Event ID: 416939
Event Started: 9/21/2005 7:52:01 AM ET

I want to welcome you to the meeting of the RAC, this is a brief meeting this morning.
We have our usual business two presentations and one protocol to discuss publicly.
We are going to begin with an introduction of the new Recombinant DNA Advisory RACmembers by themselves ap Dr. Patterson will introduce people.

Good morning.
It is with great pressure that we welcome several new RACmembers and have them introduce themselves.
By name, by institution and by area of expertise.
Dr., could you begin.

My name is Steve, and I am from John Hopkins. I am a clinical trials or bio stat tition.
That is fine.


My name is Nick, I am professor of the university of Minnesota and my expertise is in and HIV .


Clinically trained in endocrinology, internal medicine and doing experimental research for the last 20 years with particular interest in neurologic diseases.

Robin Shapiro.

I am a professor of bio ethics and study of bio ethics at the university of Wisconsin and I practice health law.


Louis, center for the virus research at the center of California, Irvine.

The next order of business is that we review as is our standard practice the rules of conduct and conflict of interest.
So I will read into the the statement.
Being a member of this committee makes a special government employee and there by subject to the rules of conduct that apply to government employees.
The rules and regulations are explained in a report titled.
Standard of ethical conduct of the employees of the executive branch.
You received a copy of this document when you were appointed to the committee.
At every meeting in addition to reminding you of following the ethics rule.
We ask you to take to make sure conflicts of interest are addressed.
Before every meeting you provide us about your personal, professional and financial interests.
We use this as a basis for assessing if you have any real, apparent already potential conflict of interest to be objective in giving advice during committee meetings.
While we waive conflk of interest for general matters because your ability to be objective will not be effected by your interests in such general matters, we also rely on a great degree on you to be attentive to the meetings to the possibility that an issue will ar roiz that could effect or appear to effect your interests in a specific way.
If this happens, we ask you to recuse yourself from the discussion and to physically leave the room.
You also required to recuse yourself from the protocol review process when you have a real or apparent conflict of interest involving a particular protocol.

As sawlz frks you have a questions about the conflict of interest we will be happy to answer them.

Next, we will discuss the menes of the last June 15-16, 2005 RACmeeting.
Doctors Albelda,.

Dr. Nemerow and I reviewed the minutes and found them to be accurate and complete and reported that to Dr. Patterson.

May I have a motion for approval?



Thank you.

And now, we will go through the process of approving the minutes.
Dr. Powers?

Miss Shapiro.

Dr. Heslop.

Dr. Nemerow.


Myself, aye.

Dr. Feeroff; Dr. Rosenberg.



Due hiewrs.


And thank you all.
I obviously did not ask the new member to say vote, because they weren't here during the last meeting to validate the minutes.
Next is our gene transfer safety assessment board review and Dr. He will beld da, if you -- Albelda if you would like to start.
Sorry, Dr. Heslop is going to report.

There were visits to, in your annual report and five M one responseness.
The M C one responses would highlight two.
To protocol and present it at the meetings.
The first was to fix ate nine which is the Parkinson study.
Administrative responses and protocol 669 had responses to direct comments.
We had a couple of comments about long term follow up.
Three growth trials.
567, and 604, which were having a reduction in long term follow up with the discussion of the F.D.A. and three oler trials that we wanted very long term fl update at that.
291 which is taring, and 391, and 465 which was an alpha study.

Dr. Albelda.

We received 18 total, A events, A being serious.
Adverse events.
Eleven of them were new initial report s.
Seven were follow up.
None were thought to warrant public discussion.

Any discussion or questions by RACmembers?

All right.
We will proceed then to a presentation by Dr. Natasha Caplen from the national cancer institute.
Dr. Caplen.
Thank you for joining us this morning.

Is there a microphone?

All right.
Thank you very much indeed.
So my name is Natasha Caplen.
This is part of the office and technology partnership with the office of the director program for cancer research at the NCI.
I have been working on RNA interference now for about six or so years.
And it has been an amazing journey for most of us that became familiar with this field around six years ago.
When it was first described around that time, that there is a response that induced the very specific silencing of genes as a result of the introduction of RNA.
Up until that it fulfilled then a were aware of the fact that what we thought was RNA encoded proteins and that was the first indication.
That RNA had an important role in gene expression.
In the last few years, wheef seen a whole body of data from multiple different species that have shown now a specific molecule double strand RNA and can express which is only six years ago and now today within the community that you are reviewing, the first trial that has come to you, it could be used to regulate or alter the expression or the presence of a transcript exploiting this methodology.
What I thought would be useful today is the mechanism in which the mechanism is now exploited in the field of again --gem Nome Mcs and applied in an again nomics and this is known broadly and has a number of associated pathways.
Is involved in an inDodge news mechanism that controls jean expression by silencing.
It involves a decrease in gene expression as the root of regulating gene expression.
There is a requirement at some point in the pathways that have a dribble strabted molecule or molecule that contained a double strand.
It is associated with a number of conserve pro treeps that combine preteens and biological roles include the proteins that encode.
M R As and formation and the silencing of self fish genetic material.
In addition, and in particularly, in plants and other organisms it may well form a part of the antiviral response.

So RNA interference and its broadest pathways can down regulate and RNA or a protein level.
Whether at which level this occurs is determined by a number of different features.
Many of which we are still trying to understand.
There are undoubtedly characteristics of the molecule which will decide which regulation is mediated.
Some of the proteins differ that insuring the down regulation of the different levels and definitely a role in the small molecule with the target sequence that will determine the mode at which this regulation occurs.
It is this feature that is going to be a major aspect to what I will discuss today.
So just to familiarize yourselves with the associated pathways.
So when I am and most of us came into the field, the methodology or the what we were studying seemed relatively simple.
What wre wroo dealing with is single strand mole ciewl and invert brats and double stranded RNA that was discovered by a number of different groups.
Was chopped up by an enzyme nope as dicer know the very characteristic small enter interfering RNA and it has a very characteristic structure.
Three primed over and these were incorporated into a very large protein complex and into the silencing complex or risk.
These were then able to locate a sequence that was with this small RNA and choose site specific and sequence specific chiefage.
You would see clear raj of any interests trns crypt between one strand in this case it is the antisensor guide strand and exactly five or 10 nuclear tie-ins from the sense.
If you had a complete sequence match.
You would see a specific cleavage.

What was found shortly after the discover of this, this was part of an inDodge news mechanism.
What we know, there is a lea large number of species.
Between 200 and 500 species in the cells of what are now known as large precursor pliemary micro RNAs which are large transcripts which is in the genome to form the RNA and typical short, hair pin structure which is processed by a related molecule to dicer and combination of other proteins and this is processed into this structure for approximately 70 70 to 80 nuclear ties with a hair pin which is from the nuclear site toe plasm and processed to give molecule a very short period of time that is Akin to RNAs and then also introduced into risk.
What is different is that their micro RNAs primarily thought to induce a repression of expression through interaction that involved the blockage on translation.
However, this is not an all or nothing procedure.
It is not the one trigger cleavage and one will trigger re pretion.
There is an interaction between these and how this occurs is which of these reaction is the most dominant is still very much a matter of speculation.
Most recently, it is shown there is well probably in this process of site toe plas Mcbodys and during this process.
Finally, there has been found a row for some of these proteins and suddenly some elements that are very kin to this process in the form of hem crem continue.

It is being done in low organisms but it is undoubtedly that this Americanism involves a way of triggering changes at a genomic level too.
So, where have we been exploiting this?
Now, what you have is a way of controlling gene expression primarily in this case by trying to endeuce cleavage of the trens crypt and it as powerful tool and the Ma malian cells and three main molecules that are being developed over the last five or so years that allows to come at a pathway that allow us to exploit the process.
First was synthetic.
Very Akin to those that are naturally occurring. The second was hair pain RNAs and subject to the protocol discussion today and most recently has been slightly altered hair pain RNAs that include some of the processing and sequences that we know are involved in the micro organisms and increase the ability for them to be processed into these types of structures.
So, I just wanted to clarify.
This is going to come up at another couple of points exactly what I mean in a difference between cleavageage repression.
If it is thought the sequence alignment is of importance and probably not the only part that is an important in determining whether cleavage or repression occur.
It involves the relationship with different relationships and different localization within the cells.
What is important, we now know if you have a complete Al linement of your S RNA and here I am representing the guide strand of a S HR N A or from a micro RNA flanked by an R.N. A, it adopt matter at this point.
If you have a complete match.
You are going to assure maximum cleavage.
You will see this and relationship will be the most dominant.
However, we know now you only require between 12 or seven nuclear ties of Al line meant for there to be a sufficient for minimal but potential cleavage.
So it is only a requirement for a relatively low alignment so some interaction between this and the target within a risk complex.
Similar story occurs in the case of repression.
Here, what is thought to be primary.
It is not the entire story.
Here you have alignments that involve bulges and misalignment.
It makes it difficult to understand how the interactions occur.
But there is a great deal of work being done on this at the moment.
There are probably more specific alignments that are more facilitating for this process than others.
Again, there are probably max mal repression type arrangements that you will see, there are also minimal alignments that will occur and cause repression and it is thought this may be as few as five tide title alignment.
This is dominating the field over the last three or four years, as it has developed. I will come back to this in a couple of poans during the presentation.
I just wanted to make everyone aware of some of the differences between what I call the X triggers and how this is relevant to the discussion subsequently.
So the triggers that we have understood today are micro RNAs.
Naturally occurring small species about 22 nucleares in size. They enter ak with many of the same proteins as you see when you use ex oj news but it is primary thought they target and decreasing level through aggression and there are cases in the literature where a micro can cause a cleavage and its means. The gene expression through transcript interaction appears to be with the process.
Including differentiation.
And there are now increasing number of studies over the last year that show and ber rapt expression of micro RNAs.
May have an important role in cancer very many.
So any disruption of micro RNA processing as a result of exploitation has got potential issues and has to be taken into account.
What we have been doing in the context of the gem Nome Mcstudys and what they have been thinking of, is the use of ex oj news triggers for RNAi.
That is hair pin or S RNAs and I just represented them here and S RNAs and we want them to go into the same waghtway.
Whether they are processed or not.
What is absolutely key is that you want to make the most efficient risk complex that you possibly can.
That will target your gene of interest and a great deal of work over the last couple of years has gone on to insure, that you include the risk complex will preferentially load the guide strand corresponding to the transcript that you are trying to regulate and this will give you a functional risk complex that can disrupt your target transcript.

So, there are two principle RNAi vectors that are being, and these are the two primary once that I am going to mention today.
Double stranded nuclear types and usually consist of a strand which consists of RNA and have an over hang of two groups.
The antisense antistrand is cog cognizant and again have D T T and X Y in the primary, and commercial entity that produced these.
These are highly efficient.
There have been safety issues related to the delivery of pro clip cal trials. But that's not relevant to today's discussion.
Hair pain RNAs are a stingle stranded molecule that requires enter molecular and these can be introduced cellularly and therefore may ex employed additional features of the Americanism that S RNAs are unable to.
What are the primary applications to date?
Most of this has been in the context of fiscal genomic analysis so the ability to knowing down a controlled fashion.
A study.
And we have been able to take the studies that occurred and put them into a cell culture system.
This has applied to board organisms and he is definitely moving know the whole genome fields with humans.
They are using S RNAis and volume I haddation of molecular targets.
The development of mobile systems and the ability to be able to knock down a gene for a very long period of time and create a whole new cellular system and finally the trance generalic animals and introduction of these during development.
R three embryonic stem cells and actually generate animals where there is a control knock down of a goan. The advances of these.
Thee could insure spatial or temporal down regulation.
It has been very quickly the move to conduct experimental as a therapeutic approach.
I just wanted to show how quickly.
So the very first studies of RNA and interference in the cells were published in 2001.
But by 2002 you were seeing the first models and this is a down regulations in a mouse model system.
It may not be very clear here.
What they are showing is the wrath could suppress the in this model system.
Very quickly they saw the therapeutic this particular mechanism.
It is also applied to genomic disease where they were targeting a transcript in var sell la attacksia and following deliverly with an adknow virus and change the markers of this disease in a trance generalic model of attacksia.
Very quickly there were a number of studies and this is one of many studies that show SNRNA could be used to inhibit HIV infection or replication.
It may not be clear.
But there are points where it can be use tod challenge H I vi by down regulating a September tore.
You have the potential to stop the infection.
Because it can no now longer enter and you can challenge the transcript itself during the life cycle to trigger a degradation of the RNA genome.
There are a number of studies conducted to do this.
They have been published.
I will not review them today.
They will come up in the discussion.

To say this was very quickly adopted as an approach to this.
What I wanted to do is highlight some of this to the mechanism and how complicated it may become in the process of moving RNAi to the clinic.
I I am highlighting four things here that have occurred to me and many others within the field.
What if you were inducing RNAi with in cells in a clinical situation, what do we think will be the short or long term affects on of RNAi and will it disrupt RNA processing or other features.
What do we think will be the affect on nontargeted transcript and we now know that the sequence align menes and other features can load quite quickly to off target effect.
What is the activation.
Any chance for activation of nonspecific double cellular related to ante virus response in the cells.
What can occur, so as an establishment of mutation L is a selection against, can the RNAi effect a molecule, that be triggered and altered so there is a resistance to it.
Specific nuclear type changes can change the RNAi effects so it will now not chief.

So what about the effects of RNAi.
Can the RNAi machinery be saturated.
Can this alter it to such a degree that now it can now flow longer process N R A I.

They have used multiple RNAi machinery.
Large genetic screens have now been conducted that would let you look at multiples at a time and doesn't seem to be any growth effect.
However, the study that is limited and really will not be complete until we can do full transcripts on the next type a.m. sis.
This is probably not feasible at this time whether it is appropriate, it it there is going to be some discussion and whether we look at the RNAi pro feels, did we see any profiles changes in those.
Does this indicate we are not disrupting it therefore, it should be okay.
Finally as long as the functional analysis is okay.
Does it mean it is not effected.
You are maintaining differentiation which we know is an important marker of the action of micro RNA.
It is not enough data to be able to say conclusively what is the maximum dose or the minimum dose at which we would be disrupting the machinery.
But collectively. We can probably say at this stage.
There is minimal chance in that.
There does feed to be emphasis on this point as other markers are not going to be feasible.
the RNA on most is complex issue that is going to face anyone attempteding to use RNAi and the optimization of the effect tore and potential for using expression profiling to see whether other genes are effected.
Other issues related to that.
They may give us some clues and not going to be, may not catch every feature.
There are limbs to how it can be used.
Finally can we use functional analysis.
Are there any real changes on any other nontargeted transcripts and having a true effect on the cell.
So, one of the feet fewers that--features that have moved in the last couple of quloars is one -- years is one aspect.
It is the optimization of the investigator of trying to make sure you are just so you are increasing the probability that your complex form and not a complex that ain contains the strand have other series of target effects.
This is may increase your efficacy and not necessarily a marker of eve fik Cassie.
What I found by number of groups that primarily worked for the doctor.
There were certain features of the guide strand.
In this case, where I plotted the ante sense.
And the five primed certain dynamics you could represent as related to the Delta G for this molecule.
Where it was shown at this end there was a relatively low Delta G effective and they were in micro RNA and increase in it which then comes down and still debate at what is happening at this end of the molecule.
What it was showing.
What actually happens when the molecule is being loaded into risk.
Risk probably tests either end of the S RNA or micro RNA and unwind that that is eastiest to unwind.
Unwound the eastiest which is relate tod a lower Delta G at this end.
Then you can design a little bit.
By just changing or looking at the certain dynamics of this to increase the likelihood of that loading.
There are probably other features.
This is not an all or nothing.
We certainly have a number of S RNAs that do not obey this rule.
Which I highly.
But we do not know whether this is influencing the decrease or changing any potential up target effects from other, if the sense strand had been loaded.
Where with most sort of speculation occurred first to risk of the up target affects is threw the expression profiling.
And this is a study done from rose setta where they took SRNA and it may not be clear here.
What I am trying to do is point to, there was found, a number of genes that had an alignment with the element of that sequence within it and S RNA that showed they were also changing in their expression profile.
This led to a number of studies.
Discussing there were or how one could predict what the target effects will mean.
So, though, the interaction that we are targeting is the predominant one.
There is the potential for a significant background of off target interactions.

I just wanted to show one example to make this slightly clearer.
This is where we designed.
With S RNAs that were designed to very closer related genes and A and B.
Which they set in a molecule level and attempt in this study to knock down both jeans by designing and we are trying to find out what was the minimum require men to get.
What was effectively a target affect.
So what we did is designed S RNAs to.
I hope this is very clear.
This is where we targeting N M A 2 or 231 B with SRNA that were cognizant and these chiefed successfully.
One was specific.
There was no alter ration in the expression levels of enemy 230 A but there is one SRNA and had three miss matches with that.
With the sequence, with enemy two and this did give a significant decrease in N M E one levels.
This is the point.
It is very hard.
It is not easy to predict which Al linements will lead to this level of KHRA*E vaj.
-- cleavage.

This is the important aspect here for this discussion.
Is that we are still not quite sure what is the minimal level of alignment that will lead to some sort of suppression or cleavage in the context of an up target.
Undoubtedly this will be the predominant reaction.
If you have 100% cleavage, this is the reaction.
This is the minimal or the potential relevance of any other interaction that can occur.
So, I think just did an analysis to show an example related to the crin cal trial that will be discussed later.
I have no idea at this point whether this will be remotely functionally relevant.
An indication of what may occur in a cellular system.
This is taking the target sequence for the S RNA and it will be discussed later.
When you do this against, you find this type of match for this particular target sequence.
That is a calcium channel dependent gene and has an exact match across the target ap release small mismatches here.
In the lab, we will predict this might act as an interaction that could result in a cleavage or repression.
This will have no consequences.
I put that out because it is possible.
We did another example here is a search against.
I stress this is a buy--this would not occur, and may have no functional he will men at all.
But it is to indicate at this stage we can't really predict what it is with these interactions will do and what degree they have importance.
So that may be that actually the function analysis is only, is the only way it is occurring.
You can do as much profiling as you like.
As much analysis as you like.
But it remains intact is this the best measure that we are having minimal nonspecific effects.

Finally I just wanted to mention obviously RNAi has the potential, of activating nonspecific double RNA and the cells have extremely sensitive responses as part of an antiviral response and these frequently result in a nonspecific decrease in protein translation and through different mechanisms.
What we do know is the size of the double stranded RNA to main and concentration of that may play significant role as to whether these are activated or not.
In what cells they see these may also be critical.
Some sequence motifs and primarily these are studies using SRNAs and maybe more pertinent for the cellular responses than others.
But it is again.
Still a very much matter of debate.
The key proteins that we know involved in the responses are double strantded PKR, or OSAone and other related proteins and more recently.
The reSeptember tore in these processes.
This is still very much an debate.
But it seems this may be cell type specific.
So in any particular clinical setting.
It may be looked on a case by case basis.
Finally I want to cover this through mutation.
So we know through many studies in our labs single nuclear changes at specific points in the RNA sector can completely eliminate RNAi and there will be pressure against any S RNA induced cells that is a result of the mutational link.
Due to singular changes.
This I believe has more significance with respect to self fik Cassie -- efficacy and now it could have an interaction.
And different target interaction and related more to he have fik cast that any change in the -- efficacy that it can interact with at a transcript level.
So, the very brief overview, RNAi has proven to be a potent gene silencing as a and Dodge news mechanism which we are trying to understand at a great detail.
It is at a different levels and RNA and protein level.
RNAi has been exploited a a functional genomics tool and actively used in most labs within most labs certainly in the U.S. that are doing biological analysis and most scientifically active labs.
But the move to clinical adaptation.
They set realistically.
Any effects op RNAi critical role in normal cellular processes.
Any effects on cellular transcripts and men mizization of any specific effects.

I will now take any questions related to that.

Thank you very much for a lflly -- lovely over vow of this -- overview of this.

Any questions from the RAC.

Do you know any examples where they tried to down regulate a gene and created a tumor or killed a cell.

You can very easily kill cells.

Inned vert tenthly.
-- Inadvertently.
This would interpret the data sets in relation to what we specifically targeted but there are a number of, increasing number of papers in the literature that are saying that larger screens where they have had hits that come out of those screens.
Now depends on.
Not defining what the hit was because it was part of large screens where that is now not a target.
They have not been able to reproduce that looking at other RNAi genes and in those cases they will be regulated to cell deaths and I know there are cases that are not yet reported but in the discussion of that occurring. The no, I have not heard of that specifically.

What about the, I like the idea of doing search.
Now that we have the genome.
Have they done that and gone after the possibilities.

My understanding is that, analysis has been report ed but the biological data testing is not yet in the public domain.


Dr. Feeroff.

Is there any evidence in either pro generator or stem cell populations that the genetic conveyance at the level of changing from a structure is a function of when those pro general it tores are formed and with respect to newly formed or differentiated products of that pro general tore stem cell?

Not that I am aware of at this point.
Most of the studies conducted locking at het crem changes in yeast or very low studies at this stage.
A lot of interest in moving it in the cells but I don't have data that would support that.

So, how is this system if you have a SRNA expressed in a cell.
Can a subsequent RNA be ineffective.

No, you can put in multiple RNAs and select silencing of different genes.
So in the context of this, if you have a SHRNA and the HIV can generate a mutation that is an escaped mute tant.
Now the cell is producing an ineffective S HR N A and so if you come in with a second one.
The machinery is producing this, so the second one can be as effective as the first one until you get resistance.


Do you think there will be a time when there is no reward.

At what point you would saturate out is the key question.

But it would, my supposition is toe moment people are discussing and it is not yet done.
You can put hundreds in any one time.
Without effect on the RNAIand maybe other context but not at this point.

Good morning.
If you could comment on the trance miss sieve character of the RNAi and absence in the cells.

So, in C and--will is evidence for an inheritable form of RNAi so we can be silencing from one generation to another.

These appear to be under the influence of specific proteins but do not appear to be present in the cells at all.
There is no evidence of any spread of the RNAi effect where in C elegance there is a spread of that effect and the proteins are being slowly defined.

So in the context of studies to date.
The proteins that appear to support on genetic analysis.
That does not occur.

Other questions, comments.
From the audience.
Questions or comments?
All right.
Dr. Caplen, thank you very much for joining us.
We will call you on you again when we discuss the protocol.
We will take a break.
I want to remind everyone.
If you need transportation, at the end of this meeting to an airport, please sign up at the desk outside this room.
And please sign up during the break.
We are a bit early so we are going to reconvene at 9:15.
Thank you.
[ Break ] .

The protoe Coll discussion is titled safety and feasibility for AIDS lymphoma using stem cell treated with Lentivirus vector encoding multiple anti-HIV RNAs. Dr. Krishnan is the principal investigator and we'll lead off the discussion doctor?

Thank you Dr. Wara. My name is John sayer. You know, I'm the project leader of this protocol and I come here with some of my very talented colleagues. John Rossi will be joining me to discuss the research side of the safety data that we have been accumulating. The PI, is Amrita Krishnan, a translator from the city of hope who has extensive experience in the area of transplantation ever AIDS lymphoma. Larry is in charge of manufacturing and regulating affairs. He's the senior VP for applied technology and Dr. Linda Kuto, who is with benitech where she's director of preclinical development. We want to tell you basically three things in this short overview. Why do we want to do the study? What do we think it will do for the field? And why is this so important? What about the vector design that we have created? What is so special about this vector? And what do we know about the safety of the procedure? Well, first of all, the goals of the study are simple. The study is one of safety and feasibility of stem cell therapy in the setting of AIDS lymphoma. Using stem cells treated with a Lentivirus that includes multiple anti-HIV RNAs. We are looking at safety in terms of adverse events and HIV infections and the feasibility of the stud, in terms of the quantity of the cells that are derived from these stem cells and the duration of their circulation in the periphery, the blood and the character of the vector marked cells in terms of integration analysis. But to begin with, let's talk about the national for doing this study. It's really to advance the field of the management of the HIV infections. Not to advance lymphoma therapy. The problem with current multiple drug antiviral chemotherapy is that HIV is still detectible in the tissue and when you stop therapy, the virus recurs. There's a potential for resistance to occur as it does in some patients. Some patients have serious side effects. For outpatients, it's expensive and to propose that you will take key chemotherapy for a lifetime is a difficult thing to grasp, especially when there's a worldwide problem. So gene therapy is a way to do this. It can modify the need for continued antiviral therapy. The development is the reason to initiate this. We think it's the next step towards the eventual development of a genetic therapy for aids. The study will provide important information needed for determining the safety of this relatively new vehiclor, Lentivirus vector that has been used already and reviewed by this committee in the past. But it is a vector as it has considerablepotential for other areas of gene transfer research. Now, why have we chosen AIDS lymphoma as the setting for the trial? This is an important point as well. When you are transferring genes for the treatment of HIV, exvivo, the manipulation of the genes exvivo, you can use T cells for blood progenitor cells and we have chosen them because they are multiple targets for HIV infection. The problem is that this requires mylo oblativ to optimize the grafting of cells. There's few patient populations in which you can begin the evaluation of a new vector by introducing myloblativt therapy, perhaps not even heart failure patients. But there is a population who are going to undergo myloblative treatment and that's patients with aids lymphoma. So the setting for transplantation after dose intense therapy for AIDS lymphoma we believe is not only ethical but a scientific appropriate clinical setting for the evaluation of this new genetic vector. This is actually done and you are having just the treatment of your AIDS lymphoma you undergo collection of peripheral but progenitor cells as you finish the last cycles of chemotherapy and we collect the cells of apheresis type, with an ater Reese is collection and they are chryo preserved and then there's a dose intense chemotherapy shown here. And the cells are infused almost as they are rescued of that high dose therapy. We will interpose a collection of two extra apheresis and select the progenitor cells. Then when the patient is completing the dose intense therapy, these cells will be thaws and transduced with this vehiclor and infused. That's for the duration of expression of the transgenes in the peripheral cells. Now why is this study important to the field? The study will provide some indication of whether the use of Lentivirus vehicl vector -- vector is seasonal in gene therapy. If the vehiclor is effective in the setting, it will have considerable application in other settings that would be appropriate, using hematopoiectic cell based gene therapy. You can imagine the treatment of sickle cell anemia, would be considerably advanced if this Lentivirus is shown to work in HIV. What is special about the vector? The vector is derived from HIV in such a way that it can not replicate. We have removed the accessory protein that contributes to meth genesis. And there's a self-enacting Lentivirus. It expresses RNAs that inhibit HIV on a replication and this is the first use of gene transfer of RNA interference as you have heard as a strategy in clinical trial. And this slide here just shows -- as you probably already know, this is HIV itself shown ski gnatically with -- schematically with its accessory proteins and other proteins necessary for replication. Large areas of the envelope have been removed the LTR has been altered so it can no longer function. And we have approximately 10% of the original sequences of the virus and the vector. We then introduced these three transgenes, we call them, the SH-1 which is the small hairpin -- this hairpin RNAi to this overlapping region here. So it hits both tap and rev. We have TAR which interacts with tap and interferes with the function and serves as a decoy strategy and HIV replication and we have a rye -- ribosome 2. And one of the vector integrates, the integrated copies are shown here, and this will be the sequences that will be analyzed in the intergrant analysis interim study. The rationale of this particular antiviral design is that the -- as you heard SNRNA is an appropriate inhub thor of RNA pro-- inhibitor of RNA processor. The problem is it's very specific and does force induction of resistance. And when you singly -- it's been well described that you can -- the virus can circus vent a single SH RNA. This is the RNA element that we are adding no this, because of its excellent ain't -- antiviral events, it interacts with tar and tap and it's expressed in the nucleolar to achieve optimal effect. As it the tap comes into the cell in the virus, it's siquestered in the neucleolas. As previously described. And the ribosime knocks down this protein that's important in the secondary receptor entry into the cell. So that's the overview of the study. I'm going to go to the discussion of what we believe is our information on the safety of the approach. Excuse me. the transplantation procedure itself is well known to us. Dr. Krishnan and her colleagues really Pioneered this area, and they -- in fact Krishnan has the largest experience in this field and it's been shown to be incredibly effective in the treatment of this fairly difficult-to-treat disease. We have used the study designed in the past in a study on the evaluation of ribosomes used in the same settings, for AIDS lymphoma patients and Dr. Krishnan was the PI on that also. So we believe that personnel involved and institution, which is one of largest transplant centers in the country is I am meantly cap imminently capable of doing -- imminently capable of doing the work. We looked extensively at the SN RNA, especially the off-target effects and microRNA are influenced by our manipulation and I will ask Dr. John Rossi to join me on the podium to describe this detail -- these results in more detail. John Rossy is an expert in the field.

Soil focus on the safety issues and the actual issues that we addressed experimentally are shown here. So the first issue that I will address and a lot of these are nicely brought up by Dr. Caplen in her presentation are the significant alterations of microRNA profiles. We have all heard that microRNAs are indodgionous that regulate gene translation. Are there significant disturbances or immune activation suggesting a non-targeted genes? And I'm going to be talking about, in particular, the potential for Interferon pathway induction by hairpin type of RNAs and there's a strand of the SHR RNA risk Anna Tasha brought this up as a potential issue that has to be looked at. So actually it's easier for me to point it here. So are there significant microRNA profiles and we addressed this question by using a microRNA ray, specially available microRNA ray which has on it, highly consevened -- about 2 -- conserved about 250 known microRNAs and most of them are represented cross species. So the array had representation from five different species, human, monkey, and some other organisms but I can't remember. Each specific microRNA is present in triplicate and you get fairly nice statisticalal evaluation. In using this array we have used not the same construct of using the gene therapy, but we thought we would push the issue further by using a construct that discussed three separate hairpins each targeting the HIV but at a different site and one of those hairpin RNAs will be used in the clinical trial. And what we did is analyze the microRNA profiling from this vector containing these three separate RNAs being expressed from the promoter and CMET lypo cytes. And what we saw, in both systems was that two microRNAs seemed to have some statistically significant up or down regulation. One is call MUR 224, which was upgraded by two standard deviations by the mean of the control which was the vector transduce cells alone and MRNA 227, which was down regulated relative to the control, again being the vehiclor. Now, we went -- vector. Now we went the next step and asked -- so these are microRNA array analysis. We don't know what the target for either one of these ar. And 337 turns out to be predicted microRNA. It's not known if it's a real microRNA. But we did analysis to see if we could actually see these up and down regulations at the level of RNA blotting and the answer is we did not detect any changes at all. We couldn't detect 337 RNA at all. And why it comes up in this this type of analysis we are not really sure. And 224 remained unchanged. So we have actually not stopped there, there. We have gone ahead and HTK 293 cells, again we saw similar profiling, with a different approach, that we could see 224 and 337 being altered in one fashion or the other. So we have actually tried to look at this more includively and 293 cells to see if we perturbed these two different microRNAs by knocking them out using anti-cell nucleotides and there's been no functional consequence that we could look at and we have looked at cell cycle analysis and potential arrest. So to the west of our knowledge, there is no major prohibation, which this was a prominent member. So the next question is: Are there significant distributions of cell functioning targeting non-cellular genes? And as Natasha pointed out, there were a couple of published descriptions of danger motifs in sn RNAs. These were identified by different method that is by taking synthetically manufactured sn RNAs and delivering them either to cells and cultures in particular invee voluntary, in animals where the cells will actually take up this RNA. And what was shown is that these two motifs, in particular, are active in inducing type one Interferon responses. There was also a single report that SH RNA induced Interferon output induction, and this was actually retracted in a publication about seven months later with we realized -- the investigators realized that their construct had an aberration start site. So many people don't know about the retraction, but this is in the literature as one -- the only case that I know of where an SH RNA is shown to have an Interferon alpha induction. And then the Interferon genes be activated in CD 34. So these are the questions that we are -- we have addressed. So the method of experimentation is to collect CZ 34 positive core blood cells and we have done this with antologous but I will show you the data with the core blood cells. The transduce these with the vector. This is not the clinical vector. We have to use an EGS marker so we can follow the transduction vector. And then positive cells by flow symtometry. And give -- and then once these are obtained, again, we go through a fact analysis looking for markers, and Interferon induction by alas -- by assays. We looked at the cytoyinese reheat and macrofaction. We did take synthetic SH RNAs, and one of them turns out to be -- actually, two of our SH RNAs we have tested in the lab, contained a UGU motif and that turns out to be immunosynnatory when presented in vitro. But we wanted to look if we were able to do the same thing. So this is just a slide showing the results of that assay and so this is -- these are done in multiples. So we actually have about six independent confirmations of that data. We have looked for induction of Interferon BETA. MXA which was a prime indicator of Interferon one pathway, OSA 1, which Natasha mentioned and OSA 15, which turns out to be important downstream factor from interfere ons pathway induction and P 56 the same. So if we just add Interferon to the CD 4 cell population, you can see that we get activation of all the genes in the Interferon one pathway. Seize cells are responsive to -- these cells are responsive. But if we look at the construct, the triple construct in particular, not one of those genes is activated there. May be marginal, with a different SH RNA which is irrelevant to this study and just the SH RNA that we are using in this study doesn't have any infear ons induction pathway and so show that the genes are being expressed we did a northern gel analysis. And you can see the SH RNA. So they are actually being expressed but they are not inducing the type one Interferon pathway genes. We next asked if the cell differentiate normally in culture into monocyte and macrophase linages. This is before sorting. You can see that that there's about 35 or 40% transduction efficiency in the 34 positive. And if we look at just the total population of different markers in this initial population of cells that are not transduced in this case these are the various markers that are assay for and this is the distribution. So these are unmanipulated, ungenetically manipulated cells. These cells were sorted for expression and we looked again for the various markers and the various constructs and you can probably read it better than I can here, but you can follow the triple versus the vector alone. and this is looking for either monocyte or macro page and the triple construct shows no significant difference in any of the marker differentiation pathways from the controls. So we feel quite confident from having done these types of analysis multiple times that differentiation ex-viv ofo -- ex-vivo is not expressed by expressing the SH RNA to the at that time rev target. Now the other question that Natasha brought up, and I think this is an important question as well. You have two strands that can enter in here. Of course we want the anti-strand to enter because that's targeting HIV but there's also a potential that a tox strand can enter. When you express sh RNA, using the u6 promoter and I will show you that data. The way we assay this is we put a large segment of the HIV target into a vector that has a reporter that's fused to an untranslated region where we can argue our SH RNAs, knock down the target in the three primary ETR and actually results in known or reduction. So we can put it in two different orientations. This would favor utilization. This would enable as say for utilization of the antisense strand and enable to assay for the unite saition of the sense strand. and so this is -- in is as saying for -- we can -- assaying for. We will focus on the ones that I circle because there are constructs irrelevant to this study. The antisense strand which is targeting the actually at that time rev sequence. 90% or more in this particular assay. But when we put the target in the opposite orientation and assay for the sense strain activity we see that, in fact there's no knockdown of the sense train whatsoever. So this says that that sense strand does not enter an effective risk complex, and therefore, can not target an RNA. And this actually turns out to be an interesting case for all the SH RNAs that we looked at that come from pal three promoters the sense strand at least in our design of construct does not enter into risk. It could be something to do with the initiating triphosphate that is not present on the bod Tom strand or the way the sh RNAs process, but we don't know at this point. But we think it affords another safety teasure that we don't have to -- feature that we don't have to worry about the sense strand being an active component of risk tox summarize, all ofure ex-vivo expression no concerns. Our microray analysis shows no Rell significant regulation, in the microRNA profiles. Two microRNAs that did show profiling up or down rent ration couldn't be reproSeussed at the northern level. We are not really certain what to make of that, but we do know that 337 is still a punitive microRNA. The clinical vector expressy macrophases this was submitted to you in part of our publication by Leo. I didn't talk about this but we showed that it has normal marker differentiation, but also are able to package and tie with e. coli, with a normalphagocytosis assay. We will look for T develop differentiation -- T cell differentiation, using the clinical vector to transduce CD 34 cells and let them mature using feetlesinusliver. They were able to transfer at the same efficiency as the vehiclor cells and that these -- unlike the control vector which is highly sensitive to SIV this continues with the resistance of the high challenge. We have done a number of experiments with Alex at UC Davis, using each of the constructs that we employed in this stud bibut individually not in the context of the triple and she has injected the Lentivirus targeting at that time/rev or the CCR ribosome in fetal mechanic in unite ore row. But we -- in utero. But we did see paragenal, but the monkeys came to term normally. All the markers indicated there were noor was no toxicity and they are alive and well several years later, at least some of them. And I'm sure we would see high levels of expression of the egsp marker which has lasted in one monkey throughout the three years. So in summary then, we think that we have performed a number of at least functional safety studies that give us a high level of confidence that these genes are not going to cause any overtoxicities.

In conclusion, then, this proposal will evaluate the new Lentivirus vehiclor through antiHIV rans and the goal is to treat AIDS through again therapy. We'll look at RNA interference and will have application to many other gene transfer studies that we would like to apply RNAi therapy in the future. And this is particularly appropriate because it will inform future trials of stem cells that have application to other types of diseases. Thank you very much. We are present.

Thank you very much. We'll proceed with our discussion. And our usual process is to ask you, doctor, to stay at the microphone, and Dr. Rossy, we probably want you fairly close too so you can help answer questions. We will go through the formal written reviews one by one, beginning with Dr. Caplen and ask each reviewer to restate their initial concerns, and then to simply say whether that concern has been satisfied, ameliorated either in writing or by your presentation this morning. And if it has not been fully satisfied, to state that as well. And we'll proceed through the formal reviewers. Then open the protocol up for discussion by the full RAC and then move to the public and ask for further discussion. So we're going to begin with Dr. Caplen.

So in my review of this particular protocol, I focused on three particular aspects and I want to stress at this point that I chose not to deal with the potential efficacy of this assay SH RNA. That was not part of the main emphasis of this pilot study. So the three concerns that I had were will the SH RNA assay affect the cellular role of the RNA? Will the SH RNA interact with other transcripts in such a way to induce an off target effect? And will the SH RNA or other RNA evidence within this vector trigger a double global RNA response which normal results in cell deaths? So in the initial material that was supplied to the reviewers, the investigation presented some evidence that long-term expression of this particular SH RNA would have no overt effect on the cellular role on the RNAi on machinery. The specific evidence including data showing that transduced CD 34 positive cells differentiated normally after the introduction of the SH RNA vector. Subsequently in today's presentation, they have also shown that the profiling of microRNAs appears to be -- remain remarkably stable. So they have indicated that there are two microRNAs that show some changes in expression levels. They are indicating that although they are extending those studies and looking at that further. What my concern was with that, even at this point, is whether there's sufficient preclinical data to satisfy us that there's no long-term effects on the RNA machinery in those cells where there's been a transduction of this vector. Next point was will the sh RNA interact to produce an off target effect. The data that has been presented by this group has been based on function analysis that those sells that have received the vector system show no specific changes as a result of that change, though they differentiate normally and they are expressing the appropriate sort of markers. The question remains, though, whether there are still some cellular transcripts that would be affected by this -- the sh RNA included and whether additional analysis needs to be done with specific cellular transcripts that could be analyzed si matically or some microarray profiling. In my opinion, the use of microexpression profiling may be somewhat misleading. There's a great deal of noise in those types of assays that may not be ideal and there may need to be some specific follow-up on some particular transcripts that we can at least make some prediction of. Final point was, in the context of whether any sh RNA or it should be said some of the other RNA elements within this vehiclor which is quite complex, will trigger a global double stranded RNA response. The group has presented within the information that was supplied to us quite a great deal of data comparing the sh RNA vector with an s RNA system showing that in this case, it seems to be minimal triggering of those Interferon responses, and I think the data presented here is as extended that and seems to be approaching that in quite a logical fashion. So there probably is increasing amount of evidence that these particular -- this particular sh RNA would not trigger an Interferon response that would be detrimental to those cells. The question would remain whether that is still sufficient in the eyes of this particular committee. So those are the principal points that I saw in the review that I was concerned about in the context of RNA interference. I think the group has addressed most of those concerns. I think the remaining feature is whether we can make any assessment of any potential targeted effects in my more detail.

Is there anything specific or more specific that you would like to comment on or suggest? What I'm hearing -- I just need some help with this, is that additional work on the potential off target effects is important.

That is my -- I think that's my feeling at this stage. There needs to be additional workup to ensure that if we can predict at all, any cellular transcripts that would be affected by this particular sh RNA, should be followed up on.

Thank you. Let's go to Dr. Rosenberg.

I'd like to thank the presenters for their presentations, which I found to be clear, and also for their responses to the questions I raised and the additional material that you all submitted which was very helpful. I think most of my concerns have been addressed in the presentation, but I will go through these for the record. My first point related to whether the transduced cells would retain function and whether the presence of those cells might, in some way, affect the efficacy of the transplanted cells, its function of the transduced cell was impaired, and I think you have spoken to that. I do wonder, however, you certainly shown us data on macropage and alewded to data on T -- alluded to data on T cells I wonder if you have dateo on other lineages and the reconstitution of those that you might be able to tell us about.

The answer to that is we -- we only have the s KIDU data that shows T cell differentiation at present. We have in vitro data that you have seen with macropages and it really will be the clinical trial in which we'll be looking at whether any of the other markers will -- any other lineages will mark in the clinical trial.

And there's no potential to model that in an animal system to get those other data in your view?

I think -- I would have to say there's always that potential. Could you model in an animal, but then it would be different than the actual system that we will be studying here. Sow never know what -- you -- so you never know what. You always come down to the final question: What will happen in the human situation? We think the SKIDU prediction is the progenitors do function well and induce differentiation, in which we have different plans of differentiation, which would be abnormal.

Thank you. My second point related to whether or not the transduced cells would engraft as well, better, less well, than non-transduced cells and clearly, as you have pointed out in your response, you will determine that in the course of your study. But, again, I'm wondering if you have any information from the SKIDHU experiment that might speak to that point.

Again, we don't have any more data than the fact that these cells will function in that SKIDU model. The previous experience with transduced cells, with retroviruses, they will engraft in the short term very well. We don't expect an effect on the grafting.

My third point related to the fact that some of the material originately submitted was too small for me to read, but you certainly submitted larger versions. Thank you very much. I think also you have commented on my fourth point related, again, to the impact of the vector on function of normal hematopoiectic cells and I think you have presented data on that. And my final point related to the emergence of mutants that would resist the effect of the RNAs that you are introducing, and you comment that viruses I didn't completely understand your response. You commented that the viruses, I believe, were not fit, and were out competed with a wild-type virus. But I may have misunderstood your response.

Well, yeah, actually, I can give you a somewhat extended answer to that question. So we studied this quite extensively. In particular, for the sh RNA 1, in under challenged conditions we do periodically get resistant virus that comes out. We have sequenced it and analyzed the point mutations. They are all within the target sequence of the sh RNA. The interesting aspect of this -- and we can look at the relative mutants, which is sometimes as much as 9 to 1. We clone out the individual PCR products of that infection. If we take that virus, and then the wildcat comes back again to dominate it. Suggests the mutant is an unfit form of the virus at least in normal conditions. In the clinical construct, we have actually gone extensively in terms of finding viral mutant. But because there's three different genes acting and blocking viral entity we keep the replication quite low, at least in monocites and macrophages. So firal breakthrough has not been observed by --

And the mutants that you found do they replicate in cells expressing the hairpin RNA better than wild type? I missed that in your --

No. Well, in the cells that are expressing the hairpin RNA, they will replicate. That's when you see viral breakthrough of p24 measurements.

But -- but are they fit in that circumstance?

It's hard to say that they are fit. It's just that they replicate. We believe that they are not being challenged with an antiviral and they mutate quite rapidly or the population that takes over is the non-mutant population.

Thank you.

Are you satisfied with that response? What I heard is that there's a concern remaining that cell lineages other than the monocite macrophage line, that the differentiation of other cells lineages might be impacted by the transchain.

I think that would remain an area that would be useful to investigate. Yes. And I think my other concerns have been addressed.

I just wanted to be sure. Thank you.

Can I just say one thing also because it came up in Dr. Rosenberg's review. The cells that are placed in the patient at the time of transplant are two kinds. The unmanipulated obviously are the therapeutic ones and all lineages will eee volume of from those. The question of the research is whether we can get linages, multiple linages by vectors and that's done by the study about. We don't believe that we will alter the differentiation of other lineages that are necessary for the patients health.

Thank you.

Let me pursue that just a moment, though. The work that you presented to us is specific for the monocite macrophage linage, correct?

Yes but in addition we have showed the T cells in the SKIDU.

Okay. Thank you.

And we have actually looked for plasma cites in that population and they are present in the normal ratio.

There's some data we did not show you, that was just new and that's Carl June has taken this clinical vector into his system we expands T cells using a stimulation called a C.D 3, CD 28 stimulation system for use as immunotherapy, and in that setting, these cells expand equally well with the control cells. So at least T cells can replicate very rapidly, but, again, it's the T cell system and that's all.

Dr. Powers?

Thank you for very comprehensive presentation, and it's an exciting first step in the research study. My review is confined entirely to issues arising in micromedical ethics. My comment -- biomedical ethics my comment is that the informed consent document seems a bit complex. Admitting, in fact, you have a very complex message to convey. And so I appreciate the inherent tension between the desire to be comprehensively informative and being comprehensible. I'm not quite sure you have struck that balance entirely right. The response was that the -- the IRB would be your source of assistance and that could be perfectly fine, and many institutions IRBs are quite good at that, perhaps even more helpful than you would desire. But nonetheless, in some settings, the IRBs don't have quite the resources necessary. So we often, in these instances recommend a couple of options for you to consider if the IRB is not all you would want it to be. Many medical institutions have some expertise available for really crafting these kinds ever things, and so I would encourage every -- every person who comes in with a very complicated research protocol to look for that kind of expertise in your institution, take advantage of it. The second thing is that we also appreciate that the informed consent document is just simply part of the process, of informed consent. And so it may well be that in instances where the document itself remains rather complicated and perhaps unavoidably so, that the process of the personnel involved used to convey the information could cure those kinds of deficiencies. But I'm quite happy if you are, that you can do that job and you will get a little bit more refinement in this territory.

And so this should remain a recommendation, that the language in the document be simplified and that the presentation of the informed consent document utilized a non-involved individual, is that what you are saying?

Yes, I think so. It still -- it still needs a little more work. I know the aim of getting an eighth grade comprehension level is an allusive vein at best in any of these. This one is a very difficult document, but a very difficult message to convey. I do appreciate that fact.

Thank you. It will take me just a moment. It's always complicated for me to do six things at once, so I apologize. I had a number of concerns. Most of them related to the selection of the specific patients who are going to undergo treatment. And the control Orrelltive control of HIV in -- or relative control of HIV in those subjects. But to begin, I had a couple of preclinical concerns which have been satisfied. The first was that I asked that either in writing or today that the NB drawing and the new mouse data that supports the strategy be summarized and that was done for us very nicely, showing that the transgene that's proposed, especially in the new mouse model is effective in that model and that's as much as I think we can expect. I asked whether there was any possibility of vector recombination with wild HIV, in such a way that the HIV path thisity, the wild path HIV pathogenicity. The issue of combination was addressed in Dr. Rossy's presentation and I'm satisfied that that is extraordinarily unlikely. I asked that the data summarizing the safety of a Lentivirus into atologous for transplant into those who have underlying immunodeficiency, from the myl ablation and from the HIV 1 itself, be summarized for us, and that -- the response that I received was in writing, and is fully adequate. The subjects themselves will remain on their anti-HIV therapy, their anti-retroviral therapy, and they will be monitored closely. They will be begun on proPhalaxis for opportunistic infections should their CD 4 count fall below 200. And I think that response is adequate. I asked that there be some discussion -- and I'm still -- I would like a little more discussion of this, actually, of why the individuals selected for inclusion in this study are going to be allowed to have viral loads up to 50,000 copies. That's a rather vigorous -- it's a large amount of HIV. And it seemed to me, at least, that the likelihood of adverse events would be decreased if the subject undergoing atalogous transplant cell transplantation had somewhat more controlled HIV. I'm wondering what your response to that would be.

I would ask Dr. Krishnan to come up and help me with this answer. I'm an infectious disease person. My only question is. What is the level that would make a person feel comfortable, obviously undetectable would be best.

Oh, we won't get there.

These people come through, multiple cycles of chemo which is usually chopped cycles which is a four-week cycle. They may go through two or three of these. They are quite sick and not taking their antiviral medications regularly as they would otherwise because of IG upset and that sort of thing. So their loads are if youating significantly. The hematologists have a problem of bringing this person who is approaching at least a partial remission, quickly to transplant. So you don't have time to manipulate the antiviral therapies, like you would like to do, in an HIV patient. There you would like to change the -- if you could do a genetic analysis of the voir us you would get, the -- virus, you would get, the genome type, you would get a good response of what the result is to this many nip ration but the clinicians don't really have this option. And I would turn it over to Dr. Krishnan.

My question is really the latter, does it really matter? And I guess if I had a goal, it would be less than 10,000 rather than less than 50,000 copies, merely because the relative well-being of HIV-infected individual is remarkably improved if one can get relow the 10,000 copy level. And I understand that that may not be possible. Dr. Krishnan.

I think we approach it in two ways. When we did our first gene therapy study with did he set the bar, it was kwai a high bar. -- quite a high bar. We used less than 10,000 and certainly we had minimal infectious complications. And subsequently we had 26 patients in total and we loosened that barrier really because we found, one, exactly what the doctor said. Those patients are not taking their antiviral therapy, and they need it for two months for this transplant. Number two, wee look at the ater Reese is, as -- ater Reese is, as its own safety, it will not mobilize an adequate number of stem cells to safely perform a transplant. We have just had that experience in a recent patient. The viral load has never really gone below 50,000. The lymphoma is very well controlled and actually in remission. We tried to mobilize him and we could not mobilize. So that patient obviously would not go on to transplant. And the last point, really is the issue of therapy with chemotherapy, as you are aware. This data, looking at the epox Rituxan, some suspend it by four months, and, again when they resume it, they haven't seen. So I think more importantly is screening patients for active infection and opportunistic infections is also another Safeguard we have built in.

Okay. And while you are -- while you are there, could my -- my next question had to do with CD 4 counts because you don't build in, in the protocol -- I understand what you are saying, that the apheresis procedure may on its own control for this, but in the protocol there's no inclusion/exclusion criteria based on C.D 4s.

Again, I think that is based on our prior experience when we did our first again therapy study. We had a C.D 4 count of 100 and we realized that was fairly arbitrary because we made that CD 4 count at the time of lymphoma diagnosis. By the time that patient has gone through -- these are very heavily pretreated patients and the CD 4 count is long gone. And then more recently, there's a paper from an Italian group that has treated 17 patients this. They actually had no CD 4 counts as a cutoff and they said quote/unquote no opportunityivic infection and we found it was more on an effective proPhalaxis, not the CD 4 count.

Let me see if I can try this another way because I'm still not quite comfortable. If a Phase I study which is looking at a new vehiclor, an anti-vehiclor, into -- that's not a new vector, about you a new use of a lentivehiclor itself, wouldn't it be wiser to include slightly healthier patients. It doesn't have to do with the CD 4 count of 50 or 30. It's more about how can you best guarantee that you will have usable end points to validate what you are trying to do in the five patients that you propose the procedure in. That's my intent.

I think that's what we struggle with. Our issues have been more with the underlying lymphoma in that sense and less the infectious risks. The electrotransplant screening does a -- the pretransplant screening does a good job of weeding out those patients. They get fairly intensive chemotherapy. First, that's a pretty good test of their immune function and tolerance before you go on to a transplant, and I guess my other point is we have four opportunistic infections in our patients. All of those were not correlated with CD 4 count but when they stopped proPhalaxis for a variety of reasons.

However, you know, I don't think we want to go at this very much further. There are opportunityivic infections such as PML where there's no proif I lax is. I'm -- proif a lax is. I'm concerned that this may not be, as currently identified the optimal patient population in which to ask the very important question that you are trying to ask.

And a appreciate your concern. But I do refer to you and the committee, Dr. Krishnan's survival curve. And I presume you have seen that and most impressive survival curve. It is a flat line at 85%. and this is using the method that she has developed. That's not looking at CD 4 counts, allowing a certain level of HIV loads. I think that our -- you know, I can say that because I'm not really part of the transplant team. These people are really good at what they do! and it is the selection of the patient -- that probably is as important as anything. And more important than CD 4 count and the viral load. I think that's all we can say. I think we would have to expect the community to rely on that record to reassure you that the patients are not going to be unusual, and also the experiment is not at risk. I don't want the experiment to be at risk. I want to see those patients at two years and three years and four years to see if the viral marking is still there.

That's really my concern. I am certain from your survival curve that the majority of the time this procedure is not going to be harmful to these patients. I'm comfortable with the risk benefit, however, you're only enrolling five patients. And that's a limited number anyway. If one ever or two of those -- if one or two of those subjects prove to be in the 10 to 15% who don't do well and prove to have very high viral loads and very low CD 4 counts, you would not gain max mall information -- maximal information that's my concern ?iesm guess the only comment I can make in terms of our survival data is really most of the mortality has been relapses of lymphoma and not so much from the HIV. And I think that -- you know we have little control over this as far as we use our standard lymphoma response criteria that we use for HIV negative patients too, but really most ever those early relapses have been from relapse lymphoma.

Okay. I think we'll close that discussion. I suspect it's a difference of opinion and I would just respectfully request that you rethink the issue of a limited patient number and the increased risk with that limited number of one or two disastrous events that might be better controlled if you had a firmer boundary on your patient accrual. That's all. Okay. I'm going top Ott discussion -- I'm going to open the discussion up to other members of the RAC. Dr. Simon, do you have -- go ahead.

Just to go back to the earlier point of looking at transcripts which a blast search would show have some homology to the targeting sh RNA. Is that a valid -- is that a useful criteria. In som senses you are pioneering two things, first SH RNA in clinic and secondly, just marking the Lentivirus of stem cells. Do you think that's a fair -- a fair request, I suppose?

I do. I do. And I -- I think we'll immediately go back and look at -- we'll do another blast search and see if we can come up with more candidates and anything that looks like a promising target, we will look at. Presumably it has to be something that's expressed in this matter for the population for it to be a VAM I had target.

Right. And -- valid target.

Right. And I understand that there was another strand with on SH RNA strand, and also spectrum basil.

The other strand definitely is not functioning at risk we really looked hard at that.

In the experimental basis that you provide that other strand you have expressed in the UTR, so in that sense, the accessibility issues for the target are altered, right.

We put a big -- in the context of a large piece of RNA to circumvent that problem. So we actually took a region that covers three of our targets which is several hundred base parafragment and put it in a free frame UTR.

So just for some clarifications of the R RNAs that you annal hissed -- analyzed, the 250 that you used, how many of those are human?

There were 250 human micro RNAs.

Okay. I thought there were 250 total. And then you saw that there was a variation in three idea on the chip but not confirmed by normal blasts but your HIV was expecting three shs in that context. So I suppose it would be nice to know that the one that you are using doesn't actually have an effect on the chip.

Correct. And we -- we can certainly go through and do that again but I think the important point is that even when those three were expressed, those two that are operated by the Mike other RNA array analysis are not detective in terms of prohi baition by their better means of atall sis. In my pound, the -- the issue here is that we are actually not very high -- we're not getting high level of expression of any of these sh RNAs in the context of Lentivirus vector encoding. We get 1.5 intergas station per cell so it's a single copy gene per this entire genome. And so the level of that RNA are very small. One thing I think Natasha alluded to as really these are all concentration dependent. If you go back and look at the Jackson data, if they titrate those down, they completely use these off targeting effects. Only really when you have the hyperfunctioning RNA will it maintain itself. So we think that -- I definitely will go back and look at these targets but I think that because we're expressing such low levels of the sh RNA it really becomes less of an issue but also becomes more of an issue for our breakthrough which has not been something -- without resistance, just the ability of virus to replicate the sh RNA. And that does happen with a single enteryentity. That and that's why the triple construct is so important. It also gives much more complete protection from viral replication in these cells. So just kind of -- that's the answer.

And the -- back Dr. Naomi's point, you did a molecular clone?

That was from from the viral infection.

There could be a wild type --

There was a wild type in there. You always see a mixture when you do these ratios. -- we completely lost the mutant on subsequent challenge to just 9 T cells.

I misunderstood you as saying the mutant reversed it back to the wild side as opposed --

No. No. It just lost its replicative advantage.

Reporter: And in the SKUDU model, you are putting back CD 4 cells and Naomi's point about where you spike back untransduced cells to ask about the efficacy of repopulation by the transduced cells?

So this work is being done actually by a collaborator at Colorado state university who has the SKIDU model operative. I don't think he's done that particular experiment. Basically what we do is compare the repopulation by vector back bone transduced vs. us is vector backed stem cell transduced and look at the ratio of GSP positive T cells that come out of that population and that's never really shown a significant difference. And of course we need to challenge those T cells with HIV to see if, in fact, the ones that have the therapeutic genes are functionally active and the answer to that is they were.

Two more questions. The first is -- and, you know, there's a controversial issue about the WPRE elements in HIV and in the context of EIA and SIV, so the -- the WTRE element that you are using, does that have a presence?

So the -- the only -- the reason that the WPRE element is still in there it discretely terminates transcription quite effectively, writing the string of reviews that are after the gene of interest, there's no read through into the WPRE element. We can't detect anything other than the transgenes being expressed from these vectors when we are intergated. I mean it could be a low level of read through but it's undetectable and even the syn elements are not completely silent. But in our, you know -- would you have to do quantitative RTPCRE to pick up those almosts so we don't know what those means as far as potential dangers are. We have plead it as safe as possible as we can, and translate.

Okay. And the final question, just as a clarification, you mentioned looking for the detection of an assay and postinfusion of 612 posttreatment. I'm sort of confused. Given that the patients already have aids, you -- how are you going to look for a replication of the Lentivirus? Because they should have a replication of the Lentivirus.

Yeah, they're active under the nomenclature problem here. We have clanged. What we will be looking at is the presence of vector sequences in the patient's own HIV. So it's a HIV recombination that we will be looking at. We have been calling it RCL but in the latest version of the -- of the pro toe cot,-- protocol it has been reclassified to a HIVrecombination.

So Dr. Simba, are your comments and concerns a wildcatly addressed? Thank you. Other -- Dr. Duhave you ever.

I very much enjoyed your presentation. Thank you. And I was -- I'm very pleased to hear about the experiment that you have with call June that relates a little bit more to the functional activity of the T cell lineage, not just the differentiation. I did have some questions, though, one was just a point of clarification on figure 6 and the additional data. Is that a challenge with NL 43? The reason is because most of the challenge data is with a five years in virus.

Yes, the T cells, from the SKIDU system and the question is obviously what strain was used by Dr. Akinan.

In the papers he used an L 43 but I don't know if that's the experiment.

I think that's in the paper.


So that's an L 43 challenge.


So let me just -- one thing that's striking in many of your challenge experiments is the very low MOI for the challenge. What happens when you do a high MOI challenge in these cells?

I -- I would not though those are low Ms. They are -- MOIs. They are in the microgram levels. People are usually showing nanograms so we have five lots of difference between the control and the vector. We have done those experiments that we have actually gone from a .01 to a .05 MOI change and what we see is that the sh RNA alone or SH RNA 2, or any sh RNA is less effective at those concentrations especially when it's ingrated in the Lentivirus vector, as a single or double. But the triple construct maintains its inhad a big activity. So -- inhad a big. Activity. So that was one of the criteria that pointed us in using the triple construct as opposed to any single sh RNA strain. We have proposed to our cells that we would go with a single hairpin strain, so we could test the HIV protection from RNAi and of course there are a lot of reasons not to do that meanwhile viral mutations but also we know the higher the challenge dose, the more the chance of getting breakthrough. I guess part of the answer to that question is that we also use this rye -- ribosome that knocks it down, by 85 to 90%. So we are blocking viral entry quite effectively. So even at the higher dosages this is C.CR 5 or endotropic viruses we are not seeing any major breakthrough at the triple construct but we are getting fairly robust of viral antigen with the other constructs of those conditions.

You are not seeing breakthrough with the GFRL or one experiment with the NF 43, but one question I did have is the extent to which the sn RNA is conserved between different strains and I wanted you to make a comment on that.

Yes, sir. We have actually done by-- biochromatic. I think 500 isolets and then you start to get smaller numbers that have pew gasations or variations within that target region. I think that will be the case for, many transcripts. So we chose at that time and rev because those are small proteins that have very little room for changing I prune knows says and they -- immunoassays and they could change code usage. I think the probability that there will be a virus that will emerge that will have mutations in the target region are high. I don't see that as not being the case. But hence having other antiviral genes present to add schematic protection. On the other hand, the -- you know the fitness of those vee viruses may inbound -- viruses may be in question as well. So changes in the at that time region, at least in our hands seem to not result in voir Russ replication competent as, you know, as starting the wild type virus.

Let me start questioning the patient selection. Have you given thought to the possibility of excluding individuals who you know had x4 virus on board and whose viral sequence might be divergent at the S RNA sequence because then they would be potentially resistant to two of the three targeting approaches, more or less from the get. Go and I realize those patients are probably quite rare but should your sample size be smaller, that might be a complication.

Let me answer that question. Again, we are forced by the tempo of the clinical scene, primarily in this regard. We will characterize the HIVs, certainly retrospectively but we did not feel we could do those studies as part of an eligibility screen within the confines of the timing of the patient access.

So in other words, the -- the -- because of the --

I think if this makes it to a Phase II trial would you definitely exclude patients who have --

Is there anyway -- just to focus on perhaps the x4 in particular, right there's certainly been the notion that the use of R 5 blockers could sled to the selection of x4 virus and it's conceivably more pathogenic S. it possible to have that data at least, initial information to try to select?

I think that anything is possible with enough effort. And it's a reasonable consideration.

I mean, I think I would at least like to ask you to consider. That I mean I'm certainly not sufficiently expert in that area but John Moore, people like that, might be people to discuss that with and determine whether that would be worth to implement a basis for subject collection.

-- selection.

Again, I think your concern is that we would coerce a selection of an r --

No. I'm not suggesting that you -- no, I'm suggesting that you may have a low level of x4 virus present and would you drive that to become the dominant form in an accelerated fashion and my understanding is that has been a concern with some co-receptor blockers pharmacologically. Thank you.

Dr. Piantodosi.

Yes, thank you. I wanted to offer some suggestions for improving this study design, and none of what I'm going to say will be really be overtly in the form of a question but I would be happy to sort of hear responses along the way. And so I'm going to give you three or four specific points that caught my eye in reviewing the study. And the first thing is when I attemptrd to characterize the goal of this -- attempted to characterize the goal of this experiment and the purposes or the nature of the design, and my concerns or suggestions in this point are typified sort of by the title. You call this a Phase I pilot safety and feasibility study, which I would consider to be three mutually exclusive and partially conflicting terms. And I think that it reflects how the investigators have tried to craft the study and so those three terms appeared in the title. I would argue, in fact, that none of them are correct. For example, phase one is usually an investigation of dose versus safety which is not. There's no dose question here at all. A pilot study, as far as I know is an ambiguous determine that has never been defined anywhere. Y certainly couldn't de-- I certainly couldn't define it. Safety and feasibility is something that's a little closer; although as clinical outcomes safety and feasibility, I think are beyond the scope of the purpose of this study, at least as far as I understand it. In fact, I think this trial is almost a perfect embodiment of what my closest colleagues would call a translational trial and that has certain characteristics that -- that I think if we pay attention to those characteristics it will point to some improvements that are possible in the protocol. First of all, it's small, and that's clear that that's your intent and that's an appropriate limitation for translational trial. It tests emerging therapeutic concepts from the laboratory which this clearly is, and I'm willing to stipulate that this is a great idea and that the investigators are superb at what they do. So I have no concerns about any of the underlying biology or execution of the trial. The translational study depends really fundamentally on the measured biological outcome that's thought to be an irrefutable signal of the disease and the paradigm of the disease and the efficacy of the treatment. And I will come back to this point, because I think that you have articulated that in the discussion here today, but I don't think that it appears explicitly in the protocol. The worrisome thing from a meth logical point of view that a translational study confounds three things that it's really important to ultimately separate. First is the accuracy of the disease paradigm; the second effect is the target validation; and the third effect is the therapeutic effect of the treatment. All three of those are compounded in a translational trial, as they are here. That's not a criticism. It's just important that they be separated very clearly in the protocol list as to how you are going to prevent, for example, an inaccurate target validation or the wrongA dime of disease before masquerading it as a therapeutic effect. So I think we need to be very explicit about what sort of study it is, and what the goals are and to be very explicit in the protocol about separating those three effects. So --

Excuse me, could you repeat the three effects for us.

The disease paradigm, in other words how does the disease work? And how does the -- how is the target valid, and how the therapeutic effect works. So if you are wrong about any one of those three, it may get confused with the other one. And that's very sipical for I think these extremely early developmental trials. So that leads me to be concerned about the study outcomes and you have proposed, I think most clearly in the protocol that safety and feasibility is the connect outcome and those are clinical outcomes, but, in fact, I'm -- I find them not well-defined in the protocol write-up. Safety is characterized a little bit in terms of short-term and long-term outcomes. But feasibility, as far as I can tell is never defined. You went through great pains in the presentation to say that we know how to do the transtection an -- transfection and that works. So that's not the feasibility study. We know how to do the bone marrow transplant, and so that's not the tooblem. Transplantation is an established therapy in this setting. So I think that feasibility doesn't -- doesn't really fit, according to the current write-up. And I'm concerned that five patients is too small a number to establish the -- either one of these objectives. We heard some concerns from Dr. Wara about sample size and I will come back to that again in a second. So I think what's missing from the current protocol are the following points: First, a clear articulation of the essential biological questions at this stage in therapeutic development and I said biological, not clinical, because I don't believe that the clinical outcomes are the ideal ones for where you are. Second, an explicit definition of a failure. What I mean by that is what could you see in your cohort of patients that tells you that that this is not a good idea or that you need to go back to the laboratory or you fide to -- or you need to take some step off to the side? The real purpose of this study is not to convince us all that we should be treating AIDS, HIV/AIDS lymphoma with this viral vector but the real purpose is to lay a foundation for some subsequent experiments. So we have to anticipate what it is that might indicate this would not work. So you said, for example, one of the things you hope to see were multiple lineages marked by the vector. So if you failed to see multiple lineages marked by the vector, what does that tell you as being the worth of this therapy? That would be an example. a third thing that I think the protocol should contain is a display of possible outcomes for the study and what experiments would be needed, depending on each of those outcomes. In other words, how does the information gained from this study fit within the developmental paradigm for this treatment? Some of those outcomes would tell you to move ahead with traditional types of development, some of them would tell you that you need to go back to the laboratory and so on. And so I think the protocol should indicate those. And then finally there should be a thoughtful discussion of how many subjects are appropriate, and moreover, what is the -- what are the properties of your experimental and the decisions that you are going to make, based on these data as a consequence of how many ever subjects that. Is I lrs interested that I -- I already interested that I didn't think -- I already hinted that I didn't think five subjects was enough. That it will be desirable clinically or biological. Here's an example, suppose that you observed some event in all five out of five subjects, or equally noted in five subjects. The precision that that would inform you regarding the next 5 or 10 or 50 subjects is only about 50%. You would only be reasonably certain that half of the individuals are subject to that event. Even though you saw it five out of five times. That's probably not a sufficient precision to tell you about what's likely to happen down the road with this therapy. Maybe it is, but if it is, then I think the protocol needs to articulate very clearly why that rather imprecise amount of information is sufficient for the developmental needs of the therapy. I think it would make more sense to imagine a translational trial where you would have roughly three times as many subjects, perhaps in stages of five each and that's far more likely to provide you with enough precision to make a decision that you have to. So finally, I just wanted to give you a sense of how serious I personally think these concerns are. That's just one data point from somebody who is a meth methamphetamine -- methodologist, and I do sit on my institutional review board and my seriousness is that they are quite serious and I do not yet think that this is a VAM I had research design, until you incorporate some -- valid research design until you incorporate some of the points that I mentioned about methodology and try to sharpen up the way that the protocol was written. Thank you. That's the end of my comments but I would be happy to try to clarify any of it, if it would help.

I -- I thank you for those well-thought out words of advice. We actually did work on how to assess feasibility and we have the text in the protocol, which obviously doesn't convince you. The question in our mind is: Will this vector appear in the progeny cells. If you had five out of five that shows the vector appears in all lineages, I think that would be -- that would move the field, no question about it. If you had none appear, no sector marking in any of the cells that you follow, and you can measure expression levels, et cetera, et cetera, that would move us as well. We would do what we did with the retroviral vector in that same setting. It does not work in that particular model. So I think that from our standpoint, those are the kind of observations we would learn from this, and that's the -- the feasibility assays that we do, the measurement of cells that are marked by vector and the amount of transgene that's expressed, is something that would be better if you had 15 patients, of course. So that's notpossible within the limits of the reality that is -- it's such an expensive study to do. But if I could just say one other thing. And then the question is: Will this vector recombine with the wild voir Russ? If the patient's -- wild virus? If the patient's wild virus does not have any vector, it will move the feel as well, it did not recombine. But if it does recombine that will certainly have an effect on this field. That is your vector now can interact with the virus. Those are the real issues that we are attempting to -- to address, and I understand that from a disease-specific standpoint, I -- I accept everything that you said, but I would like to hear what you had to say relative to our goals.

I'm not the least bit uncomfortable that those are worthy measurements in which to take and evaluate the effects of the therapy. What I'm not comfortable with is whether or not you should be moved or the field should be moved based on five out of five events or 0 out of five events. It's the quality and inprecision that the decision is based on, not the nature of the measurement that you are taking. So that's -- that's what's making me uncomfortable. You said that you would go in a particular direction. If you failed to observe any events in five subjects, and I -- I guess what I'm trying to say is, is that an appropriate -- or is that property appropriate for making such a major decision? And my instincts say no, it's not. Because -- it could easily be that 100% is the correct event rate and you have observed zero perfect. That's not a very unlikely circumstance with the denominator of only five. That's what's making me uncomfortable.

Is anyone else uncomfortable?

I'm not sure if this solves the problem. The focus is on five patients but it's not really five events in those five patients. You have, I don't know how many stem cells you are infusing into a patient. It's maybe that number, and you have been looking for one event out of that number of stem cells that you can detect, you know what the sensitivity of the detection is. The same thing for the recombination, it's not five chances at recombination, it's however ever many stem cells and wild virus have got mean those tells. Those are the -- stem cells. Those are the number of recombination chances that you have for detection and that's a large number. You have a good sensitivity in terms of your assays for detecting a single event. Again, I'm not sure, honestly, because I'm not expert in that field whether that's a legitimate way to look at it. But I think that's what they are saying when they say, yeah, we have a sufficient sensitivity to equivocally determine that. Is that a fair statement?

That's a fair statement.

Well York want to engage in a full debate about the statistical properties about these types of hisured outcomes. I would just point out that the cells within a person or not, independent of one another, and so the fact that you count thousands of cells doesn't suddenly convert you from a denominator of five to a denominator of thousands. It does not work that way. And I think some of my colleagues around the table have pointed out concerns about eligibility and the characteristics of these individuals that might sort of throw you a curve ball and my quip before this experiment, subject to, are quite different or unusual, or somehow not using the same disease paradigm that you built this study on. But in any case no, matter what kind of outcome you have, it's clear that you don't have much leverage. You are making huge decisions based on five subjects. And I'm -- I'm just concerned that the properties of those decisions are not really up to the task of -- in this setting. That's all.

Thank you. Other issues for discussion? Comment by members of the RAC? Ms. Kwan?

I just wanted to go back to a simpler form of the questions raised and that's Dr. Powers' comments about the informed consent document and I believe in the preliminary discussion, before the decision to make full public review, that I also had read the informed consent document, and found it quite complicated and I think one of the responses that we got back was that these -- this particular group of patients was probably very used to reading and understanding highly technical information. And I would like to try to change the view you have in writing these informed consent documents. I'm going to go with the premise that perhaps you already have kind of a standard informed consent document for the fundamental treatment for the lymphoma. And that this is a second document to supplement it because you are doing a second thing. If it isn't going to be a separate document, I would recommend that it would be a second document. And on that assumption, I think it might be -- make it easier to write a clear, understandable informed consent document if instead of seeming to make your goal an eighth grade level description of a high complex scientific procedure, instead you looked at it from the potential participant's point of view, and tried to describe for the participant what is it that is going to make his or her particular treatment than somebody just ordinarily going through a transplant. Are they going to be less comfortable? Are they going to be at greater risk of an untoward event? And if so, what type of additional discomfortables? What additional risks are they taking on and not attempt to describe the scientific process which they probably are not as interested in as how will they feel? And what are their -- how are their chances of recovery or non-recovery affected by this additional labor of -- layer of experimentation? And I think that would be more useful to a potential subject.

I have a few other comments. I got side tracked on the subject inclusion/exclusion criteria during my earlier discushion and failed to -- discussion and failed to put into the record my comments on informed consent, which I -- which I do want to do. My first comment was not precisely what Ms. Kwan just stated but very close to that, in that if the document is to be left as is, then the components dedicated to stem cell therapy, which is part of therapy for the patient, be very specifically separated from the research component which is proposed. And the investigators have provided us with a sentence which I believe moved towards that goal. I also asked that there be some language on the consent regarding the timing of the proposed pregnancy tests to assure that if any women enroll, they will not be pregnant. As well as a requirement of an obla Tory agent for any woman who does enroll. The response I received back was we will abide by the IRB decision. Some say it's insensitive to a young woman's feelings as she sents a potentially sterilizing procedure such as this. Although I certainly agree with that, it's essential that in this cape of protocol be -- this kind of protocol, there be careful wording in the consent, not only suggesting but actually stating the methods of birth control that any woman subject will be asked to undertake. So I would leave my comment in. And finally, I asked that language be included in the consent regarding the requirement for anti-Fungal and PCP proPhalaxis, as well as the IBIG infusions in the subjects and that language is going to be inserted, as energy understand -- understand it. And then I have one other question to ask you. Could you just briefly address the biosafety and public health precautions that you propose because you will be working with Lentivirus in HIV-infected subjects and for the record, we'd like to know what you are proposing. Thank you.

The patients are housed in a bio-- in a setting that is a Hepa filtered room that meets all the universal precautions that we normally use with any patients undergoing a transplantation. The institution also has a biosafety official who oversees inactions of any type of reagents that are produced in a B.L 3 lab and this reagent is produced in a lab that's similar to a BL 3 lab. I would say that these patients in the transplant are protected from contact with others in various ways, primarily through the air flow that's in this hospital. It's a new hospital. It has quite impressive Hepa air flow, which is an OR level flow in the rooms and in the corridors outside the room, in fact. And when they go outside, they do wear masks to protect themselves from other patients. So I think that there is no major concern in that regard. In terms of patients spreading infection.

Additional questions regarding biosafety from the RAC members? Thank you. Dr. Caplen, you had one final comment.

Yes, for the regard, I would also like to indicate that I had asked the informed consent form to include additional details about the RNA experience, the novel approach, the new approach and that they would be potentially, if this goes forward, the first patients receiving this type of approach.

And that was agreed to.


Thank you. Okay. Are there any other comments from members of the RAC? Members of public? Dr. Borrow?

I just had a couple of comments on the informed consent document. One was a common comment, is the use of the term "Gene therapy." I think we generally recommend that that be "Gene transfer." I had also a question about the alternative treatment section. It states here that you have been informed that alternative is not to participate in this research protocol but instead had the chemotherapy, including the transplantation with the genetically modified cell and I wondered in that was a typo and if it was without the genetically modified cell.

Yes, that's a typo.

I think there were some other typos similar to that early and the words missing and particularly on page -- well, section two, it says the virus to activate its own grown, and another one was inserted into using a particle, so I think there's some words missing or --

Actually, on that line, I know it's sometimes very difficult in doing these documents to try -- technically these are not genes. That they are actually putting in these are RNA transcripts. They are not coding for any protein.

On that note, are there any further comments from the public? All right thank you. I will then attempt to summarize for you what is left, and what the members of the RAC should be voting on. And I have divided this into preclinical, clinical and informed consent comments. Under the clinical -- the preclinical comments, just a moment. One, additional work would be useful unitelizing biochromatic and microray analysis to assess the likelihood of an after target interaction between the processed sh RNA in the transgene and the cellular transcript that speaks to Dr. Caplen's remaining concern. Two, it would be useful to determine the effect of the transgene on cells other than of those in the Mondayio cite microphage lineage. This speaks to Dr. Rosenberg's remaining preclinical concern. Clinical. And this is where we may want to insert some modifications so I would like all the RAC members to listen carefully. One, it would be useful to reconsider the inclusion/exclusion criteria to better assure that the outcome variables in the five patients to be accrued or optimally interpretable. Please reconsider -- and I have carefully used the word "Reconsider" inclusion criteria that restricts enrollment to patients with a more well-controlled viral load, query, less than 10,000 copies. As well as a CD 4 count, greater than 100. The reason I have inserted the queries is I don't mean to impose my or our definition, merely to ask that the investigators reconsider the patient population of the size that they currently plan to accrue. Any thoughts about how we could reword that? All right. Two, it would be useful to reconsider the exclusion of subjects with x4 virus rather than CCR 5 given the limited number ever subjects to be enrolled. That's Dr. Duhurst's comment and, again, it would help to bring this to a more homogenous patient population. So that the outcome variables could be more carefully assessed, perhaps. Three, concerns remain regarding the study design, which would be improved by "A" clear articulation of essential biological questions; "B" explicit definition of failure in clinical terms "C" clear articulation of the possible outcomes of the study, with clear statements regarding subsequent stebs to be -- steps to be taken from these outcomes to inform the underlying scientific paradigm; "D" reconsideration of the number, of subjects to be enrolled in this translational trial to better assure increased precision. That's terrible. To assure increased precision to guide subsequent decisions regarding future research. Do those work? All right. Informed consent document. We had a number of comments on the informed consent document. Beginning with Dr. Powers. One, further asitance with the document to simplify the content and wording as well as conveyance of the consent by an uninvolved individual. Two, the protocol informed consent document for the proposed gene transfer should be separated from the informed consent document for the therapeutic stem cell therapy, for the stem cell therapy. I think that's better. That's consideration of. That's not a mandate. Three, please include specific guidance for women subjects regarding recommended birth control, both off la Tory and barrier. -- ovlatory and barrier. Four, please utilize the term "Gene transfer" rather than "Gene therapy" throughout the document. That's what I have. Are there additions, changes, for members of the RAC? Dr. Powers?

I think there were a couple of additional risks that you and Dr. Rosenberg subjected to include in the informed consent document. -- suggested to include the informed consent document.

I'm satisfied.

I thought there were a couple of residual risks that you wanted communicated.

No, I'm satisfied. I think they were responded to regarding my concerns. I just feeded to read them into the -- the public document.


The public record. All right. I would like a motion for approval. Second? Dr. Powers second. And we will vote beginning with Dr. Powers.


Ms. Shapiro re-

Aye. [ Voting ] All ayes.

Did I miss someone?

No. We can't have them vote. All right, thank you very much, and I want to thank the presenters. This is a complicated protocol with many pieces in it, and our goal is to assist, not to -- not to prevent or retard. So thank you very much for really straightforward, clear presentation this morning. and we will -- where's my -- should we go directly to that? Okay.

Dr. Wilson has wonderfully agreed to talk to us this morning about a user's guide to FDA draft guidance, again therapy clinical trials observing participants for delayed adverse events. a subject close to our hearts. Thank you Dr. Wilson.

Okay. Thank you. Waned to thank OBA and the RAC for actually providing us an opportunity to come here today and provide a summary of this guidance document. It's my goal today to hopefully answer some of the questions that might be in your minds, and obviously answer any others that might -- that my presentation doesn't get to. The main question being what are the key events that led to the development of this guidance and what are the scientific reasons why we think some gene therapies carry long-term risks and how does one term, if you are a sponsor, whether long-term observation should be performed in your particular clinical trial? And then finally, what exactly do we mean by observing participants for delayed@verse events? If you go over a little bit of history long-term follow-up, actually dates back to 1993, when the FDA September a letter to all sponsor -- sent a letter to all vectors for life long guidance. And this was then republished in 2000. That was really based on the concerns that were raised by an experiment in monkeys where it was shown that if you are contaminated with RCR, that you have a risk of developing things like lymphoma and so that raised some concern about long-term surveyance of these subjects and, of course, this has been more recently high lighted by events in clinical trials with we see insertion of genesis in the absence of RCR. So the next major event was in 2000, we issued another letter to, in this case, all sponsor of gene therapy, clinical pro toe call and this is really done due to observations that we had made in review of data submitted to us over the previous again -- ten years that subjected there was a lock of good conduct study monitoring and that concluded that there was incomplete guidance among sponsors to doing long-term follow-up T. wasn't a requirement at that time. It wouldn't have put you on clinical hold, in other words. And that then initiated a series of BRMAC meetings where we brought up discussion of vectors and other types of protocols. We had three different meetings where we discussed this with our advisory committee, the biological response modifiers advisory committee and the transcripts are available, if you are inned in reviewing many, many pages of discussion, but it culminated internally with us in 2001, really trying to come to grips with a variety of recommendations that were made by the advisory committee and we implemented finally the recommendation to perform long-term follow-up of subjects in all clinical trials regardless of vector. Well, over the course of the next several years it became clear to us that that wasn't the most feasible, nor scientifically justifiable for us to take. And so for that reason, in June of last year, we worked with ASGT and other members of the gene therapy community to cosponsor a workshop on long-term follow-up of participants in human Genome research and the main focus of that workshop was really to focus on the lack of the scientific basis for these recommendations, the lack of details for how to perform long-term surveyance, the 2001 recommendations were never written in the form of guidance documents. There was a lot of conclusion on the part of sponsors about what we really wanted and discussions also included, these consequences for long-term surveyance. So with that in mind then, after going to that workshop, we, internally, tried to take into account all of the discussion and input we received at that workshop, and the output is the guidance document that was issued in August. And we focused an two main properties that we think have the potential to cause delayed events. And the first is related to persistence of vector sequences. The non-persistence per se, perCifuentesence in combination with either integration or a vector which has the potential for lateeny is reactivation, and then secondly the concerns raised with transgene specific effects. So in terms of perCifuentesence, we feel that long-term risks will be influenced by the macro influence. It's clear now that if the vector integrates you have the potential for persistence vector sequences. We think that vectors that have the potential for latency is with. We know that herpes virus upon reactivation, do carry the risk of encephalitis, the immune status of the subject may also play a role in the outcome of long-term risks. Transgenes specific effects are harder to characterize, but clearly, if the transgene itself has tumor genetic effects as in the preclinical study that would be of concern. If the particular context may induce an autoimmune disease, for example, an immune genetic disorder where the patient may not have seen that particular protein before, or expression of a gene that normally would be tightly regulated and may also cause long live term effects, likewise jean expression in a tissue where it's not normally expressed may be a concern. So how do we take all of this together and then assess whether or not a particular protocol should be subject to long-term follow-up? And what we have tried to do is come up with a fairly simple sort of question/answer Al Gore algorithm to address -- algorithm to address and the framework I will show you here is also in the guidance document. The first question is: Is your gene therapy product only used for modification of cells. If the answer is no, then the next question is you are using an invo and is there vector sequences. If the answer there is no as well then we feel the risk is low, and long-term follow-up observations may not be necessary. On the other hand, if you answered yes to either of these questions, we then trigger into two more questions. Are the sequences integrated? Does the vector have the potential for latency is and reactivation? No to both of these questions, again, we feel the Riis sk low. -- risk is low. Yes to either of these questions, then we feel clinical protocol should include long-term follow-up observations. And the guidance document, we provide, a number of details about how one would perform preclinical suddenies to determine persitence -- studies to determine persitence. If essentially a biodistribution study, which people have been doing for numerous years and is probably not very different than what they are used to, the main point is that you need to include multiple time points. We have defined persistence in that -- in a manner that is somewhat flexible, because persistence, per se will vary depending on the particular animal model that is used, and so we don't want to say, persistence means longer than three months because that may not be applicable in a different type of animal model. So instead we said that it is present throughout all time points and study without downward trend over seven time points. Now, the question of integration or latency is. If there's no information that your particular vector with regard to eat of these points, then we recommend that you could look into this as part of your persistence suddeny if it's tease -- study if it's feasible. Otherwise for most of the commonly used vectors we took a position that most of these vectors have a lot of both preclinical and clinical data to support whether or not they have a propensity to integrate. Now, we deliberately used the word "Propensity" because we are aware that things like plasmas and anti-viruses can on rare circumstances at low frequencies ingrate. So we recognize that -- integrate. So we recognize that this is not always an absolute yes or no. But if it's not designed particularly to integrate, we are saying that it doesn't integrate, and therefore, no long-term follow-up would be required. The two exceptions of the most commonly used vectors are herpes virus because we know that has the potential for latency is with reactivation and retroviruses and I want to make clear that we used the term "Rhett sew virus" -- "retrovirus" which is the family retro virada which includes the Lentivirus. If you have no information about your particular vector with respect to its integration potential or latency is potential, and you can't perform these studies in preclinical models, then the default would be that you would have to perform long-term observations. Now we do also include some exceptions. So if you do a study and there's evidence for persistence, no integration, but, again, this really is more of a transduced specific effect that you are seeing your potential for long live term toxicity, then we would want to follow those patients longer term. And likewise, potential for autoimmune response, and then also coming back to table one the commonly used vectors, we are also aware can be modified in a manner that can change that integration potential. You may change the mode of delivery of a plasma. People have made insertions into the antivirus that are now integrating vectors and so on. So clearly if you are doing something that deliberately changes that quote propensity, then that would change the answer to those questions. We also include in the guidance document explanation of some special considerations when retrovectors are being used and this is really just written guidance on what we have been implementing since January of 2003, in response to the ex-clinical trials and we apply these recommendations when the retroviral vectors are used with high replicative capacity and long survival. So things like hematopoiectic stem cells. So, for example, if it's hematopoiectic stem cells we would be looking at peripheralblast. We recommend that you test for retrosequences every six months for the first five years and yearly. If no vector is detected then you can discontinue this assay. We also included in a section some information about key points that we think need to be included in the informed consent document, in order to accurately reflect the risk of cancer. This was something that's been highlighted by our advisory committee discussions as well as other advisory bodies. If vector is detectible, and at least 1% of surrogate cells, then we also request that you develop a method to assess the pattern of that vector integration site. This is really looking at whether or not you see a polyclonal or oleclone ol' or monoclonal. If it's going to the monoclonal pattern, we would want you to compare those to the human Genome and determine whether or not the gene could be affected and do closer study for signs of malignancy. Now let's go through the clinical ramification of long-term studies. Again going back to the framework of the yes or no answers. If you get to the bottom and determine that the criteria do suggest that your vector or context is associated with a higher uncertain risk, that information about the product as a whole indicates a need for long-term follow-up, then this section would apply to you; however, even when you have nose conditions emerge, if the clinical context is such that we think or you think that the studies has no scientific value, there may still be exceptions to doing this. And we think that the determination of whether or not a particular study would have scientific value of doing long-term observations are related to three major characteristics of the population. If you are doing this in an end stage cancer population, with a short life expectancy of 3 to six months, doing long-term follow-up is probably not feasible, first of all, at no going to be inform -- and not going to be informative. And certainly they will likely have multiple morbidities and other carcinogenic aagents. So we think interpreting events in this type of population make it's very difficult. Now, if you are doing follow-up, the recommended duration is for 15 years; however, again, there's some exceptions and there may be reasons for doing shorter Durations. -- Durations. For example, if there's information following these patients and vector persists for a year and you can't find it past that, then we feel there's no need to continue following for 15 years. Wakelize, the -- likewise, the translation of transgreen would be a key determine meant. Again exposures of the study population, expected survival rates and other relevant factors that you think may be important for consideration would impact how long you are doing the follow-up. Now, during the first five years, the observation is a little bit more intense, with a more passive method for the subsequent ten years. It's very important that they find information to include systemic case histories. We can't compare anything in two or three years if we don't know what the patient started. With it should include gene therapy events and related exposure to mutagens and it should be obtained by visiting a healthcare provider annually where that would record new malignancies and other hemotoe logical disorders. These are the type of long-term events that we think are most likely to be related to gene therapy pro toe protocol. And we think -- protocol. And we think it's very important at the outset to make sure that study participants will understand why it's important to do long-term observations and hopefully take an active role in cooperating with the collection of this information. Now in the next ten years, if as I mentioned a little bit less intense. It's an annual contact with specific screening only if indicated. And just continuing to follow up as indicated by results from previous years. So, again this comes into whether or not you see evidence for vector or transgene expression and other information you may have collected. If an event occurs that is serious, you need to report that to FDA using expedited reporting as designed in the CFR. Otherwise, this information can be compiled and reported in an annual report. If there's an emergence, an adverse event, thin we would probably -- then we would probably work with the sponsor to determine the additional follow-up that might be required, and, again, if it's technically feasible, we would recommend testing for vector sequences at least annually. Now, one question that has already come up is -- well, I have an NIH grant, it's for five years and how can I use 15 years of follow-up. I put this up as a reminder because actually after some of the trials, with this game about, the national center for research resources worked closely with the NIH to develop a mechanism for doing long-term observations of NIH-funded studies. And they provide support for subject visits through their general clinical research centers, blood sample storage is I willble through the NGVL at no cost to the HIN-funded studies and clonality is also provided at no cost by the NGVL. And I provided the url which has detailed about this particular program. I wanted to reiterate, finally, that this guidance is a draft guidance. What that means is that it's open for public comment for 90 days. So I would really urge people to review it, think about it, see what works, what doesn't work, and provide us your feedback. That's really valuable because we will read those comments and -- and revise the document accordingly. And you can do it by one of two ways, old-fashioned way, by written comments in the mail, or electronically. And then finally, if you have additional questions, as you go through the document, please feel free to contact me if I can't answer it all for me to refer you to the right person. Thank you for your attention.

Thank you very much, Dr. Wilson. Questions or comments from members of the RAC. Dr. Duhurst.

I think this will be very useful for people doing vaccine studies because there's a great deal of confusion!

I should clarify, actually in the first sentence of the document, this does not per taken to vaccines! [ LAUGHTER ]

So is there any guidance for those people? Because they really have a lot of confusion, even, you know, at different IDCs there's questions about whether these regulations apply.

Dr. Sentic.

This only applies to products that are in the office of cellular tissue and gene therapy. We really don't have any control over guidance written by vaccines and currently, there isn't a long-term follow-up study although there is follow-up by vaccines with particular sponsors. So, we just really can't comment about that, and so if you have concerns I would suggest you contact that office. I mean, ray legally to don't mean to push this aside. You have to understand this guidance document only relates to the products in the office of cellular tissue and gene therapy.

The reason I bring this up is because I'm not talking about your extended vaccine, but the vector vaccines.

Still, but they are regulated by a different office within the center for biologics and they take a different approach to these issues. It's -- it's a different patient population, for one thing. So in some cases, it's quite ofstified, to have a different approach to how we view these issues.

Just one suggestion I might make is once do you finalize the guidelines that you at least notify the IBCs sort of nationally about this guidance because they may find that helpful in trying to evaluate whether or not a sponsor is supposed to be providing them with long-term follow-up. Because the way the guidelines are written right now, the NIH guidelines the impression that many IBCs is they are supposed to get long-term follow-up but what you are saying is in a sense is one should be rational about that and one should have a scientific basis for requiring it.

We hope so.

And that might be helpful for the IBCs to understand.

Just a quick comment. OBA can take on the response of notifying the I.BCs and is willing to do that. I think we should proceed.

I would also like to point out that we have made this available to almost every society that we know exists, that this would pertain to. We are getting this out as fast as possible and plan to make everyone aware of the availability of this.

Dr. Heslop.

I would like to say that it's a very helpful document that clears up a confusing area and I would like to thank you.

I thank you for the record too. I have one -- and I will come to the public in a moment, sir. I have one question that occurred no me as I was listening to your presentation. With regards to NCRR's generous offer to pay for and sponsor the long-term follow-up, it says that the GCRCs will support the subject visits. Do you know whether they will support the laboratory tests?


Routine laboratory tests.

Yes, it does.

It does.

And that's in the document?

Yes. Yes. If you go to that urlT describes everything that's supported. I just didn't put that much detail on the slide.

That's terrific. Thank you very much.

Dr. Abell dah.

I would like to say as someone who used short-term vectors that you made things clearer.

Ms. Shapiro.

Does the document address at all the participants not to comply with long-term follow-up?

We're certainly aware that that is an issue, and that's why -- obviouslyT can't be coercive, but we feel that it should be incorporated as part of the informed consent at the outset of the protocol to make sure that they are aware of this and to try to encourage participation as best they can. We do have a number of sponsors who are compliant with long-term follow-up, who have trouble tracking subjects or subjects to withdraw and, clearly, that is their legal right and we certainly acknowledge that.

Member of the public who had a comment. Sir?

A member of the public is one of the authors. I'm Dan Rosen Rosenbloom I'm the clinical reviewer who participated in it and I wasn't sure if everybody was getting the message that it's the sponsor's responsibility to examine the product and write the letter to the FDA going through the arguments either in favor of or against doing a long-term follow-up study which we will review and concur or not concur depending on information which we may or may not be able to share with the member. Your question about following every individual, this is a plan for a study, just like the protocol itself. And not everybody is going to comply. But the reason that we put in patients being aware, what we meant is to put that in the informed consent for the study that it is our expectation that you will be followed for 15 years, and this is what we mean by that, five years of observation, so that the subjects who are probably much more motivated to do this than you might think will provide us with the information, because the goal is to get the information. Okay.

And one final question for those of us on the RAC. I read that this is a preliminary document.


When it is finalized, will it be all right to put it on the OBA web site?

I -- I would assume so, yes. And it would certainly be on our web site as well.

Can we put it on the web site now? Just for our use, our internal use?

I think you can. I mean, I will check on that, but since I sent it to a number of institutes, like bs for instance, the American society of gene therapy put it on their web page or at least announced it. I mean, maybe what the best is to announce it and put our site.

Yeah, just link it to --


That would be fine. It would be just be very helpful to us as we review future protocols and consents. So thank you. Thank you all very, very much. Any other comments from members of the public? RAC members? All right. Then the 101st -- that's a long number! RAC meeting is adjourned. [ End of meeting ]