Good morning. Please standby for realtime transcript. We will reconvene in a moment. I'm just waiting for the slides to be loaded. Good morning everyone. We're going to reconvene the December 2005 RAC meeting. Good morning. I have two announcements this morning. Quiet. Please sign up for transportation to the airport if you would like transportation at the registration desk outside the doors. If you drove to the meeting today, please go to the registration desk for your complimenttary parking vouchers. On that note we will open this mornings's session with the transduction of upper airway ep they'll yum air ways in human placental Dr. Dusty Miller will present. Dr. Miller? >> Is this on? Yes. >> Thank you I didn't see you behind me. >> Moira L. Aitken is also here. >> Will Dr. Moira L. Aitken be presenting this morning? >> Yes I'll start out with the vector presentation and biosafety data and Moira will follow-up with clinical presentation. This is recfor--i'm going to review a little bit so everybody's on the same page. It's a very similar AV recfor that has ACDA there are no other genes and then it has the AAV terminal repeats. This is packaged--this is basically an AV2 vector much like what you've seep in the past. It's packaged using AV two rep the only difference between the ones in people is it has a AV6 capsid. Why we are excited is the lung ducktion that's we put anywhere from 100 to 200 micro liters into the 2 nose inhaled spoontaneously. It was AP positive cells. You can see a lot of transduction in the airway. A closeup up here shows many are transduced in the airway the major in the lung. There are a few blood vessels here, here and here and prangca ma is down there. This is where we want to be for CF. The nose of these animals is transduced all this black staining is AP positive cells in the nose. They get it by naisele insulation by this end we can show good nazele transduction. This is experiments that we did with human naisele polyps. Here I show AV2 with con pair son the trabs transduction with AV6 is better in the human samples. To back up a little bit we have proposed an AV2 trial and now we're proposing an AV6. One of ideas would be to compare these in the humans. We would like to know which vector works best if either at owl in comparison what we've seen in mice and rats. We've done a preclinical safety study with AV6. These are the details here. Wee used rats. And the reason we used rats instead of mice we you used them both in the first study with AV2 the rats give normal levels of serum parameters and are generally healthy. We had problems with the mice in the last study where the liver enzymes bobbed up and down. It want related to the vector administration but you couldn't tell. We used rats 150 grams when we start. Administered these doses her 0, 3.3 times 10 to the 9th. 2.2 times 10 of the 8th vector gene Nomes per gram. CF subjects are quite a bit lighter than normals. We're 3 planning to do 0, 10 to 11th in the first group and 10 to the 12th vector genomes. 2 times 7 vector genomes to the gram tenfold less the max there. We examined them day 6 day 60. We looked at DNA the lung, blood and go nads and did historiology. All of these were uneventful. The serum chemistry was find expect one parameter and the CBC was normal range. Histology I already sent that to reviewers it was fairly complicate but nothing that suggested a toxic vector response. There was sign usite tus and things like that. I don't have times to present all of that. This was one area where we had a serum chemistry fortunately in the right direction for us. Here's the date for the serum chemistry in the blood. You can see in in group everything was in normal range. No difference between the different controls. What I'll showing is control, control male, this is a low does male group, high does male group, control female low and high dose. There is no effective dose towards higher enzyme levels here and they're. That's day 3. Day 60 we had one group a bit high control males showed high ADAST. None of the vectors showed normal level over that. that was the only variable that we had that was significantly different between the groups. As far as the DNA I have to explain the chart. We looked at lung, blood and gonad and the des sick Nateors--designateors is the high dose female 1 through 5 in the animals. Here's the high dose male and we looked at high dose for genome 4 transfers. You can see high levels of the DNA in the lung. It's variable with the section of the lung where it winds up. Sometimes you get a low darkly stained sometimes not. You see high mass DNA in the lung. If you look at the blood then, two animals were positive at purple bars. The rest of animals were at the background levels I indicated that as 14 copies per microgram DNA that was limited detection. In gonad all gonads were negative we were below the limited detection of SA14 copies per micrograms in DNA. In the day 60 animals, the lungs were positive, less so than day 3 as you would expect for DNA and all of the woods and gonads were negative. There was no signal above the copy per microgram protection of as say. Given we're going if the lung very little escapes in the blood and we can't detect anything in the gonads. That's it for the biosafety vector design and Moira was going to present on clinical aspect of the disease. >> Thank you very much for your review of this protocol. I'm in the department of medicine in the division of critical care medicine. So unlike your other protocols that you are reviewing at this meeting Cystic Fibrosis sis is not a is--epithelium cells of the organs but the vast majority patients die of the respiratory disease in this sued we'll use the sinuses as the lungs because of easy of access and it's been used in other gene therapy trails. As I said, the survival of Cystic Fibrosis is improving with treatments of anti-botics and patients survival remain 35 year-old and 5 the need for novel therapies in cystic fib brose sis. So or ren at a time the protocol will be presented in the next three slides. Our objective was to determine the safety, transduction efficiency and emmun response associated with the delivery. A double blinded placebo control randomized trails. The escalation to the second level will only occur if if there no safety concerns at any stage during the first dose level. The sample ceise there are two dose cohort per subject cohort the placebo 3 will perceive the vector and total of 8 subjects if both levels are study. The subjects are patients with Cystic Fibrosis if will be dealt until 50 years old and subsequent can cohort 10 to the 12th vector genomes. Our outcomes from the study are primarily safety. Secondly looking at the serum neutral lizing anti-bodies and he have efficacy outcome is--subject inclusion criteria aged Cystic Fibrosis aged 18 to 50 years, their lung function has to be relatively good. We're taking 40% predicted lung function at screening. Their oxygen saturation has to be above 90% predicted and serum negative for the A section and use contraception for the studied that's 190 days and have to give written, informed consent. I haven't outlined the exclusion criteria which is longer. In the preliminary work-up towards the study we looked at how many people are serum negative or positive. 70% of adults patient are serum negative of AV. This f study of almost 200 patients. The first three have Cystic 6 Fibrosis and we looked at a group of normal adult. As you can see in the Cystic Fibrosis adults which is the this population the pink or purple darker purple are AV6 coming here about 25% and AV2 is more like 27%. We're being conservative in saying the catch of the patients will be about 70% of the adult patients. We are planning to deliver the vector as we have historically with other vectors to the in fearer ter banant this is showing the mild. The cat der is placed 2 to 5 back from the noztral we know it's expressed in that area. 100 to 300 micrograms is delivered over a 30-minute period where the patient either upright or in a semi-recome bant position to that they're comfortable. An issue that came up that there may be spillage up here to the ol facry never up here--there is no AP expression in the brain in rats. Again in preparations towards this trial we wanted to make sure we could get adequate biopsy samples and sampled the in fearier ture banant for five patients for other reasons this good sample tissue was obtained I'm grateful for these samples from the University of Washington. You can see the epithelium is shown here and no in dodgeious staining. We feel we will get it for our measure N. response to some of the questions that came up from the committee, the sample size justification. This is a proof of concept study to determine if AP can be delivered to the upper airway. We think it's a pilot study to guide the AVCSTR trails, future trails. It's a 7 machine nisic studied nod powered to produce a statistically result. The reason for this AV6AP is not a therapeutic vector and wanted to limit the subjects in this precious patient population we use. In summary the number we chose is a compromise to ob taint enough scientific data but to minutenize the number of the subjects and my patients with non-therapeutic vector. AP expression, we predict that expression of AP will increase over time. We're doing our biopsies at day 7, day 28 and day 112. We predict day 28 and 112 will be similar but more than day 7. We predict that the transduction rate will depend on the vector dose, a higher dose will have for transduction. So the transduction rates will be at the higher dose level. AP expression, we're going to primary outcome measure is a percentage of AP positive epithelium cells. We have designed a strongly positive response. If any of the three naisele biopsies taken at each of the study days will be greater than 10% of the epithelium cells A. dose will not extend on the lower dose transduction rates. So if we had a relatively high transduction rate of 50% we would want G on to the higher dose. We do have a proposed amendment if 2 of 32 study subjects are strongly positive, in other words they have more than 10% transduction, the study will be considered a success and no further patients will be enrolled. However it no subject or one subject is strongly positive at the higher dose then 8 we wish to recruit three additional subjects, that's a total of six we wish be able not just study three with one placebo but have three at that higher dose. We would like to the committee's permission to recruit more patients with a equivocal dosing. Our future studies with AV6 is we would like to procedure to AV6AP in the lower airway recognizing the upper airway may not be a good surrogate. We would like to procedure with AP6TFR with nazele and ultimate goal is AV6CFTR into lung as a potential therapeutic measure in patients of Cystic Fibrosis. That's what I have to say and Dr. Dusty Miller and I have happy to try and answer any questions you may have. >> Thank you both very much. We're going to go through our formal review process and begin with Dr. Dewhurst. >> I would like to thank you both for a very informative presentation and also to the responses to the questions we raised. I want to preface my comments by saying this is a very different trial than some of the ones we heard yesterday whether there was a lot of discussion of potential benefit and risk. Obviously in the case the benefit is very indirect benefit. There is no potential benefit to any of the study participants here. And the other aspect of the study that's unusual is your initial choice for the surrogate tissue for the one you're interested in down the line. It's some of those areas I would like to discuss I think. And some of the future plans to get a better feel. 9 Some of your responses indicate requirement for more information from regulatory bodies. I think it would be helpful for us to understand a little bit more where these studies are at and how compareable those data sets are in this particular study in terms of design and that sort of thing. For the record I had a number of questions many of which you have answered very satisfactorily. And the first one is you eluded to in the presentation related to the potential of transduction of cells in the ol facry bulb and transduction of cells in the central nervous system. I think the responses were adequate and concerns resolved. A second question and general interest about the AAV6 receptor and you indicated there's nothing known about that and that's perfectly acceptable. I had a question about focusing exclusively on the nazele route of delivery in your response you indicated that you're considered at some level that the nazele epithelium is all the regulatory bodies will allow you to examine. I'm curious if you can elaborate on that specific term. Clearly your long-term goal is to go in the lung. One of my concerns is what you eluded to Dr. Moira L. Aitken is the precious population you have one of the potential risk is these individuals my convert to AAV6 and a consequence where your vector prove in the future were they exposed to the AV vector may find themselves at a disadvantage down the line. I guess I wanted sort of some feedback from the both of you. So there's some historical 10 context to that. We initially propose to the norms and look at transduction rait in the nazele epithelium and lower airway. The tissue is more relevant to our ultimate goal. You can argue both ways. We had the same problems with CF patients immunizing them against A6. It's a subjective position and basically the RAC says we do CF patients that's why we're here with CT patients again. That's a reasonable argument I agree for you the down side for the patient is not what we'd like. >> In terms of--you know one of the issues at the heart of protocol is the extent to which the nazle epithelium is valid surrogate for lower respiratory tissue. At least in the human it sounds more or less an unresolved question. Is that accurate to say. >> It's a epithelium and looks like air ways there have been a lot of studies done in the nose with add know viruses. You could measure from there to here and we can go through a successful trial and measure a change in the epithelium. It's a reasonable epithelium. The problem with the lower air ways it's more invasive and harder to sample the tissue and quantitate what you have da duced. Maybe you would like to follow up on that. >> Absolutely I agree with dusty. I agree with him completely. The issue with sampling the lower airway did the that the patient have to undergo bron cose Topi and requires seddation. We have shown and phase 1 trial of that many of the serious adverse events were thought to be 11 secondary to the bron cose pea procedure itself rather than delivery. I actually think using the nose as a surrogate is a very good first step toward that. and regarding the serra conversion which I also had an issue with, the consent form as proposed puts that in there but I feel strongly that we would like to procedure with this and--precede with this and speaking for my own patients I feel they would like to proceed better for some to convert than for us not to proceed at all. I'm not belittling that issue. It has led to hours of discussion internally in our own group. >> You may remember too targeted genetics are conduct target 2CF in humans. We don't know in the transto transduce cells. We no very little from the study as far as why it didn't work. CF disease is highly variable and hard to get an efficacy measure that you can use on limited patients. We would like to know it it work first and then transfer our measure to CFDR. >> All right, thank you. I had a question for plans direct comparison for AEB2 that's contingent as you indicated upon resolution of the questions from the FDA with respect to animal studies on that. I wonder if you would add on that. >> We're in the sticking point with the FDA in terms they want an additional animal study which we don't have the money to pay for right now. It's a huge studied takes a lot of effort and all the controls and laboratory procedures and so on. Our position is sort of AV2 has been used in people to a large 12 extent over 100 people why you would want for animal studies. It showed no bad efforts but the problem was we had variable levels of--they bounds around with the mice it was a problem too. >> I would like to kind of make a comment here. I think there's a number of issues involved here so I'll try to point out maybe 1 from the FDA's point of view is that although--i'm not going to go into discussion about animal models for AAV2. I think one of problems the FDA had is if you're doing a direct comparison you do exact protocol where you look at them and the issue will come back up again with a lot of the comments I've read it looks like you have to two separate protocols if you're comparing them why you aren't comparing them together. >> This is a new one for me. Doing them together would be I think time tense difficulty because of where we're at. If we can't get one protocol approved it's hard to do two comparisons. I think one of the concerns of FDA they felt that our data suggested AV6 was better vector and why do AV2. Let me finish. On the one hand we have the scientific question of which one really works better in people. On the other hand we have it that AV6 works better. I think the argument could be made to with AV6 and see what the result are. >> That's fine if that's what you feel go forward with the AV6. I have a question while I've got the floor. Again it looks like you're going to do there in sort of the multi-steps. First go in the 13 nazle epithelium to look for the transduction efficiency and not tell you if it's hitting the target tissue in the lung. The next step is the lung. Now you have the two separate patient populations to have potential to convert and not until the third trial that you go in with the therapeutic gene. Is there a reason why with all safety, safety studies is there a reason why you wouldn't go forward with the actual therapeutic gene. >> Part of reasons for not moving quickly is does not fit well in the AV vector it's too big. Tough do something to solve that issue. The way we approached it it so use two vectors to allow a combination to give the full length CRT gene are promoter. It was never clear in my mind whether that worked at all well in subjects. So we got a little bit of problem with CFTR approach. We're working on that. We've it going well in mice and better in rats. We're doing the CFTR studies. May be the best approach would be to take the dual vector approach and put if the nose and show it works in the people. What you're asking for was the original RAC review why don't you do the whole thing from the start. I think we have to to take a step-wise approach. >> If the actual product changes dramatically or enough that the FDA things it has changed considerably you would have to do animal studies to do know I'm thinking to udefense--why to the develope that system first? >> If we wanted to do CFTR we have the two-vector system. We can bed with that if we get in 14 the patients we can't so the affect we don't not what the problem is we don't know if Idaho a blocked poor CFTR expression. It's a measured approach. We may not go to the lower airway with the AVPA vector. I think it depends on the results of nothing whether the nose is transduceable with this vector. It's epithelium makes CFTR and a reasonable circuit for the airway. We know in the mice and rats we get more transduction in the nose with higher doses and more transduction in the lung. We may be surprised. I mean if the results were negative in the nose, who it suggest we ought to go to the lung? We don't know until we get more results. If we don't take that first step, it'll never happen. I mean have you arguing we should to with CFTR right into the nose. >> I'm not arguing anything it's your trial I'm asking questions why this approach. And my concern is what you see in the nose may not have any relevance in what happened in the lung because you know with the most of the case with AAC2 there was not transduction in the lung so even if you see it in the nose it doesn't actually say it will be transduced in the lung. If you get positive that doesn't necessarily say you will get positive in the lung. I guess I'm going in circle here. I would like you to think about how you are designing this experiment doesn't mean you will get what you are. >> Just to show transfer. One of things I approached with targeted genetics why not put an AE2 vector at the same time of 15 the other vector to express AP. If we could show AV2AP is safe in the nose we might think of that as a marker for the actual gene transfer. Do you see what I'm saying? If we had a CFGR vector we put the lower airway and spike a ten fold lower dose and get a transfer of that. >> I don't want to turn this into a an argument. This is clearly a discussion between you and FDA. It just wanted to bring it out because I had questions and I hope you'll be able to help answer it. >> I certainly appreciate the discussion because it highlights the issues in a protocol and that's what we're here to discuss. Dr. Dewhurst I'm turning to you as first pry primary reviewer. >> That summarize the comments I had wren and the patient enrollment and numbers many we have a number of members in the committee here who may have enput on that I'm not an expert on that area but I would be interested in hearing what my colleagues have to say. I don't know that I have remaining questions or concerns myself but I am interested in sort of hear what the group feels on the broader issues of the nazle target, the use of the vector that is not the end game versus CFTR those broad issues. The broad issues as you eluded to as a combination-based strategy which is different than the strategy you're piloting at present. At least in my mind it seems reasonable to want to validate the activity of the AAV6 capsid transduction and go to the strategy. It's a reasonable approach and I'm 16 interested to hear what others have to say. >> Doctor Helen Heslop. >> I would like to thank the investigatingors for their response. My first couple of questions were regarding the the clinical study did he seen and I asked for om rationality in choice of sample size and discussed in more detail how the study would be used to guide future investigation. I think that was done in the supplement they provided in the presentation this morning. I asked for more justification of the choice of study design between the C bell and study agency and eng that was addressed in the presentation. My next question was biodistribution after nazle distribution that also I think was addressed in the presentation. My final question was same one raised by other doctors about whether there was preclinical data to suggest the result with the nazle delivery can be extrap lated for future studies and I suspect there's nothing else to add. >> Thank you Dr. Helen Heslop. Our third reviewer is Ms., quan. >> I think a lot of the questions I had was raised by the previous two reviewers. My concentration was on trying to get these technical explanationses that were ch changed at the table into the language that is understandable by potential participants in particular because as the discussion showed it's possible that subjects that would agree to participate in this kind of trial which has no possibility of a therapeutic benefit, 17 whether they could understand it is position they're being placed in with the potential of not being able to benefit from a treatment later developed. And I think with patients of this type where it is not an immediately fatal disease but certainly with a very difficult outcome, a lot of the potential subject population might have a very al true ristic view of participation and I think it requires more really careful explanation of just where this particular experiment fits in a long-term model for trying to find a real therapy for this disease and to really ensure that the person is able to wage weighing the farther force before agreeing to participate. In looking at the informed consent document I really don't think it does enough to really layout those things in terms that a layperson could understand. And as we have had other forums here regarding informed consent, I think probably there also needs to be instructions that are delivered orally and with films and or with other things that are not just confined to a paper document to explain this. But I think the discussion here at the table really raises an issue that speaks to a role for the RAC an evolving role for the RAC. I think at this point we really need to look at the possibility of the RAC's role in developing study design models so that the statistics and the ethics can be kind of summarize and put into sum guidance documents similar to the informed consent guidance that the RAC did within the last five 18 years so that each of these kinds of po sal--propo sals especially combines the risk of the participant not being able to benefit from the future therapy that these kinds of things really would best be handled if there were some concentrated focus on that to bring together and provide the entire community with consolidated document or guidance to help people make these disution decisions in a more informed manner. Part of my recommendation is not just for this particular study team but to really urge strongly the study design work group really be staffed and assisted in doing some very important work that I think will help the entire community. And I would like to hear a little bit more discussion here at the table with regard to ethical considerations for this particular study in light of statistical weakness and the potential for hindering the participants from further benefit. >> So the ethical question seems to me to be how--in the context of a study which is in ways uncertain, then how does one consent a subject--now this subject is an adult not a child, so the consent, the entire consent the ethics of consent are different but how does one do that when it is very likely the that the participation in the current study is diagnose to exclude the subject from what we all hope will be effective and efficient future studies that's really the quandary I think. >> Yes, and the further dimension being that eats not 19 clear there's sufficient statistical power to make the study move the idea forward either. >> So Ms., Kwan, would an informed consent document satisfy the quandary or dilemma for you? >> I think a more thorough discussion that's what is currently in the informed consent document bold also strongly suggest that there also be a plan in face to have face to face discussions in addition to the written consent document. >> Thank you. >> A couple of comments with regard to that. >> Please. >> I keep on hearing this statistical power is weak in therefor the study is not worth doing. How would you have us design the study in light of limited patient resources and valuableness? It's a test bodies like this want it test you if you don't have the answer you're in trouble. If you can tell me how you would like to power this study to we can get statistically meaningful results without damaging more human subjects than we need to, I'd be glad to consider it. Especially if you're thinking about writing a white paper saying this is way you design a trial how would you redesign our trial to get a statistically meaningful result. >> Dr. Miller I'm not sure we can provide a specific answer to your question Dr. Steven Piantadosi asked to speak. >> Let me answer our question with a couple of questions. You state in the procall the primary purpose is safety. I'm going to put aside the considerable 20 issues about context of the studied subjects sur rendering and their eventual possibility of having a therapeutic vector administered. Let's suppose there is a safety concern that is yet unobserved, I mean that's the purpose of doing the trial. What would be an unacceptably high frequency of such an event? >> Well we design it in phase 3 but Moira could address that better than I can. >> What frequency would be high for this yet unobserved safety event? >> So as--as this committee may or may not be aware the Cystic Fibrosis patients because of their multi-system disease have many adverse events. I believe the number for phase 1 study was 242 adverse events. On a shorter term study. I think when we're looking at this small number of subjects, we're really looking at serious adverse event three or four toxic cities and this would will immediately reported. >> You are referring to the adverse events that's a consequence of the underlying disease. I'm asking about intervention, the gene. If you atribute a particular unobserved event to that gene what would be an unacceptable high frequency of the event? >> It defends on if the event is death it's 0. >> Okay, let's take death. What's an sun unacceptable high frequency of death. >> Patients with Cystic Fibrosis do die. >> They do not die from this gene. >> They do not die from this gene. I think it's so often 21 with experimental treatments not so request gene therapy. The University of Washington is one of the factor development centers for a multitude of experimental treatments. And in shows studies sometimes the protocol design allows the site principle investigatingor to determine if an adverse event is likely due possibly due to the therapeutic intervention. So I think that as a site PI I am extremely cop servetive about that in my possible or probable. I think with any especially with this particular step toward a therapeutic adventure, we would be, I would be extraordinarily conservative in my approach as an adverse event. Not only for the reg tribodies to--regis tree bodies and report back to the five regulatory bodies but individually one would be very conservative about any adverse events that one thought were possibly due to that or even a relevance. >> I accept that but you're answering a different question than I'm asking and I can't let you off the hook because it gets to the fundamental principle about a study design. You're saying what you'll do if you observe thing and the attribution of the events. What I'm asking is if you have a solid attribution of serious adverse event of the nature that might terminate the development of this vector and you attribute that to the vector reliably there is some frequency that you would observe that that would be unacceptable clinically if 25% of the subjects died for example from a cause attributable to the vector anaphylaxis or 22 overwhelming systemic in fession or a mucous sinus attributeable to that. How frequently would you have to to see a disasterious event is a high frequency 10%, 50% of the subjects? What would that answer be? >> I think it's interesting because if we're talking about four subjects of whom three are getting the vector. We're talking about three with a dose. A statistical answer--this is a machinistic study. I don't understand your question. If you're asking about statistics . >> Let me ask if T about another way. Forget N equals 3. What if you had a thousand individuals treated this way so you had a very large number. Just do the a experiment in your mind. If you had very large number subject and you observed the certain frequency of adverse event. >> Inflammation in the nose? >> I'm talking about you would see as unacceptable developing term Nateing event 10%, 20%, 30%? What do you think it would be? >> I'm going to have to ask for a clarification. You mean of this particular study? Or in future studies? >> I'm talking about this product, this vector. There is some unacceptable rate of complications from this that would terminate development. That's my point. >> I understand. Let me help Dr. Moira L. Aitken just a minute, just a little bit. In this with this particular product with the AP gene, there is really no hope for any benefit; therefor, at least if I 23 were answering the question I would say I wouldn't accept very many if more than one serious adverse event directly attributable to the gene and gene product because there's no possibility of benefit. On the other hand if I put myself out five years and there's another transgene out there that has a high likelihood of cure, I might be a little--I would have very different answer. But if Dr. Steven Piantadosi is asking about this gene product to the upper airway, then would you be able to handle or tolerate as the hands-on physician, more than one? >> I think that's an excellent response to that. I'm just thinking ahead. A serious adverse event might be a hospitalation of a patient and--for example, in doing these who were performed by an Oto largeologyist would not accept a nosebleed. I could imagine there could be serious adverse events. Should we stop the trails because of that? That-- >> You're free to define adverse event anyway you want. After all that's the objective of the trial and safety. >> It's clearly outlined. >> I understand. I think the point operative point from my prospective of dope--dean sign your tolerance of serious adverse events is low you have a particularly low threshold. And whatever the complication is that you haven't yet seen you don't want to see very many of them because of the context, right? >> That's right. >> But the point about study design is that the study doesn't 24 permit you to rule out relatively high frequencies of side effect. For example if all three subjects are treated absolutely safely, you only rule out frequencies as high as 70%. 70% of your subjects could in fact in reality in that larger 1,000 patient study that I high hype sides. 70% could see it and 70% of them could die. >> That's true. >> And you could get three that don't. >> That's correct. >> That's the problem with this study design. That's the problem. >> But phase 1 trails are often like that. >> I understand that they are. But it did you want help that phase 1 trails are lousy designs. It doesn't help. >> But I think, if I could interrupt here too, I think there's a corollary in the other direction because with only three subjects that actually received treatment you could have one adverse event resulting in the person dieing and it might represent really a very, very low frequency. You could throw out the entire results and say no, this is unacceptable we don't go forward with this and it actually could be the road map to an efficacy shus therapy. The things that could go wrong you could get a death among the three that actually received it. It could actually represent a much lower incidence than apparent. You can stop the entire study because of one death and lose a promising therapy. That's the other part of it. >> Of course we can always be 25 victimized by chance in either direction. We need to plan the study that we are not victimized to a great extent which was any concern with the small sample sizes. Don't get me wrong I'm not suggesting that you have an early developmental study that has dozens or thousands of individuals on it. What I'm suggesting is the difficulty in the investigatingors in thinking this way has led them as is typical to use a design that is an off the shelf design that was never intended to answer the basic safety questions at play here and it's made worst by the fact that the subjects who participate in the study are probably giving up their ability to derive a benefit from a similar therapy in the future should everything work out exactly as we hope. So what I would look for and the way I would turn around the question that you asked me about study design is that the protocol needs to reflect some of this kind of thinking and if all the more critical in this context. It needs to reflect your ideas about what you might miss because that's the error that's going to be made here is that we're going to miss something important based on successful application of a treatment in three subjects and how we interpret the absence of information when we have a denominator of three is exactly what the problem is and it's exactly the flaw in almost every subject--every study that's reviewed by this committee. >> Can I respond to that briefly? >> Briefly. 26 >> If I see where the problem is and I understand where you're coming from we listed safety as primary outcome. I think the reason we listed it that way we want to sort of state that primarily we are concerned about the patient's safety we don't want to allow adverse events. Our primary goal here is not to compromise the patients but really the primary scientific goal is to figure out transduction rate of AV6 vector in epithelium. I think that's our primary goal I think with three subjects in each group get that information. That's what we're after we like what it does in mice or rats and show if it works or doesn't. If we find safety problems. >> Stop. >> Okay. >> Go ahead I don't mean you stop I was trying to figure out what I was going to write. Go ahead. >> Really primary outcome measure the reason we're doing the trial is to figure out if AV6 works in epithelium. >> I understand that and I'll be quiet after this comment. You can't get off the hook in a direction either. Because if you have three out of three successes then you only ruled out success rates in excess of 30%. That may or may not be appropriate but it cuts both ways. >> But there we're talking about strongly positive is greater than 10% I expect we'll see some transduction rate if all subjects. That's what we see in the animals. We get a quantitative number. >> I gave you three out of 27 three. >> Dr. Steven Piantadosi I would like to reaffirm Ms., Kwan's comment is that this trial is another example, for me at least, where the current construct for phase 1 and phase 12 studies doesn't work. We desperately need for a field like gene transfer a new construct that's one of charges to the RAC to develope such new construct. I would just put it out to our entire group. We've been talking about it now for the last two sessions. We've had two out of four studies this time. We're at directly applies and we need to move forward so that we can provide our investigatingors with answers to questions like the 1 Dr. Dusty Miller proposed to us. There are other questions Dr. Strom you had hur hand raised earlier. >> Thank you let me state for the record I'm not an experiment in the CF transfer in this area. But I'm an oat to largeologyist. Having said that I think it's unfair and wrong to discount the import of the epithelium on this disease as an intern price in and of IT. Folks with Cystic Fibrosis one of the most devastating things that happen to them is their nazele systems and often difficult to control and disease. If you can help their symptomology you can help them whether through better irrigation, exposure etcetera. Look at the nazle epithelium I would encourage you to not say it's a certain mark but perhaps I'm biased a target point in and of itself. If anybody has biopsied a turban Nate, you get a certain amount of bleeding. It bleeds like crazy in CF. I 28 would say that it should be a limiting factor because you'll exclude everybody. You're going to get bleeding. It's eas to control and I would just stop with that and know that it's going to happen. >> Dr. Albellda. >> I'll put on this hat instead of cancer hat. I think the nazle epithelium is the good target to start with. Eng there's lots of data. You mentioned some of it. They are sillyated epithelium. If you can't affect the nazele epithelium you won't have success with the lung. Eng it's a reasonable place to start and probably easier than a bron cose top pea. I would like you to think about and person opinion to discuss the idea of using normal subjects rather than CF patients. I got some sighs, I'll bring it up. There was very good point that you sensetizing a rare population group that's not eligible to do it. I think when you're looking lung gene transfer in CF patterns it may be different than normal patients. The nazle epithelium I don't think is so different. If you can't transduce normal patients I think it will tell you right away you better find another serum type or something else. It is a big controversial. There could be big side effects. Is there are another way besides biopsies to get reasonable data? Maybe Dr. Strom can answer that. You do want to look at the surface epithelium and I think you can get a decent amount of cells if you do scraping N. normal patient I would imagine a lot less bleeding and 29 inflammation. It's difficult to recruit CF patients. They're sicker. You're going to have much--as you mentioned the adverse event is nightmare because they have million things going on. I personally don't like the randomized then with one placebo patient and three treatment patients the statistical power is 0. Why bother doing it? May be with the CF patients with lots of side effects with one placebo I don't see how it helps you and puts the other patient the risk for bleeding and complications and especially CT patients without benefit. We can discuss this. The FDA may have strong feelings about this. I don't know. This is ideal study in in my opinion that is strong trial. There hasn't been any type of complications. I will view this as very low risk and low enough risk that I think normal volunteers would be a reasonable group. Now people may disagree but it's something to talk about. >> Dr. Strom what about scrapings. >> You can do brush buy ob op sees--biopsies in the nose. I'm not familiar with how much epithelium you would need in order to get your answer. I can't comment in the terms of sheer volumes of you can get epitheliums with the nose brush biopsies. >> It's nice with the sample you can see what percent of the surface epithelium what cells are involved with AP you can tell what cells are positive. With brushings I imagine you get a mix of cells. >> The bleeding is usually very 30 easy to control. I mean it's not a--it's usually not a problem. So I think you can do it. You are going to get bleeding whenever you take a biopsy in that area but it's reasonably easy to control if most situations. >> You get about with brushes I've even done it to myself as well as others you get to about 10 to 15 cells we can get a costain with sigh careton. If we are going to do with limited number of patient we would to get the ideal--you can see it clear but the others would lose--the idea with the biopsies is to get more specific which cells are being transduced in supposed stem cell out of the basele cells or is a termly differentiated cells. That was the idea behind the buy ob sees. I personally would still prefer them. >> I don't think it's unreasonable. The other problem you will have CF patients are nazele problem and trying to find the turban Nate is not a trivial task. You can get pollps all over the nose and identifying the turbannate is different. Your going to have a hard time. >> With delivery with one of the half dozen centers in the United States that can do the difference measurement so we're--we're familiar with that area. And then the Otto lar renologyist is who you use and we're getting the most expertise in that area. >> Doctor are you comfortable with the response? >> I would like to hear comments are the panel what do they think about using normal patients. 31 >> No, I was going there next. I meant to the response to issue of brush biopsy? >> Yes. >> Thank you, let's go next to the Dr. Arrestbellda's second comment was that prepares the investigators should reconsider conducting this study in normal people rather than in patient with CF. >> Can I ask a question? I should be more aware of that. Investigators said they came to the RAC and the RAC were the people who we said that you should not use normal subjects. I actually. >> That is correct. >> So I actually applaud this trial. I think the problems with gene therapy is we don't know when something doesn't work, why it doesn't work. It this seems to me a very good step-wise approach to try to work out are we getting transduction, it's not in the best tissue and so on. Then we go on from that stage to the next stage and so on. Could I ask the investigators if you had the choice, would you use normal people or CF? It sounds like you wanted to use-- >> I think we're right in the middle actually. We argued a lot--we talked about this a lot before coming here. That is somebody could argue well when we you'd the normal we haven't got the right target epithelium if you cop pair AV6 to AV2. May be AV 2 work better with something under stress. If you take from the stained standpoint that AV6 works we can't do trails and take the best horse and ride with it. It would go best with norms and say does 32 this work well and move on from there. >> At the risk of the RAC is pushing you from here to here and killing you on both issues. >> Patient accrual would be much faster we wouldn't do CF patients there's a lost of pluses going with normal. Human biology is human biology and receptors that are there are probably not going to be that difference. >> I'm not expert in the normal epithelium and CF in anyway and I can't address that. It does seem to me that Dr. Albellda's point is a good 1 you can tran duce the cells the issue of normal subjects. >> To the RAC's point the last time I apologize that I didn't review the transcript. But there I think the main issue is an ethical one which I actually personally did I agreed with. But just to revisit that so that the viewpoint was that ultimately if we succeed with this venture that it would benefit that patient population and therefor that patient population should provide the study subjects. Not--to my argument was of course that 1 in 25 patients are carriers so actually you know I am a study subject because of you know my ethnicity. But that was--so that was the argument, the ethical viewpoint of RAC last time. >> I do apologize I should have this in the memory and being on RAC. >> You weren't here this is four years ago. >> Oh, good that get me off the hook. Can I ask at that meeting was the ethical issue of what's 33 coming up now if you have patient who is enrolled on this trial with no benefit then excluding them from any future therapeutic benefit was that considered by the RAC at the time? >> It was. And so the other argument had taken precedent over that. So Dr. Miller and I at that time wanted to go with norms and we put that in our protocol and we were instructed that could not happen. >> May be a good way to procedure is you consider us using normal. We don't know how the rest of regulatory agencies might behave. >> It's just suggest. >> Could I offer a potential compromise? You know the debate about about normals might lead you to say the study should include both and safety information gathered in normals and small cohort CF patient as well. >> Start with perform and move on? >> --normal and move on? >> That's a possibility it would meet both of objectives and of course you would get double the criticism probably. >> Let's go quickly to Dr. Powers. >> Ms., Shapiro and I were thinking through the framework here. Mar of what's at stake I think if we can try to lay this out as this, when you have a non-therapeutic or something that's not going to benefit the subject population in adults, you either then got to have some generalizeable knowledge and all the things raised become very important and as Ms., Kwanponed out you have to ask if some of 34 the basic powerings not really going to be there is there something coming forward that you can describe as generalizeable knowledge? The more problems you get in this clinical design the weaker our case is on the benefit side. That means all the kinds of risk on the other side you can tolerate far less. You know the scales get tipped. Weak generalizeable knowledge, weak case for going forward. Less ability to tolerate downsides. In this case you now have a down side for you know it's may be speculative because it depends on what the success downstream at least some groups of people in the CF population are under a new disability. They are persons who will no longer. They will sacrifice their ability to take advantage of whatever benefits might come downstream. A bit speculative non-the less a--none the less a real worry for them. If you were choosing between normal subjects and those with CF you would first of all ask yourself the question in the choice of those two populations do I loss or gain anything in terms of any capacity to generate generalizeable knowledge? If it turns out for example that the in choice of two populations you didn't have any difference in generalizeable knowledge then it might seem reasonable to ask normal volunteers to participate in this and then you certainly don't have that added down side. I don't know the answer to that question. But it seems to me that's the framework for answering it so that if the CF patient population still 35 provides you some prepares weekly general riseable knowledge not nearly as good as we would hoped for as committed on in term of safety and benefit. That therein lies the frame work and Ms., Shapiro you may want to pop in there as well. We were mumbling among ourselves. >> There has to be a way from FDA just a brief--what would be yours--in this is a first take. >> I'm not making any statements about this. What I'm going to say for the record the FDA entertains all sound well designed clinical trails. If you have a justification, a rational that is is found for why you're going to go forward whether it's with whatever patient population and I think the most important thing here which you know I didn't state but I think everyone understands is that many ways we can get around some of the safety issues is by doing good animal studies which will give us a better feel up front. I know there's not always great models but that helps us make decisions when we have uncertainty like this. So the questions I ask today are not meant to state that the FDA doesn't like this protocol. That's not what I mean here. I based on your presentation I had questions I wanted to ask. I will reiterate we will intern tab taken--entertain any well-designed clinical trial. >> Ms., Shapiro. >> I entirely agree with Dr. Powers eloquent explanation and even one-step further. Not having been a part of the RAC--I guess part of it now. [ LAUGHTER ] 36 >> We'd still like you to comment. >> Okay. I'm not bound by the institutional need for the institutional memory however it's generally sometimes said when you have research that offers to prospect of birth benefit--benefit that you should use the affected condition in other words you should enroll those who are affected with the condition. I myself have big problems with that approach in terms of justice and terms of autonomy. In terms of justice they are burdened and sick. In this case we go beyond that because they are going to perhaps be precluded from something beneficial down the line. I think the therapeutic misconception istively in the population in that they may be reaching out for anything that's benefit in research. This case the exact opposite is true I worry following up on Ms., Kwan's comments if they really understand what they're giving up. I have a problem with that approach. >> Two comments another rationality you--really dusty and I talk about the issues over normals versus CF. The issues to do CF patient is we don't know about the transduction--inflammation, infection and mucus there. The normals may not be a perfect substitute because of that. Us that one thing for that. And then how we would obtain consent has come up a couple of times and as we've been through three regis tree bodies I can't see what it would be. The normal we do many. I guess we have 20 activity RIB studies currently. 37 The normal is going away from the principle investigating for to obtain consent. We have a research coordinators a nurse to try to meet unbiassed consent. Having kind of overheard him we retry to verbally go through it, they let people take their time and with prior gene therapy trails not all of them but some of them we have consent monitoring. Again it was a nurse, a different nurse calling back in three to five days and making sure the patients seemed to understand. I'm not sure how effective that was. We actually--I didn't try Dr. Robinson at Harvard tried to study that. Those two--actually for any study we don't have the PI we try to have an unbiassed person as we can ask for--to get permission because a lot of these you know FDA you know approved trails may not be therapeutic. As you know most of them don't come to therapeutic intervention. >> We're going to have to move on very quickly. Dr. Naomi Rosenberg. >> I wanted to ask briefly. You mentioned one reason in your last response as to really there are scientific reasons on which you might base the need to use patients as opposed to normal subjects. You did mention one. But if you could just say a little bit more about that aspect. >> So--well, so the AV receptor is the basele lateral part of the cell--AV2. And the AV6 is the at the surface. With AV2 if you look at the area epithelium in CF biopsies it is quite disrupted and you can see the membrane. AV2 does get in 38 there. However there is a inflamation in the CF airway Dr. Strom pointed out with the enzymes there and mucus theoretically stopping transduction. It didn't in the in vitro studies of the pollps that Dr. Miller pointed out. With AV2 it seems that the CF patient actually may be a better model with AV6 because of where the receptor is normal may be a better--you know a better suetable but it's not--suitable but we still have the airway inflammation issues. As Dr. Strom pointed out almost all CT patients have very abnormal upper air ways. In the small cross section I did all of them had very abnormal CT scans of their upper air ways. It's a very unusual CF patient you you want very mild genome-types that don't have upper airway involvement. >> You are next in line. I'm going to ask everyone for very ters comment. >> The advantage of doing norms would be the ability to find out to extra extent preexisting immunity precludes administration of this spector. Is really known how much the preexisting immune response prevents administration of the virus. >> We have done that the study in the mice and administer one marker and come back with another and ask what the transduction rate and with AV6 we got 50% transduction although developed some anti-bodies against the virus. It's not known in the people. Those would be nice additional studies. It would be great to come back in some of these 39 individuals a year later and ask if you could readminister the rye rye Russ--virus to the nose. >> What's the duration? >> More than a year. AV is quite nice and gives a long-term expression. >> Preexisting immunity does that affect the duration? >> No we haven't looked at that. We only looked a month after the first dose we haven't looked a year later. >> The point I'm make something we're--this sounds like there is on a assumption that treating CT patients who lack immunity is going to come pry mice their--compromise their-- >> We got a strong response you couldn't administer AV 2 there was no transduction in the second round. We haven't looked at all those 78ables. Maybe AV8 would be good for the patients that immune to AV6. I am there are others to try. >> Dr. Weber you've had your hand up. >> Just a quick question. The study will be terminated as I read it correctly if off greater three toxicity is that your stopping rule is that correct? >> Due to the. >> Due to the vector rules yes. >> Dr. Howard Federoff. >> A tiny point. Isly any reason to believe that the enzyme that you choose to express will be active in the respiratory epithelium and might alter the function of those cells. >> We are transfer heat Foss so tase it's express and adult tissue. It's also true if you don't heat and activate the lung there's massive AP staining. It was hard to get the good 40 conditions to see the heat. We think the small amount of additional AP is not going to be an issue. >> Other burning comments? Okay, comments, questions from the public? I'm smiling. We have a private joke. All right, I will precede then and read what I have gleaned from this extraordinarily productive discussion. And to reassure the presenttors you are presenting a study trial at a time when the RAC has been fairly intensively considering clinical design issues your presentation entersecond with obviously some topics we are considering. The discussion was not meant to thoroughly question your study design but rather to raise questions that we all feel are incredibly important in 2005. I guess I should also add because the RAC said one thing four or five years ago I hope that does not mean that we will consistently be held to what may be an old standard. We're trying desperately to move forward especially with issues of trial design and risk-benefit analysis. So I really appreciate your participation. Let me read. General, the unusual aspects of this trial is written included the absence of direct benefit to the subjects, normal subjects, and the use of a quote surrogate and end quote nazle epithelium rare lower air ways. The investigators plan a sequentially approach to the trial of the gene with an effective vector in a lower airway. The RAC asks the investigators to reconsider the enrollment of normal subjects rather than adults with CF to 41 this trial in order to alleviate the likelihood that adults with CF who enrolled this a study would be excluded from future studies. Ethical contract should balance the risk of participation in future studies for CF patients in the and the possibility of therapeutic misconception with the possibility that patients rather than normal subject will provide the most scientific data. As an aside I don't think we should instruct based on the discussion we had--I don't believe we should instruct the investigators to do one then or the other but rather raise issues for reconsideration and this is all up for rewording. This is possible that the same study could initially enroll snowstorm subjects and if the vector proves effective progress in adults with CF. I tried to summarize what where I believe the discussion led us. Would you prefer to reword this section or would you prefer me to read the remainder. Read the remainder. I have nothing under preclinical. Under clinical. I repeat myself here whether the nazle epithelium in man is adequate surrogate marker in lower air ways with CF I put unclear. The RAC remains concerned cat role of the proposed trial and planned sequence of tude for patient with CF is somewhat unclear. RAC endorse the proposed amendment. We did not say this but I want to bring it up for consideration endorses proposed amendment to allow the addition of the three-subject cohort at the highest dose level should the only one in initial cohort 42 in the nazle mu cos sa. That was a amendment that Dr. Miller put up on his slide that would introduce the additional power in terms of real goal of that study. To Dr. Steven Piantadosi's comment that the current design allows 70% of larger group of the subject although none of the three proposed--the RAC requests that the investigators reconsider the study design in light of unique characteristics of this trial. Please interpret the quote absence of information end quote likely to result from this trial. That is all I have. I have nothing on the IC. Dr. Powers. >> Just a thought that may be we, I don't want to keep the researchers getting bounced around too much. >> I don't either. >> May be the language would be something more than reconsider it sounds more directive to an outcome. >> give me some words. >> I would say we encourage the investigators to undertake a careful reevaluation of the risk benefit ratio with an eye toward subject selection predicated upon what we yield the greatest generallyized knowledge. I can only speak at the rate in which I think. >> And I can only write. >> We could encourage the investigators to undertake a careful evaluation of the potential for generalizeable knowledge. Based upon the subject population--the subject selection criteria choice between normal subjects and CF-affected subjects and to consider on the other side the 43 adverse consequences associated with--i'm not sure what the right word is--making it difficult for future subjects to take advantage of future benefits. Just something a little-- >> I can do that. >> Something a little weaker than the urge to reconsider sounds like we're being for directive than we really are. We are at less I think a clear point of view. >> Agree. >> Dr. Vial. >> Could you reread the bit where the RAC is concerned about the step-wise progression whether the nazle epithelium is good place to start sort of thing? Maybe. >> No, I know your question. Unclinical my first comment was whether the nazle epithelium in man is an adequate surrogate marker in lower air ways with subjects in the CF remains unclear. RAC remain concerned that the role of the proposed trial in the planned scenes sequence. >> I see what you're asking. I don't agree with it. >> I would like to say and applaud that is step-wise to build it up earn that what we're concerned. >> I think I can just delete it because I did applaud in the first paragraph which the much straighter. I tried to record the nuances of the discussion and this does not make sense. Thank you. Okay so the first few sentences remain as is. I'll try to modify the next to begin with the RAC encounseling the investigators to undertake a careful evaluation of the 44 potential for generalizeable-- >> Generallyizeable knowledge. >> Based upon the subject selection cry tear ya--that is the chance between normal subjects and CF affected suggests and rolls forward from there. So we have taken out the language that says we're concerned by we want you to redo and I'm much--if you all agree I'm much more comfortable with this. All right. I'd like to ask for a motion for approval then Dr. Powers a second, Dr. Dewhurst and we will take a vote. We'll begin with Dr. Naomi Rosenberg. >> Aye,. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. >> Aye. Thank you all very much especially thank you both for coming and for participating as actively and as willing as you both did. We hope it will be useful. We'll take a break for 15 precise minutes. >> We are not in quorum. But we are below quorum. the only person who I am aware of that won't be joining us is Dr. Heslop. So, I think we are good enough to go now. The second protocol for discussion this morning is 727 clinical translation of a mammaglobin A DNA vaccine for breast cancer prevention and therapy. Dr. Gillander, Dr. Gillander from Wash U will present. >> We are having a technical issue here. >> I understand. We are on pause. >> All right. We are off pause. &%FO &%FO. >> Thank you for your introduction. I am William Gillanders, 2 associate professor of surgery at Washington University and I am a breast cancer surgeon and I am background in tumors and I would like to present today our proposed protocol of a mammaglobin A DNA vaccine. And just, I already found the reviews of the RAC very helpful and made some changes to the protocol and to give an overview, before I give a background of mammaglobin A, we are doing a phase of mammaglobin A vaccine. And the broad objective of the study is to identify a safe and immunological dose of the DNA vaccine that can be used in future studies. The mammaglobin A gene was first identified at Washington University School of Medicine by members of our group using a differential screening approach 3 directed at the isolation of novel, human breast cancer associated genes. And mammaglobin A enclosed lipoprotein and is a member of the globin family. And forms the B. To characterize the tissue expression of this novel gene, we performed northern BLAT and fetal human tissues and these analyses demonstrate that mammaglobin A is for breast cancer. This is a northern blot analysis showing an expression of a mammaglobin A in breast cancer and to a lesser extent in human breasts. And this is conventional RTCPR analysis. Again demonstrating tissue specificity of mammaglobin expression with expression in breast cancer and normal breasts and no expression in 4 the other tissues indicated. To study mammaglobin protein expression. Control for cellar, and between tumors specimens that could be used for routine immuno analyses. We synthesized the peptide to the carboxy terminal and sequence and generated rapid polyclonal antibodies. After rigorously examining this agency. We examined metastatic breast cancer specimens and demonstrated that over primary breast cancer specimenses were positive for the mammaglobin A and for normal breast epithelium. Rare and epithelial cells are seen. And then in contrast, on the right, strong cytoplasmic for the protein is seen in the 5 primary invasive carcinoma. Then of particular note in this preliminary survey of over 100 primary breast cancers, strong cytoplasmic staining was seen in all the cases of carcinoma in situ and appeared -- In situ and was of the specimens analyzed and differentiated breast cancer. So among moderately differentiated, we saw 79% and poorly differentiated breast cancers 87%. Since the identification of mammaglobin A, the molecule has been the focus of over 100 publications in the medical literature. The exquisite tissue made this an informative molecular marker for the micro metastatic breast cancer and lymph nodes as we have recently demonstrated in a cohort study. 6 And also appears to be very informative for the detection of breast cancer in peripheral blood. To summarize these initial studies. Mammaglobin A is a novel breast cancer that has several unique characteristics that make it attractive for immune targeting. To characterize further, the mammaglobin A as an intervention, we generated CDAC lines and in vitro that was specific for mammaglobin A. We had blood from breast cancer from breast cancer patients and stimulated them with mammaglobin A. And the line that we were able to generate showed significant cytotoxic activity against breast cancer cell lines that are H2 positive and mammaglobin 7 A positive and there was no cytotoxicity that were negative or mammaglobin A negative demonstrating the immuno specificity of this cell line. And these results clearly demonstrated that mammaglobin A is processed and presented by breast cancer cells and can be recognized by the immune system. And then to determine if there is any evidence of mammaglobin reactive T cells in breast cancer patients, we performed analyses and mammaglobin reactive T cells in seven breast cancer patients and seven healthy volunteers. As shown in this figure, the reactive of the cells in the peripheral of the breast cancer patients was significantly higher than in the healthy female volunteers. 8 Just to include this, crystal structure of the MHC class one and trimeric structure to emphasize that we have now performed additional studies to study this immune response more in depth at the molecular level by identification of the peptide epitopes that bind A two and three and study the response at the epitope level. So initially we identified HCLA two and three peptides from mammaglobin A using the class one binding program from the National Institutes of Health and shown here, mammaglobin A derived peptides through A 3 but we also performed similar analyses for HCLA 2. And subsequently, we confirmed the ability of these peptides to bind the molecules in cell membranes stabilization using the efficient cell line T 2 A 9 3 and we demonstrate here mammaglobin A .35. A .37 bind to the molecule with high affinity and we have also designed which peptide epitopes bind HCLA two and three are the dominant epitopes. To further, in breast cancer patients to mammaglobin A at the epitope level, we performed the spot analyses and as shown here, there is evidence of an immune response to A and breast cancer patients and no evidence of response in healthy controls. So just to summarize these experiments. We have demonstrated that CDA and specific for mammaglobin A can be developed in vitro from the peripheral blood confirming that the immune system can recognize the antigen and from 10 breast cancer patients, they do have evidence of a higher frequency of mammaglobin A T cells. And so to target mammaglobin A, we decided on a DNA vaccination approach and these are some of the reasons why we think that a DNA vaccination approach is an attractive approach. And I will say this. That one of the main reasons why we decided on this approach is that because vaccination with the full length C.N.A. avoid the restrictions. So, there is increasing evidence that cancer vaccines may be most successful if they are applied when patients either have minimal residual disease or if they are applied in a setting of disease prevention. So one of the attractive things 11 about mammaglobin is because it is expressed near universally breast cancers, it seems it would be a very attractive antigen for prevention and this is in contrast to other antigens and it has been targeted by immune strategies which is only expressed or over expressed in about 30% of patients with breast cancer or primary breast cancers. The other -- so, if you use a DNA vaccine approach, you produce the entire protein and then it is processed and presented in the context of the vaccinated patients individual MHC molecules and this prevents the necessity to either use multiple peptide approach or identify peptide epitopes from every class of molecule. So, in preliminary experiments of this DNA vaccine, we used a 12 double transgenic model and this is transgenic for A two and transgenic for human CDA, so, these -- this is for mammaglobin in the class of human class one molecules. These mice have expression on all of their tissues and they can recognize human HLA, A2 restricted epitopes. So, we in initial studies, we vaccinated the mice by IM vaccination every week for three vaccinations and performed immune analyses and were able to demonstrate that T cells from the mice are able to HCLA two positive and mammaglobin positive breast cancer cell lines but they are unable to recognize A 2 negative or mammaglobin A negative, again the specificity of the immune response. We used a model to determine if 13 the T cells were able to mediate any antitumor immunities. So in this model, we challenged skid mice with mammaglobin A or mammaglobin B and we adopted a transfer of cells from vaccinated mice and then we measured the tumor size by calipers. And as shown here. HBL100 is an A two positive mammaglobin A positive cell line and then the M.D. A M B, 231 is a mammaglobin A negative cell line. And these are the vaccinations with empty vector and this is vaccination with the mammaglobin vector and you have tumor growth regardless of what factor you use. So this shows the specificity of the immune response. We propose a phase I dose 14 ranging vaccine safety trial of the mammaglobin A DNA vaccine and the broad objective of the study is to identify a safe and immunologically safe dose of the DNA vaccine that could be used in future studies. The primary objective is to evaluate the safety of the DNA vaccine and other objectives include assessing the immune response induced by the mammaglobin DNA vaccine by evaluation of the CDC and RAD immune responses and also, I am sure everybody is aware when we initially proposed this study, we proposed the study in patients who were, they were AJCC stage two A, Three B or breast cancer patients who had no evidence of disease. And after considering the reviews by the RAC, we decided 15 to amend the protocols. So we are looking at stage four breast cancer patients who appear to be stable either following chemotherapy or who are on hormonal therapy and are stable and these are new objectives, but since we are targeting stage four breast cancer patients in the initial safety study, we will, you know, assess the impact of mammaglobin A and vaccination on breast cancer tumor markers and analyze tumor cells and also look at molecular evidence of disease using some of our molecular essays and also evaluate the patients for times and disease progression following vaccination with the DNA vaccine. >> So again,--so, again, our new criteria, patients with stage four breast cancer will 16 be eligible for enrollment and have metastatic disease stable, follow disease stable following chemotherapy or hormonal therapy and our goal is not to identify the maximum tolerated dose. But the dose escalation is appropriate. This is the first time this agent will be used in humans. So we will start at a low dose and increase the dose. There is some evidence in the literature that there is a dose response. So there have been several publications that have shown no response at the lowest levels of DNA vaccination. But evidence of a response at higher levels of the DNA vaccine. Again, a dose escalation will only incur if we demonstrated 17 safety of the vaccine at lower dose levels. We plan to perform, I have shown you data from assays that we have done and we will perform intracellular and analyses for HLA two and three patients who identified the immuno peptides and for the sub set of patients we will be able to use HC peptide and analyses and then for the other patients we will use LE spot and expression assays. We, like I said, the advantage of a DNA approach is DNA vaccines are considered to be very safe. I think there is two issues that in terms of the safety and I think, if you read the review of the RAC and one issue is just whether or not our vaccine construct will be safe. And if there is any acute 18 toxicity from the vaccine and we don't expect acute toxicity from the vaccine construct. If you read, most of the phase I and clinical trials of DNA vaccine have shown acute toxicity is limited to site reactions and other reactions that have been described which may not necessarily even be related to the DNA vaccine and include hypoglycemia, Myalgia. Chills. --myalgia. And chills and safety which is the major concern here is what the potential is for autoimmune toxicity related to the DNA vaccine and one of the reasons why ultimately we chose to do the study in stage four breast cancer patients is that although we don't believe that there is a significant risk of 19 autoimmune toxicity, there is no animal models that we can use to test that hypothesis. So, the animal, there is no home log of DNA, and to study the immune response to the mammaglobin A DNA vaccine cannot be used to study immune toxicity. Although the tissue expression of mammaglobin A, it is, it seems to be that it is expressed. It is dramatically expressed in breast cancers and very low levels in normal breast epithelium and no other tissues. Although it suggests it is very low since we can't formally assess that in animal model. We have chosen to look at stage four breast cancer patients for this initial safety study. We define safety as using the 20 National Institutes of Health toxicity criteria. So grade three or greater toxicity and we also define the optimal dose as a dose of the mammaglobin A DNA vaccine that it is safe and the one that is associated with maximal immune response and if we see two or more doses are equivalent in terms of immune response, we define optimal dose that is the lowest dose that gives the maximum immune response. And there is one, I wanted, one other concern that I wanted to address. We have changed the vector and we have been working closely with investigators at Sloane Kettering and one of the investigators on the study is post doctoral fellow in Allan's will be and involved in the creation of the parent vector 21 which has now been used successfully in a number of human clinical trials. So we thought that and that Dr. Howt ton in Memorial Sloane Kettering is willing to share this with us and one of the reasons we chose the parent vector. It had been used safely already in human clinical trials. This vector is very similar to the vector that we used in our animal studies that have been published and using the same CMP promoter and small differences it has a resistant gene for bacterial resistance gene where as the other one. The PCI neo had an ampicillin resistance gene and we are currently testing the ping using the same animal model we used before. So, thank you and I am glad to 22 answer any questions about that study. >> Dr. Gillanders, thank you very much for a nice presentation. I would appreciate if you would stay at the podium so you can interact with members of the RAC as we ask our questions. I have one just to clarify the background for the moment. What is a life expectancy of a woman with stage four breast cancer? To put this in perspective for everyone. >> I think there was a nice study where they were looking at in the New England Journal of Medicine recently where they were enrolling patients in looking at circulating tumor cells and they all had stage four breast cancer and they were initiating a new regiment 23 of chemotherapy and it was 18 to 24 months for those women. >> Somewhere between 18 months and two years. Thank you. >> That was the median. You know, breast cancer, some live longer and some shorter, obviously. One of the, you know, initially we hoped to do the study in minimal residual disease. --residual disease and residual and this is most likely to be successful. They may not have normal immune function. There are advantages and disadvantages of, you know, each approach and then the main reason why we selected stage four breast cancer patients is because the primary concern obviously is safety and so, we want to start there. 24 Let's begin our reviews with Dr. Vile, if that's all right and then I wanted to move to Dr. Piantadosi and then Dr. Powers,. >> Would like to thank him for my replies to my written questions. Just so that I can read the questions into the record. My first comment was really that this is a long term interest of the group and they clearly identified a molecule which appears to be very selectively expressed on breast cancer. Little bit expressed on the breast normal tissue and very poorly or undetectably detailed elsewhere to make it a true antigen. I guess the second point is the outstanding issue with all of this is related to the toxicity 25 of potential, whether that is a big potential issue is unclear. And biological function of the protein remains elusive and I was wondering if there is further insight into the function of the protein in terms of this exquisite impression and indeed in terms of what toxicity and the immune reaction might be. >> That is obviously an area of research interest for our group and unfortunately, we haven't made a lot of progress in identifying the function of this protein. I mean, there is a number of speculative ideas about, but, I don't think there is really any data that we can-- >> So,. >> We don't know. >> Just for the record, you know, if you are vaccinating 26 against a protein that you don't know what it is doing on the normal tissue, you know, there remains concerns, but you expressed those in your discussion. >> Dr. Vile, how concerned are you by the strategy? >> You know, given the exquisite expression, it is only seen on normal breasts apparently and breast cancer tissue. >> Right. And I think it is important to remember, that it is dramatically overexpressed in breast cancer. >> Your RPCR, that showed normal levels. >> That is conventional. Not quantitative. I think it is, and also, the immuno chemistry is pretty dramatic. 27 >> So, I guess, my, you know, the concern is whether there are tissues that, so the concern is if this vaccination strategy is effective, essentially. >> What the investigator is trying to do is raise autoimmunity against the antigen and the consequences against any tissue that does express it, perhaps is not being detected in your studies or what ever. These are theoretical concerns which I am voicing for the record. And if there was an autoimmune activity against normal breast, I am not a clinician, I assume it would not be the end of the world. >> I did a pretty extensive literature search trying to address this. 28 >> There is no well defined, you know, natural autoimmune diseases of the breast, but my guess would be, if you had a significant autoimmune reaction regarding the normal breast epithelium and you would get swelling of the breast. Pain of the breast. You know, and you might also, end up with a premenopausal women, you might destroy the ability to produce milk in lactating women. It would be easy to diagnose because it would be the clinical symptoms would be obvious and it would be confirmed with the minimally invasive procedure such as a core biopsy of the breast. But that's what I believe and that's one of the reasons why we have chosen to document the safety in stage four cancer 29 patients is that we won't know until. >> I appreciate that sort of response. You say there is no hem globe -- mammaglobin A, are there other, does the dog have mammaglobin A? >> Well, there is no published data in any other species other than humans. There are other, in the mouth, there are other members of the secretive globin family and we are studying the members of the secretive globin family in the mouth. There is no hemology at the sequence level. But as of right, but as of right now, since we haven't identified any members of the family that is specifically expressed in the breast. I don't know if they exist or 30 not. >> Just to take it to the sort of -- >> Not that I am aware of, there is really no, there has been no formal study of the tissue expression of secretive globins in other species other than man. >> Have there been attempting to make. >> In the mouse. >> So, you know, for the record, my outstanding issues remain that this is molecule all of which is not understood and the expression of which, there is no model for autoimmune reactivity but with the published data on the vaccinations, I am sorry. With the published data on other DNA vaccines and other melanoma and the restricted expression, I am not sure I 31 would be any more overly concerned at this point. My next comment was-- >> Can I, in the response to that concern, the investigator has chosen to enroll stage four patients, does that help to-- >> Yes, express the concern. >> Yes, it does. >> Yes, it do. >> Okay. >> So my next comment was that in the protocol, the investigator described, we see no evidence of chronic or acute toxicity other than the development of autoimmune and I believe that in the, in that section, there was discussion of other trials mainly in the melanoma field, but I think a patient or somebody else reading that would get the idea that you get the idea of with the mammaglobin vaccine. 32 >> I acknowledge that probably should be revised to clarify. >> Right. And indeed, so the response was that this section has been revised extensively. I haven't seen the revision and that is something that should be done. Because the reference is all set to melanoma studies and that could be confusing. >> Right. >> My next comment was sort of, 180-degree turn, having argued about the problems of autoimmune disease. I just for more interest wondered what the status of this added would be to make it stronger and the response of the investigator was very fair. He argued that they thought about a -- >> We are interested in 33 integrating an adjuvants into the trials. Now that we have decideed to do the initial study in stage four patients. We may revisit that issue in the future. And also adjuvant of --adjuvants or DNA plasmic or other strategies to after an adjuvant into the study, it might document the safety of the mammaglobin DNA vaccine. We thought it would be, that was one of the main reasons we chose not to integrate an adjuvant into the--adjuvant into the study. >> I am satisfied with that. That's a reasonable response. So the next issue has been covered very well. About what would autoimmune manifestations of the breast, what would that be and the 34 investigator covered that very satisfactorily. And the next comment was the change of the plasmic and those -- Plasmic and the reply very extensive. Again, sort of more interest, you show clearly there are T cells in breast cancer patients. If there are T cells which are circulating and this is general immunology question. Why are they not effected. >> There is a lot of evidence that and in these patients with stage four breast cancer that have a significant disease burden. That there is an increased prevalence of regulatory T cells. Our group has published on that subject and now, there has been a number of other follow up publications that have 35 confirmed that observation, so that may be one reason is that although these cells are circulating. They are neutralized by regulatory T cells. And other studies that we have considered doing is studying breast cancer patients with minimal residual disease in an effort to determine if an increased frequency of mammaglobin A T cells is associated with better outcome. But we haven't done, we don't have any data to report at the present time. >> Okay. Thank you. These were all of my questions. >> Dr. Vile, are you comfortable with that response? >> Yes, absolutely. >> With the transgenic mice models and autoimmunity if it 36 could develop, that would be interesting for future that is more of a scientific. >> I hope your colleagues on the study sections will agree with me. >> Thank you. >> Dr. Piantadosi. >> I don't know that I will ask a lot of additional questions. But I will have some comments and maybe Dr. Gillanders will want to inject comments of his own. My initial concern on the study related mostly to design. I found myself comfortable with the biological setting in the motivation for the trial. However, in reading the protocol, there was not a clear articulation as to the actual purpose of the investigation. It was variably described and I think the investigators have 37 cleaned this up cosmetically and they appear to have chosen safety as the primary purpose of the study and it is mentioned twice in their response. One of the things that they did was to quote a chapter of my book, and I don't know whether they were using that against me or not. What I said was my perspective on the coast finding was at play, has been heavily influenced if not damaged, and I will emphasize that. If not damaged by phase I oncology studies. And because the dose versus safety for cytotoxic agents are rather artificial and simpler than others in questions, but many clinical investigators have difficulty breaking the phase I paradigm in, for me, 38 that's a very heart felt comment, I think it is the ubiquitous comment is that the RAC faces here is a very old fashioned study designed that is barely adequate for its purposes and forcing many of these gene transfer questions into the paradigm and they have avoided the and noted the consistencies in the first comment. Which I think is fine. But that's a cosmetic fix up and I think some of the really structural issues about whether this is the right design persists and I will say a bit more about that in a minute. The study design relies heavily on the maximum tolerated dose which is a, I tried to point out and I think the investigators have agreed. Is really not relevant, at none 39 of the proposed doses was there expected to be seen any dose limiting toxicity. And in fact, the background indicated with other DNA vaccines in other context, there had never been any significant toxicity and that is good evidence for the safety of this but it removes from consideration the notion of a maximum tolerated dose as the underpinning of the design. And investigators seem to agree with that. And I do think there is room for doing some sort of dose titration if that's an important question. But you have to titrate and of course the typical phase I study is not intended to do that. It titrates, so it is not a appropriate outcome I would 40 agree and I think the investigators have deemphasized it. However, they tend to insist on the structure of the dose escalation. But it wasn't clear to me what the surrogate for clinical toxicity would be. That is, how far--how are you going to pick which of the four doses is the dose optimum. So there should be some definition to offer there. Maybe I will ask a question at this point. You proposed doses ranging from 155 micrograms to 5-milligrams. What is the optimum definition of the dose? What are you hoping from the set? >> We will be measuring the immune response to the mammaglobin A DNA vaccine and 41 will be using the measures which you talked about briefly and we will measure the immune responses. Using the spot analyses. And using intracellular cyto kind and and by peptide and tetra analyses. And tetra and we will be able to with the T cells and we will define the optimal dose as the dose that results in the maximal and we may not have powered that adequately to differentiate between subtle differences and the immune response. So, it is not a -- immune response and we will be able to measure the immune response and determine you know, if it is, if there is an immune response. >> But, if you are going to use the immune response as your measured end point for 42 determining optimal dose, wouldn't it be wise to put into the study exactly into your trial design, exactly how you are going to utilize it? Because you mentioned in your protocol, looking at T regular, and L.A. spot, you have a lot of end points and how are you going to use all of those to come up with the best immune response? >> Well, I didn't measuring the T reg response is, I think that's more of an exploratory end point and then basically the other, the L.A. spot and peptide tetra and they are all looking at frequency and functional frequency in the case of the LA spot analysis of mammaglobin A T cells. I think we will be able to clarify that in the protocol that the dose of the vaccine, 43 that results in the maximal immune response to what we consider optimal. We are here to make it better. I am open to suggestions by members of the RAC and we have chosen this design. Because we believe dose escalation is appropriate. This is the first time it has been used in humans and we think dose acceleration is appropriate and we are open to suggestions from the members of the RAC and. >> Clearly the way to make it better. Is to have more subjects. Because of your statistical power goes up. Each patient costs a lot of money to do all the analyses. Hard to recruit the patients. There is usually a limited amount of time. 44 So we are in the terrible dilemma of we want to use as few patients as we can. Especially when the doses are safe. And missing and having enough statistical power to say anything. Having done the clinical trials, there is a lot of variability in the patients. They are not inbread mice. It is very difficult to get anything. That's a dilemma we face. >> And in this particular case also, there is a real desire to move beyond the safety testing so that we can start using this free agent and most appropriate context and also, you know potentially building on the vaccine in a modular fashion with the immune adjuvants that Dr. Vile mentioned and in my 45 enhanced efficacy of the vaccine approach further. So I think there is that -- >> Well, I am appreciative of the dilemma and the issue about the scarcity of subjects and I certainly would not propose a design that was wasteful of such a valuable resource and trust. But the scarcity of subjects and the desire for efficiency is not an excuse for asking the wrong study question just because it leads to a smaller sample size and that's really what I am arguing. Dr. Gillanders indicated earlier that he read chapter 10 in my book and I guess I would respond and say he should have read chapter nine. >> I read that chapter too. >> Thank you. >> Because you are a surgeon, I 46 hope you will read chapter four also. But in any case. Seriously, Dr. Wara's point was where I was going and that's it it is exception in these kind of studies to define what you mean by the optimum in a very simple and explicit way. After all, that's what the traditional side of toxic phase one does. It is a pretty lame definition But at least it is a definition. The dose optimum is characterized by the various clinical features and then you know when you have reached it. You decided on a dose optimum that is devised on my laboratory base or biological characteristics but you must define it as simply as possible and then you have to ask the 47 question. Whether the particular design that you have chosen is likely to locate that optimum reliably and efficiently and all I am asking here, is maybe this design doesn't do it. And I am pretty sure it doesn't do it in many cases. And let's not forget the primary question here which is you said safety was the primary objective of the study. Not dose optimization. And those two are only connected very artificially in the classic cytotoxic setting F you are not in the cytotoxic setting the question between dose and safety is, we know there is a dose between toxic and safety. They have a weak relationship between dose and safety so to use a design that is informing 48 about poisons that is very safe, it is probably not the right way to go. Any way, I have made that point. One other word about safety, you said on your slide that safety is the absence of severe toxicity. That's the slide that had the typo on it. Well, if that's true, then you have to provide through your investigation reasonable evidence that such toxicities are reliably absent and I don't think the small cohort sizes you propose especially at the dose optimum is going to provide that information. Now, if you knew the right dose and you put all 15 or so subjects op that dose, you might have a chance of generating reliable evidence 49 that severe toxicity is in fact absent. Couple of other points on my review. Originally safety and feasibility were mentioned as possible study out comes. We--outcomes. Outcomes and we know feasibility seems to have gone away. You said basically you took feasibility out. That's fine. Again, I think it is more cosmetic but I think I am okay with that. I did have a comment for which it has been corrected and I think the investigators response on that point was fine. And my final comment was really a summary indicating that I didn't feel able to suggest 50 firm directions for the study design until the ambiguity and the purpose of the trial had been resolved. The investigators pointed out something that I am sensitive to but only a little bit and that is that there are many other studies that have been approved by the RAC that have used a design similar to the one that they propose. And the polite answer to that is I wasn't around when they were reviewed, and they may have been right or wrong. I think it is largely irrelevant. What I am trying to do is get you to think in a new pattern. Think in a better way and design a study that really meets the objectives that you have set out. So thank you for your 51 consideration of my review and I hope that the comments that I have offered you helped to improve the study. >> And, thank you, very much. Miss KWAN. >> No, Dr. Powers. >> My only comment or query had to do with what is in the consent documents. Those comments were prompted in the initial round of review in the RAC and wondering about the theoretical risk particularly in autoimmunity and the like. And ching of vector and the patient, research subject criteria. I would say that since your own document says that risk of autoimmunity cannot be discluded. It would be to include the document to that effect. Although I think everyone 52 agreed it is up nope and difficult to ascertain. But probably either in this population. Just, it would be prudent to just add that small phrase and that is really my own further suggestion. >> I would like to open this up to discussion by other members of the RAC. >> Yes, I have three comments. The first one is that it is a minor comment. Which used HMAC, throughout the change. They are ditch and need to be changed. Major concern which is, I didn't see addressed, you have an equal likelihood of generating tolerance. What do I mean by that? Well, you have patient population where you are giving 53 a drug not necessarily targeted to the cells and if it is targeted to the tentretic cells. They could potentially induce a T cell response which could be negative. What precautions are there in the protocol to look for tolerance? I think we do more than we think. If you could comment on that, I would appreciate it. >> Well, as the protocol stands, currently, I don't think there is any formal way of assessing that kind of in a real time, in a real time basis and we do propose to measure regulatory T cells and the frequency of regulatory T cells and a lot of people believe now that may be one of the mechanisms of which tolerance 54 is induced. So you know, but we don't, not currently planning to measure that kind of in real time. >> Because if you transfactor and put it into cells and it is clear you get tolerance. That is shown by multiple groups. Now, you are take ago vector and giving it to cells where you may be expressing class one and don't necessarily have molecules on the surface and essentially you are reproducing a state where you have breast cancer cells expressing the molecules. What makes you think