang1 special was down-regulated, while ang2 was down-regulated. 1,207ing the process of angiog enesis, especially in lit trier rye lated to can desh -- literature relating to cancer. It had been shown a group that use of ang1 or overexpress of ang1 stopped vessels from being leaky and bringing them to maturation and stabilization. So we think that overall, the effect of ang1 in our product is a favorable effect, to the tiewrtd of the cells and the flow eventually by the product.

Going on to questions about clinical study, and I was confused by the patient population. And it was referd to in the scientific abstract as no-option patients, but in the inclusion criteria, the patients were described as those with chronic modification. Indeed, in the response to that concern, the statement is made that the procedure target population of the multigeneangio is the long-term patient. All right.

Clearly, there are multiple ons for exercise and pharamologic well as interventional therapies. So I'm assuming that these are patient who is would not, who were not able any of those? Can you how you would choose between exercise programs, any of the drugs, versus an intervention or surgery?

You I think all the patients that we would enroll in this study be patients who have been on medical therapy, been offered and gone through exercise therapy, counselled to stop smoking, have been evaluated for revascularization and for a variety of reasons may be felt not to be optimal candidates for revascularization.

And who, in the process, will make that decision? Will that be an investigator or someone who's involved in the study?

Our initial plan was that in the principal investigator would review the patient data and make that determination.

Okay. I talked about the risk of harvest. Clearly, a less-sick population, a harvest may not be a big deal. Cow explain how the harvest is done?

It's done on an out-patient basis, and it's associated with local anesthesia. We also reviewed the literature and provided a very comprehensive review of the relative ease that it's and the lack ever complication of vein harvesting from the forearm .

y next series of questions regardinge population of cells to be delivered and the exposures, I asked tow prolific were the harvested cell, and you provide really beautiful data shows the log-rhythmic in the culture. And I referd to a paper by David ingram from Indiana, who recently showed that the reason the cells have that logrhythmic growth is a low frequency, high proliferative regenerator in there. Can you address the question of why those cells grow like that? If you for drawing our attention to this series of papers there. Are now papers. Well, since we die light the cells 1 to 3, so this is into the great surprise to us, that can you have 40 or 45 divisions in the primary cells. But we cannot get them beyond this passage 15 and 16. So whether it's the conditions that we use or whether they're unknown cells, don't think that this is a real issue in our cells, since we provided -- we have such extensive experience by being unable to push the cells beyond passage 15 or 16, as we stated .

o there's a descending limb of these logrhythmic curves?

You that's it and they don't divide him .

nd that's 15 passages?

It's somewhere between 15 and 20 passages. And no how much growth factors you add in, and whatever conditions you change, these primary venous cells cannot be pushed beyond these passages.

And should be noted that ingram studies and others use very defined conditions and may not support the cells .

I completely agree that this may be -- we called combrex, the company that provided the cells and they use human cells. They are a very good source for primary human vascular cells. They take it from car accident victims. We asked them how old the people are that donated their organs for cell culture, and provide us this information. We use venous from patients that undergo bypass surgery, so they are elderly patient was multiple risk factors. We provided some literature to show that the number of stem cells, where there is residents tissue is reduced in these patients. So there may be another explanation why we don't see what ingram sees in his papers .

n the protocol there's a number of steps that include exposure to non-humor non-recombinant proteins.

Yes. We've been dealing with these issue was fda for quite a long time. We look for replacements for all non-human material. And we think are going to find all of them. Maybe later on we'll to find a replacement for all of them, and I think that this can be done.

So phase 1, you'll be using field serum?

Yes, right.

I raise the question, given the heterogenae of the cells, are the cells that make up the non-population or the non-smooth-muscle-cell population or you wish, none alpha acting or the cd-31, are those cells differentially transduced? Are the non-labeled cells or less advanceduced?

Well, more than # 0% stand for 3. And I think with 90% represents something like 100%. So it's very low, and it is impossible to isolate these cells and show that they transduce differently from others. The smooth muscle cells, there are several phenotypes of smooth muscle cells. And we don't know whether the synthetic or the other types transduce differently from others. But overall I think that if you look at the data of transduction, if you transduce80% of cell, you probably transduce80% of smooth muscle cells. We say we cannot sustain some of them, 5% or 3%.

So the 0% c-31 positive cell100%, is that what you said?

Yes, it's very close to 100% of facts, yes.

I asked the question about copy number, and you have done that, presentd that data with or copies of retrovirus. fortunate concerns, given the theoretical possibility of a highly plolific cell within the population, and retroviral transduction would raise the question, given the reference of the studies about -- oncogenicity. the longest an animal has been You we have followed animals up to 6 month, with rabbits. We were going to follow for 1 year. But if you look at the pathological data, there is no sign of any tumor or any outgrowths of any cell types. So I think that 6 months in a rabbit is a very long time to exclude tumor genecity of the injected cells. Yeah, we should all remember that the paper that was in the New England was 34 months.

This is true, but this is different cell types. These are fully mature, fully differentiated systemic cells, so I think that is a different arena. I had informed consent concerns; I thought the therapeutic misconception was present throughout, mentioning gene therapy and treatment.

Yeah, thank you. This was addressed, and we revised, based on the comments of everybody. We completely revised the document and I hope it will be to your satisfaction.

Was the revised concept form included in the response?

Yes, of course you have a separate file with a cover letter, showing all the changes that we made in the new document.

And the relationship of the pi to the sponsor?

The clearly stated that there is no economical incentive in this study, complete.

The Okay. I think that's it. So I'm going to actually beg of you some put my thoughts together -- (inaudible).

. Epstein, now that you've seep a model

You well, that's sort of incentive for me to apologize. I was asked rather late to be an ad hoc member of the committee, so I didn't have an opportunity to formulate my questions in sufficient time for a written response. So I do appreciate, that and I realize it's a little unfair. But let me formulate the questions I have, which I think in large part have been answered in your presentation. But my -- and had an opportunity to see the other reviewers ' question, so I didn't ask a lot of the questions that I would asked, because they were stated in the questions from the other reviewers. So the major concern that had and I had hoped that you would answer and I have in large part is not so much the oncogenecity of insertional mutogenesis, but the fact that you're putting a retrovirus into cells that subsist for prolonged of times. And as you yourself indicated in your presentation, and I quote "our data shows that transduced human cells for as many as 11 passages after transduction" so clearly these are not being dieuted out and injected in vivo undoubtedly persist for a long of time. And my concern was based on the fact that there has ban body of evidence suggesting that prolonged unregulated expression of vgf can result in he mangiomina production. And I put into the record several of those references. And I could just summarize; many of them are from the laboratory of Dr. . Blau and colleagues, who, beginning in 1,998 and up until the present time, have demonstrated multiple models that chronic exposure of vgf can cause "networks of vascular channels and me -- hemangiomas." Most relevant to your strategy is their study published 2,000, in which they transduced the my-oblasts with the transgene and injected the cells into the myocardium of mice which resulted in intramural vascular tumors. In of , surviving mice." I had an additional concern when I came across a paper relating to ang1. And this was a paper published Dr. . Yu in 2,000. And you have that paper. They were interested in the gene expression profiles of hemangioma obtain interested infants. And they demonstrated that interphelial cells derive interested infants demonstrate upregulation of type 2 message and protein, with a concominbant increase in cellular responsiveness to ang1. And the author concluded that type 2 and ang1 may play a role in the pastogenesis of hemangioma. So sort of with that pack ground, I carefully went over the toxicity studies that you reported, and you've now addd to the information, which is I think very helpful. But in the original packet I had received, you had studied 130 rabbits in your toxicity studies and you had received 51 full histopathological reports. What wasn't clear to me, how many of these 130 pathologic studies have you done for the full 3 to month s? and of those, how many were medium or high-dose of the product? That would be important. And what I had also asked, I had questions as to how, if there are little hemangioma growing over 3 and 6 months, which might get larger, longer term, issue was concerned if there was a single or cross-sections of the ischemic lag that you might miss them. So I had asked, and you replyd to that satisfily what the thoughtedology was to rule out the possibility of hemangioma. And I had suggested that perhaps a more powerful way, other than histologic session of demonstrated hemangioma would be late-term, 3 months or 6 months angi-ogram of the lag, or mri or a ct study. So if you could, if you'd like, I could reform late the specific questions.

You thank very much. I will start , all the animals that participated in the toxicity undergone full review. It wasn't available when we submitted, and it's now available. And it's in the review response. So it's all available. It is important to note that this in vitro. We know that genes transferred by retroviral investigators silenced over time. So there is silencing over time with retroviral sectors. So we cannot extrap plate for the longevity of the expression. More importantly is the data that we received just this morning, and you don't have in your report, that the 3-week animals that have retroviral sequences in the ischemic muscle -- and this was in all animals that we studied -- when compared to animals studied, the same animals, so the same dose, additionalal animals receiving the same dose at 6 months had no viral sequence, which means that cells are cleared from the ischemic muscle. So I think that putting together the pact that we have a full pathological report with no signs and we studied hundreds and hundreds of slides, slide from his each leg, we have done 72angiographys and at least 25 at months and we found no evidence of hemangimy, and the fact that the cells cleared from the tissue, the muscle, and the theoretical effect of silencing, when it's all put together, I think that concern is still there. But in our studies, we have found no evidence to the curse of this .

Thank you. You have had asked, cow provide us with complete results of the biodistribution studies that you partially reported. And in the original submission, I think you've done that; that correct?

Well, --

The biodistribution data?

We're still doing some more animal, and we keep tissues of all animals for biodistribution, so we're expecting another or animals. But the most important ones, the ones that we wantd to present here and have the data been done, and I reported them to you.

Okay. And then I had the same concern of as to the statistical, as the did the statistical analysis in the rabbit studies to show efficacy, and I think you've answerd that. The pig efficacy study, what I asked was, the mini pig bilateral ischemia model, there was mini pigs used, the krill call endpoint is flow to the ischemic, the hind limb and there was no p-value established. I wondered, was that an omission and you forget to put it in, or was there no data? The pig model?

You you again, because we finally did it 3 years after the completion of this study. So we did birth defects in the number of arteries and numbers of end flow, it was highly significant, like 3 or 4-points after the decimal.

M'hm. And just one suggestion I think, you know, is obviously extremely compelling to state that you show efficacy with the intervention. So even if you don't have any data demonstrating protein expression, something has to have caused the increase in -- is it pro fusion, or it flow that was increased in the rabbit s?

You it's d. A, b and c are correct, pro fuse, flow

In rabbit s?

Yes.

Right. I would just like to suggest, because it would make, I think, everybody feel youured: You're injecting these cells intra-arterially, they are distributed over an enormous range. And you're quite right in saying would be very, very difficult to either demonstrate the cells histologically or to demonstrate by western or eliza a preteen present. But the question still remains; how long do these cells persist? How long do they express their transgene products? And a simple way to do this, because we've done it with other, or actually with vgf, using cell, is to inject the cells intra-musclely. You could say here is where we've injected the cell, and just sample you'd have the answer, I think, to the several questions that rob had and I have. I think we would be reassured if, number , you got ang1 and vgf expressed; and that 2, it was turned off over a reasonable of time. What do you think?

We were intrigued by this issue so. In the pig model, we did inject cells intra-muscle yard. But I think this is completely different than the ones injected arterially. And it didn't work very well for the methodology. But I think that we were able to establish the cause/result in our experiments. There are different methods to show that flow, in the -- I mean flow, number of arteries and profusion is increased. It doesn't happen in any other control. So it's true; western is an extremely insensitive method. We thought of tissue extracting in doing eliza, but it didn't work because of the components of the tissues that, when we're massacring the.

What kind of cells were you injecting? And ran itgenes were human?

You yes, we extracted them with fluorescent dye, we couldn't find the cells. And when tried to assay for preteen, were these humor pig gene s? Well, all the genes that we are using are human gene, yes.

Yeah. Okay. I'm finished, thank you very much.

Dr. . Dewhurst? Thank you. I'd like thank you all of the work that you did and the additional data that you supplied to us. Before I get to the questions that I had, I just had a couple of follow-up things. of them relates to the work from the brau laboratory. In some of her experiment, she sees abnormal capillary bundles as early as 28 days. And certainly by 64 days she's seeing hemangiomas, as Dr. . Epstein pointed out. Is there any difference in the mouse model versus the rabbit model? And between the models, which is the more reliable predictor of how a human being would behavior?

I think it would make a difference if my memory did not betray me, that she injected everything intra-muscular. They direct the investment of blood vessels. When you inject it inraa-muscular, you may end up with a bundle of blood sample vessels, but there is no direction. But with arterial injection, this is following the school of -- we claim to have produced viewable and useful blood sample vessel, as we have shown .

kay. So your argument essentially is that there's just a fundamental difference in your delivery method, at least to a more physiologic process?

Yes. Just to comment about which one is more reliable in comparison to human, don't think that any is. That's why we used two animal models, to feel more confident that we can carry this to human, it's the miniature pig and the rabbit.

In terms of the written comments I had, the firs. T written comment I had related very much to a point that rob brought up initially, the question of the purity of the cell population; and specifically the notion that you could have highly proliferative, lineage specific cells in there. And I'm still somewhat confused. So the in vitro work indicates that these cells have a limited capacity for proliferation. What isn't clear to me is whether that in vitro work would indicate in vivo studies. So if you had even relatively rare progenative with a growth-promoting gene, there's certainly at least the possibility you could have a somewhat unexpected outcome in vivo. I suppose where my concern is going here, and something that's come up a couple of times in our discussion is this notion of risk benefit. That you're seeking to work with a population who are relatively stable, with a better progno sis than the no-option patients. And that may put some onus on safety considerations. So I want to leave that maybe as a comment for now.

Okay, thank you. It is a true concern, because I'm not aware of any method identify a single cell. And we may hit it with of the growth factors. But what really makes me feel very confident that we are doing the right is the biodistribution stotted they we received this morning, 3 months, there cells in the ischemic tissue; 6 month, they are gone. And the pcr that we are using is highlyWe can detect up to copies so. If there were stem cells proliferating, we would be able to detect it with biodistribution. So we have substantial evidence to show that t-cells are not there. Of course the difference is that you're working with a rabbit cell population. And in the human condition, you'll be also working with cells from the upper extremity, which could conceivably give you somewhat different results . ell, the rabbit is young and healthy, and the people we are using they have a very low number of stem cells, so it would work to our favor, and not the other way around.

Hopefully, yes.

Yes.

The second question I had related to possible effects of ang1 on neural cells because of some literature that some pro lival neurons have receptors for ang1. And you've respondd to that.

This was very challenging, thank you.

Then the third question I had related to the finding that at the high dose, you saw significant innocence of cells lodging in the lung and persisting over quite some time so. What you showed United States today was that the therapeutic dose, you didn't see that, but you saw it at the higher dose. But in the written comment, your explanation of that, I found confusing, I guess. The argument that you're presenting is that the diameter, I gerks, of the relevant vessels is sufficiently narrow that would expect to exclude the cells. But I don't understand argument falls into then this dose effect that you're seeing. I mean, the vessels are too narrow, it shouldn't matter how much you put in.

Right, I apologize for theIt's true that cells that we inject are about 20 to 30microns and capillaries are 7 mikerons. In any tissue, you have bypass. Some of it bypasses the capillaris. When you inject this very high dose, times more than we plan to to human, you saturate the tissue with all the cells that it can take. Then arterial venous bypasses open and they don't see the capillaris and they end up in the lungs. And we know it does happen in certain situations of stress, even in humans, that the tissue does not see the blood. It comes through the arteries, going meetly the veins and bypassing -- the shunt, yes.

Just to follow up on that, may assume that in humans with different levelless of vessel blockage, for instance, you may see variability, some subjects in whom.

This is a very good point, yes, because they are not unified as are the rabbit, yes, very good point .

had some technical questions -- Excuse me, Dr. . Dewhurst, you comfortable with what you've heard then about the potential for cells going to the lung and lodging there in a human subject? Or would you like that mentioned?

I'd like it mentioned, think. My understanding that could happen. All right. I had some questions about the cell isolation process, the purity of the cell, the potential failure rate of isolation, and all of that is very thoroughly addressed, thank you. I this a about the 1-year cancer screening. And I felt, I guess, that the initial approach was rather passive; that is, that subjects were to be advised to get tested. In the written comments it's clear that you're planning to do a multiple-year follow-up multiple-year cancer screening. And I wanted clarification there as to whether that is something the patients being told to go out and do; whether there is something built into the study that you're doing.

This is also connected to the fact that you may find cell notice lungs because many of these patients smokers. So we're going to subject them to a lung ct before recruitment, and probably we'd do it after 1 year. So we're going to be active in screening for cancer .

o you're going to do the ct scans of the lung?

Yes.

And is the study responsible for the cancer screening annually? Is that something you advise the patients to go do? How are you going to follow that up to make sure it's done?

You it's something we should consider. We haven't made decisions yet, but definitely we'd consider this.

Okay. Then the other questionsly related to the concept form, questions about language, clarification of risks associated with cell embolization, all of those I thought were very appropriately addressed. I don't have any remaining concerns with those issues. There were a couple of minor things in terms of the extra data that you provided. I think dollar couple of small typos in here, things I just wantd to clarify. So there's a table on page of the additional material. And if you look down at the

That-4 week sacrifice time, I think there's some typos with the standard deviations of the wait. I think there's a decimal that's in the wrong place in a of those. Because they would be

Right, sure.

But the more serious thing I was curious about is in some of your data, have an n of two animals and you're giving us standard deviation. I wondered how that was achieved. I also wonders why the innocence an ma'ams analyzed aren't always consistent between the groups. So you'll have like 6 and 6 innal circumstances and only 3 of 4 in an mean yeah, for instance.

This is easy to Spain. We've done studies, essentially. was safety, and we studied 98 rabbits. Then we studied additional 30 animal, right? Additional 37 rabbits for efficacy. And we used all the data for blood testing and for histology and for weight. that is the original for the inconsistencies for number of animals. We can clear it by writing. That's not a real issue.

Yeah. Thank you.

And I'm sorry for the -- we call it a typo.

Dr. . Power s?

Ly comments and questions from the perspective of biomedical ethics only. My had to do with the misleading therapeutic benefit throughout the concept form. The concept form was revised, as you say, very substantially, and I'm satisfied with that. I'd like to comment that I think the format from section 4 through 12, with frequently-asked questionvery engaging and I think we clearly written, a really nice way of doing things. My second set of questions had with the relationship between the sponsor and the investigative agent and the investigator. The one question thishad to do with any financial interest of the -- and that was laid out nicely in the informed concept document and the front as well I think is very useful for the subject to see, well as in your written reply. I asked also about the relationship between the sponsor and the process of monitoring or reporting of saes, termination of the study for safety reasons and the like. You provided a very leptdy and detailed, illuminating account of that, and so all of my interests were satisfied in your responses and in the informed concept document. So I it you. Thank you. Thank you.

Other comments or questions from member s? yes?

Dr. . Albelda?

I'd like to make a general comment. And I think . Simari was getting to this. In terms of understanding the mechanisms of what's going on and being able to follow some sort of gene transfer or cell transfer, and you answered by saying, well, where we saw the physiolodgeal effect different ways; that should be enough. And I would say in terms of your safety profile, that would be enough. But I think this is an ideally-designed transfer trial, but you don't understand the mechanism very clearly of how this is working, and you're not going be able to follow gene transfer or ang levels or any other type of biomarker. And what that ends up doing is, to use an American phrase, you know, you're swinging for the fences. If this happens to work, -- work, that's great. If it doesn't, which most trials don't, you don't really understand why. for instance, the only thing you could look for are physi-ologic response, and this is a phase 1 trial. Let's say you didn't the responses; it didn't work. What do you do next? Did it not work because you used the wrong transunit or the cells didn't last long enough or they lasted too long? Or they didn't get to where they're supposed to be or you didn't have any gene transduction? In the study this morning, even if the study didn't work, per say, it was really going to advance the work. Now, you're putting up all the money for the company by doing N I think you're taking somewhat of a risk, since there's nothing you have to follow, own the homerun, physiologically, if this would work. So it's not our job tell you to do this trial or not; that's for sure. But it is our job to bring these things up for other investigators. And I think rob will comment on this more. don't think the trial is designed in a way that you're going be able to know what to do if things don't work.

you investment. I would like to disagree with many of the comments that you made. We do understand the frustration. We provided in vitro data. And we provided data in animals to show that we remodelled pre-existing collateral arteries. We have anatomy; we have the many aspects of physiology. We do understand what we are doing, and we're doing it in a better way. We're doing it in a better way that any other gene therapy trial that's been done in the U.S. up to know, injecting vgf into a muscle doesn't make any sense when you compare it to the data that we presented here. So I want to resent your comments .

Well, you didn't show us where the cells were

Excuse me. Of course, we've showed you. You at the data that we provided in the pig studies. We've traced the cells in the tissue. With biodistribution, we've been able to trace the cells in the tissue. So in every study that we performed, and it's all based on our in vitro study, we see that coordinated sprouting with the smooth muscle cells contribute to the remodelling of pre-existing collateral still lotion. I'm sorry that I did not explain it clearer, but the study and the therapy is aimed at expunging existing collaterals. We inject it in collaterals that are remodelled to accommodate larger amounts of blood through the vesceral artery. And I think that we have provided data to show that this is the case.

So Dr. . Flugeman, thank you very much. I know that you've presented an enormous amount of information to us. I believe that there are many residual concerns amongst the member, and we will do our best in a public forum to summarize those in a moment. Dr. . Kwon? Ms. Kwon?

I guess my question is directed as much to the scientific members of the group as to the proponents. We have been dealing over 2 or 3 years with the fallout from the trials. And of the things that we have thought was very beneficial was that the way the trial was set up and the way that the ti's had decide to monitor and to follow through with things made it a lot easier for us to immediately know a lot more about what was going on in the mechanism, when these adverse events followed. And even in the discussion that we had this morning, it allowed us to not repress other protocols from going forward because we were more able to narrow where the risks were. And even though nobody knows exactly, it didn't result in blanket stopping of a lot of other protocols. And I guess I'm asking my colleagues here at the table: Is this protocol designed such that if in is wonderful results and then all of a sudden serious adverse events, will we be able to parge out those things that not only help the particular protocol, but limit the fallout of other, on to our investigators in our protocol s?

Dr. . Simari?

If there was a number a multi-component study, could you identify the components? And the answer is no. This is a multi-component study. And the benefits are hard to separate ain't risks are hard to separate out. That's what I was sort of feeling as I listend to the discussion. And my concern here is, I think we all look at these protocols and say it would be really wonderful if we got positive results. And, but then when an adverse event came about, I think we're, based the way it was set up, we're much better able to isolate the damage, so to speak, on other investigations. And that's a concern of mine, just in terms of our responsibility to all investigations.

Dr. . Simari?

You well, I just want to clarify your response to Dr. . Albelda. Did you conclude cell-tracking data in either model in the response? And if so, where is it in this response?

I think we included it in slides showing, many the -- in the minute tor pig cells .

t's in this response?

No, it should be in the previous document.

Cell-tracking? Because your direct response was that you have that data. I haven't seen it.

It's in the distribution .

No, cell-distribution data of intact delivered cells.

We have data to show in minute tor pig that we tested that the cells, in the muscle that we tested, the cells in the must l.

And how are they labeled ?

With gfp antibodies. don'tay. I don't know if anybody can find it. All right. Dr. . Demetz?

Dallas I have to apologize, and I'm passionate about numbers. LAUGHTER ]

When I look at your definition of objective, you talk about either a dose-limiting study or safety. And those objectives are not identical. So sometimes that language is used somewhat loosely. The design, as I understand it, is really, at focus is to figure out the dose-limits toxicity. So I think we're always looking for safety all the time, but you're really focussing on that. So I looked at your dose sclaition scheme, and it's a standard scheme, and I think it is probably okay. But I thisth trouble finding the definition of what a dose-limiting toxist stay was. You referd to nci criteria of grades 3 or 4, defined in term relevance of blood cell counts so forth. don't know if that's the right definition here or not. Now, to look at the dose sclaition scheme, you have a very detailed argument but your design is looking at a dose that roughly one-third of the patient also experience. Now, I have no idea whether that's the right target or not for this population t was designed for cancer patient, for whom, they are in serious trouble and they're willing to tolerate a lot of toxicity, and it's probably appropriate for that design. Again, don't know enough about your situation to know it's the right. I'm not sure if you'd thought about it, but you'd probably have to build some argument if this is the right target. You could use a different strategy, but you got to figure out what target you want. But you're looking at targeting one-third of the patients for dose-limiting toxicity. So there are those kinds of issues in the protocol that, I think, in several places that really need to be tightened up to make it really operational. Because once you get into the field, I think you'll hit some question marks and trying to figure out, well, what did we really mean here? And some think building it; is this what you really want? Because that's what you're going to get in termless fortunate estimates and what the dose-limits toxicities are. So that's the gist of my comments, tightening up in or places. -- in 4 or 5 places. For a phase 1 dose-limiting toxicity, you're making an awful lot of measurements. Now, I know it's complex, but your goal is to find the dose-limiting toxicity. don't that you have to do all those measurements. That may be for the next step. It's probable lay you mean question, but it looks like you're doing an awful lot of measurements here -- probably a dumb question. Thank very much .

Yeah. I would like to speak from the perspective not so much as an ad hack rac reviewer, but as an investigator in angiogenesis. I would sort of re-emphasize what's been presented by the investigators; there have been huge numbers ever therapies designd to ens Hans -- enhance collaterals in patients. And they are all built with a single protein or single gene my own per spend active is that it's too simple to try to grow blood vessels usage single fact oror cytocyme. I think the cutting edge is multiple growth factors because it's so complex to grow blood vessel, and they require so many of these growth factors. Soy think what Dr. . Flegleman's strategy, although it's not neat, I'm saying, okay, let's study agent at a time, if necessary, if the field is to go forward. So it is more complex, but I think not frivolously so, but necessarily so. And the other thing about the cells and their distribution, I understand the froes frags; I'd like to know where those cells are and our strategy has been to inject cells intra-muscularly. But I think it's interesting to inject them intra-arrest tierly to take advantage fortunate flow vessels. And once you're injecting cells into the single tree, single cells, you just won't be able to see them on histologic analysis. I mean, you have to do thousands and thousands of sections. So I think it's not as neat as have other protocols been. But those protocols have failed. And so you know, I think there have to be some -- there has to be some gamble, you know, not for patient safety, but perhaps not having all of the answers that you'd like to have.

I'd like to respond to that. There have ban number of clinical trials that have come forward that have used a non-iterative approach, rather than suggesting that they know what the factors, use be transcription factors. So there arrest number of investigators working on multiple growth factors, as we have. And the literature is full of tracking studies, both, you know, with mr using labeled cells or Stephanie's group using Indian-labeled grants on histology, where track will be different. But there are ways to use muscle trackers that are commonly done .

May I make a comment about Indian labeling? Indian labeling, when you label the arterial cell, it's within 24 hours due to the radiation. We looked at this method for tracing cell, and it's not a Ile option for us, especially in long-term studies. -- viewable option. In flag we tested, we found cells in the muscles. In high-dose rabbit, we found cells in the tissue. And the krills in the ischemic tissue and nowhere else -- cells.

Dr. . Deluka? Yes. Not to belabor the issue, but I'm still a little confused, and some others are as well respect to to this gene expression issue. The deal is that you can detect a investigator, the DNA of the virus for 30 day business rtpcr, and then by days you can't detect it.

Right.

Okay. But if you looked for, did real-time reverse transcripting pcr, you can't detect the message any time?

We know that expression needed be for a long of time; otherwise, the blood samples would not expunge. -- blood vessels.

Okay.

So to look for vgf or ang1 within the tissue, I think is

I would imagine the message would be as easy to detect as the DNA by rtpcr, if not easier, since this are more copies of the message than the template.

Well, that's something that we should consider, yes.

You know, there's unpublished data from direct gene delivery that subjects that, you know, the area of detection, this is intra-muscular, the area of detection of the DNA is broad, the detection of rna is more narrow.

of time transduction events non-productive, is that what it's saying?

You I think so.

So the argument comes to a again itsic then, right? If you put both genes? In.

Yes. I've only been shown the statistical analysis today. If I was reviewing as a pap, I'd have a statistician look at data. I would have to defer to a statistician. 1.

You any comments or questions?

I just wantd to make a final comment about the informed consent document. Although it is much improved over the fir st version that we got, think there's still a fair amount of complex language in the document would not be understandable to all subjects .

We will take ever it. -- care of it.

comments from members of the audience, the public? All right. Thank you very much, both of you. And you may sit down and we are just going to deliberate amongst ourselves to determine the content fortunate letter. So -- content of the letter. So we began with a general and simple statement. The patient group chosen to receive the gene transfer are relatively healthy. The mechanism by which the strategy may be effective is poorly understood from the pre-clinical studies. The method is difficult to disearn the relative risk benefit -- discern. Pre-clinical study, there are at least five remaining issues. I'll read it and then ask for comments after each wun. 1, in order to explore the buying of the proposed transgene, further consideration should be given to defining the role of non-modified cells of both types in improving the profusion of ischemic tissues and defining the survival of deliberate cells on the tissue. In orders, which cells get, there and how long do they survive?Rob?

A retrovirus expresses -- the details suggestd that they are available upon request, which we didn't do. But you know, are there moiivegs fortunate cells just in the process of transduction or with the transduction in a retrovirus that might have this effect? I guess I would like to see those controls if I was reviewing this as a manuscript. And that question of what level of evidence does one need to do a phase 1 clinical trial is what we're really dancing around, I guess.

All right control issues include the use a modified cell with a non-retroviral gene well as non-modified cells. 2, results fortunate pre-clinical studies would be clarified by determining the purity of cell sub-populations, so populations utilized including the cell type of the 30% of cells that are reportedly acting negative. Are the non-active staining cells transduced or cells contained within this population? 3, biodistribution studies should be completed, which track the delivered cells themselves, persistence of the cells and cell products and the tissues should be determined, utilizing rtpcr. 1.

Rnartpcr, thank you.

Real-time rnartpcr. Just a minute, got t.

My typing getting worse this afternoon. Okay. 4, pre-clinical data should utilize the proposed clinical human product. 5, dose response should be clarified both for cells and cell products for all pre-clinical studies. Those are very broad-reaching, and I think touch on the remaining concerns before we move to clinical, which, we have another set. Any other pre-clinical concerns, without focussing on specific studies?

You I think we all had concerns about the statistical data .

eah, I think I had that -- additional statistical analysis would help with clarification.

To be enrolled in the study, what criteria will be used to include subjects who isve utilized the usual treatment strategy, such as exercise; and who will make that decision?

I would, you know, in the past we have, in a case that's been gray, we have asked them to consider non-involved, a non-investigator clinician to describe the risks and benefits of this investigation, compared with other therapies.

M'hm. (inaudible).

Our microphone has gone down; I'm sure you all realize that .

There you go. Clarify the patient population to be enrolled; what criteria will be used to include subjects; consider engaging a non-involved clinician to discuss the options of, the available options with the patient, that's fine. 2, clarify the potential risk to subjects, injected cells lodging in the pulmony vascular system. 3, chair if I the plans for patient long-term follow-up and their responsibility for this follow-up. 4, clarify on whether the subject objectives include dose-limits toxicity and clarify the dose-limiting toxicity in the protocol and the specific criteria. And under informed concept, the ic be modified to increase understandability. Now, I'm open to other suggestions. This is fairly complicated. Let's take the easier areas. Is there anything people would like to include on the clinical piece? Any guidance for the construct of the clinical protocol? Dr. . Epstein?

The I would just like to suggest that investigators consider using some imaging modality late after treatment to just rule out the unlikely possibility now that there is hemangioma formation.

M'hm. And what image no dal would you --

You well, mri, ct or angiography, which ever is most convenient for the center.

Why don't we leave it open then, consider the use of animaging technique? Yes, I would certainly leave that open to the investigators. To monitor the potential risk of hemangioma. And we can exix the -- fix the wording of these. Dr. . Lowe?

I wantd to ask a question to the reviewers. One of you raised a concern about the screening for cancer on follow-up after the intervention had been delivered. I think that might have been Steve. I just wantd to ask if you were satisfied the response .

Well, I felt that it was rather unclear still. It sounded like it was something they were going to be discussing further among themselves. Soy think it's appropriate we should put something in there to indicate how we feel, at least.

Put that after long-term follow-up, clarify plans for long-term follow-up, responsibility, include appropriate follow-up for malignancies, okay. Now, pre-clinical, a little more complicated. I tried to simplify this, rather than -- and theretherefore, there are only six items.

In general, not in specific, but in general, with a phase 1 study, how is the rigor and the efficacy demonstrated in the pre-clinical studies play a role in ind for a phase 1 study?

I'm, actually, we're quite fortunate because our chief of toxicology is here.

Depends on the patient population, the mode of delivery. If you're going into the Blaine or the heart, a very invasive procedure. There are more concerns for the potential of efficacy. this case, being the injection is, the efficacy is helpful, but not an absolute to win the phase 1.

Thank you very much.

The other comments, additions, deletions? I'd like a motion then.

So moved.

Second?

I'll second it.

Dr. . Vile, weber, powers, rosenburg, Lowe, Simari, war, pat, no, deluka?

Aye .

Aye.

All right. Thank you all very much. Readjourn. We are beginning tomorrow morning at 8 a.m. sharp. Many of us have flights for early-afternoon, and it's important that we end on time tomorrow, so we can catch our flights. More important now than it would have been even year ago, because planes flying full, and it's difficult to rebook. So I'd like everyone here at 8, please. I appreciate everybody's working hard for the last few hours. This was complicated. I thank you very much. And those of you who are members fortunate rac and others in the audience, dinner tonight is at mikanos, which is on Congressional lane, we'll meet in the lobby of the hotel at 6:00. We'll be arranging our own transportation, but I understand that some of of the staff are going and can provide transportation for some of us. So 6:00 in the lobby. Thank you.