Saturday, January 12, 2008 3:45:55 PM
Alzheimer's Panel - "Relative Optimism"
This is a 32-page Report from Credit Suisse discussing all of the approaches, and even numerically handicapping the various clinical trials, by a Panel of 3 experts. One thing I have learned is that you can't always trust those knock-out mice.
As this may be the longest post ever placed on iHub, and the formatting may not work out, you may prefer to email me at my yahoo address (not hard to work out), put ALZ Panel in the subject line (to distinguish from the spam) and I will send the PDF in response.
Date: January '08
Alzheimer's Disease - Relative Optimism from Diversified Experts
*
Three weeks ago we assembled a panel of cutting-edge researchers in
Alzheimer's Disease (AD). Our purpose was to provide background on
disease pathology and identify interests in new therapeutic prospects from
the mid to late stage pipeline. This group had relative optimism on a number
of new mechanisms and saw several as potentially playing a lead role in
future therapy (e.g. WYE/ELN's bapineuzumab), but did not see any one
option as the means for eradicating the disease. The transcript from the
panel is included in this note.
*
The amyloid cascade is the point of attack for all late stage therapies. Two
panelists were optimistic that this cascade is critical to the disease course so
key drugs taking this approach should be meaningful in some way, implying
commercial promise. Even the less positive panelist felt this approach may
yield acute cognitive effects that would be embraced by the AD community.
*
Our panelists believed that therapies that reach the marketplace in the next
3-5 years should provide incremental benefits. Given the prevalence and
severity of AD and current lack of disease modifying drugs, therapies that
provide even incremental benefits (along with reasonable safety) could be
potential billion dollar blockbusters. Using these agents in combination with
each other could help to truly modify the course of the disease.
*
Our panelists expressed the greatest optimism for a passive immunotherapy
approach with antibodies directly targeting amyloid beta (Aβ). The leaders
in this approach are WYE/ELN’s bapineuzumab (recently began phase III),
LLY’s m266 (recently finished pII dosing) and PFE’s 4360366 (in pII).
*
Our current valuation of WYE/ELN's bapineuzumab suggests a $2.5Bn
opportunity (using a 50% probability of success adjustment along with 20%
market penetration and $25,000 price/pt estimates). We see this as
conservative and our panelists directionally agreed that there is at least a
50% chance that bapineuzumab (and other passive immunotherapies)
proves to be an advance in the treatment of AD. Current WYE share price
provides a free call option on bapineuzumab, with pII data in mid-late ‘08
likely to be positive and, if so, providing more positive support to the stock.
*
Our panelists were also optimistic about the potential for gamma secretase
modulators and other drugs that alter the Aβ 42/40 ratio, but felt that Myriad
Genetic’s Flurizan may not be potent enough to have a significant benefit.
Each panelist expressed some concern about the safety of therapies that
completely inhibit enzymes involved in generating Aβ (e.g. gamma and beta
secretase inhibitors) because these enzymes have other important functions
in the body, but they anxiously await more data.
Transcript from our Expert Panel Discussion
Catherine Arnold:
We're going to go over the neuropathophysiology of the disease. We're going to talk about
the amyloid hypothesis. And we're going to talk about the tau hypothesis to set the stage
for the balance of our time together which will be looking at the drug pipeline and after that
we'll go into full Q and A.
Now as far as background, I would bet everyone in this room has been affected by
Alzheimer's disease in one way or another. With that experience you're undoubtedly
aware that no approved options are available that which actually modify the course of the
disease. The roughly three billion in drug sales in this area come only from symptomatic
treatments. With an aging population, the Alzheimer's market will grow faster over the next
several decades. And estimates now are about 26 people million worldwide and growing
to about 100 million people by the year 2050. Couple this with the fact that we've got
several drugs in late stage development, which hope to actually modify the disease and
we see this as a critical investment theme for 2008 forward.
I'd like to introduce you more specifically to our panelists. I want to comment, though,
that we really worked to put together a panel of individuals with very different perspectives.
You'll hear some very strong positives and strong criticisms on a variety of topics, but we
will work through them to try to reach closure and see if we can identify a consensus.
Now immediately beside me for those of you here is Dr. Paul Aisen, who until November
was running the Memory Disorders program at Georgetown University. For those of you
who don't know, he's transitioning. He's going to be a professor in the Department of
Neurosciences at the University of California San Diego. He has a wide breadth of
experience. He's studied almost every drug that has been studied - symptomatic and
disease modifying. He should have a very broad view for us. He's collaborated and
designed a number of large NIH funded trials. He's worked with the pharmaceutical
industry.
I'd next like to introduce you to Peter Davies. Dr. Davies is a Professor of Pathology and
Neuroscience at the Albert Einstein College of Medicine. He's a chair of the Judith and
Burton Resnick on Alzheimer's disease and the Scientific Director of the Litwin-Zucker
Center for Research on Alzheimer's disease. Now he's been working in the field for
obviously a very long time. His bio says he published his first paper on Alzheimer's
disease 31 years ago. And that led to the development of Aricept and Exelon. A lot of
colleagues think of him as being an expert on tau. Although as I've gotten to know him I've
come to appreciate that he's actually very involved in other aspects of the neuropathology.
So I think you'll find him to be very insightful.
Our last, but not least, panelist is Dr. Rudy Tanzi. He's currently a Professor of Neurology
and Neuroscience at The Harvard University. He's the Director of Genetics and Aging
Research, at the Mass General Institute for Neurodegenerative diseases at Mass General
Hospital. Going back to 1987, Dr. Tanzi focused his studies on Alzheimer's disease, and
isolated the APP, which we associate with Familial Alzheimer's disease. In 1995 he was
with the identification of presenilin 2, which is obviously important to the process. He
has also been involved intimately in the discovery that the metals zinc and copper are
necessary for the formation of neurotoxic assemblies. He's been very involved obviously
therefore in the amyloid hypothesis and we're really pleased to have him here.
I've asked Dr. Aisen to start us off with a high level overview of Alzheimer's disease
neuropathology. And talk about current treatment options.
Paul Aisen:
Thanks very much.
Okay so I can summarize current treatment of Alzheimer's disease on this slide. Dr.
Alzheimer described the disease just over 100 years ago. And then we did little about it for
3/4 of a century. Dr. Davies and others in the late '70’s developed the cholinergic
hypothesis that first suggested the possibilities for treatment. This led to the first small
positive study in 1985. At that time the pharmaceutical industry awoke to the possibility
that AD was not a degenerative disease about which nothing could be done, but rather a
specific disease that could be effectively treated. This led to the approval of tacrine eight
years later and three other cholinesterase inhibitors. Finally memantine was improved in
2003. So a total of five drugs have been approved in this country for AD treatment.
Okay so I want to start by just saying for two minutes a little bit more about what our
standard therapy for AD is today. It's all focused on the synapse. On neurotransmitters
that allow communication from one brain cell to another. The most important
neurotransmitter for cognition is acetylcholine. Each of the first four drugs approved act on
the cholinergic pathways - specifically they inhibit acetylcholinesterase and thereby
increase the activity of acetylcholine in the synapse. So there's a cholinergic deficit in
Alzheimer's disease. That deficit can be partially compensated by reducing the
breakdown and therefore increasing the activity of the acetylcholine that's left in the
synapse.
Of the four cholinesterase inhibitors - again they are tacrine, donepezil which is Aricept,
rivastigmine which is Exelon, and galantamine which is Razadyne - they are all approved
for the treatment of mild to moderate Alzheimer's disease. Donepezil has recently
extended its indication to severe Alzheimer's disease. Rivastigmine has recently extended
its indication to cognitive impairment in the studying of Parkinson's disease.
Memantine is the first drug of a different class for the treatment of Alzheimer's disease. It
is also acting as a synapse but not directly on the cholinergic pathway. Rather,
memantine is acting on the glutamate pathway - a specific glutamate receptor called the
NMDA receptor. Memantine is a partial NMDA receptor antagonist. It seems to cause
some hyperactivity at glutamate's interaction with the NMDA receptor, and the net result is
improved cognition. Based on its mechanism, there was hope that it might reduce
neurotoxicity. That is cell damage. There is not clinical evidence that there's any disease
modifying effect with memantine. But clinical evidence suggests that its action on the
glutamate pathway augments cholinergic activity and provides a symptomatic boost
generally similar to what we see with cholinesterase inhibitors. But the two work well
together. So a cholinesterase inhibitor plus memantine provides an additive symptomatic
medicine.
So this slide summarizes where we are today at the end of 2007. Standard therapy for
Alzheimer's disease is symptomatic by and large, including a cholinesterase inhibitor. One,
but not more than one, of the drugs I listed. Memantine is particularly effective at
moderate to severe stages of Alzheimer's disease. It's only approved for moderate to
severe. The results in mild disease show a very small - an effect, but a very small effect.
Practice varies from one clinician to another, but a standard evidence based approach
would be to start a cholinesterase inhibitor as soon as possible after diagnosis. And then
add memantine to it at the moderate stage of Alzheimer's disease.
I put Vitamin E here as well. We used to use Vitamin E to treat everybody with
Alzheimer's disease on the basis of oxidative stress. In one positive treatment trial, that
component of the regimen has been reduced recently as we've learned about some of the
long toxicity that may be associated with Vitamin E. So Vitamin E is really no longer a
standard part of the treatment regime. Rather it's a cholinesterase inhibitor with the
addition of memantine later on.
And I would just point out that we have a huge problem in the area of Mild Cognitive
Impairment, which is the symptomatic prodrome to Alzheimer's disease. And we don't yet
have any effective therapy for the millions of people who have MCI - Mild Cognitive
Impairment.
Now just in the last few minutes I want to set the stage for the discussion of drugs in
development today. I don't think we have any disease modifying therapy for AD. But I do
believe that we're getting close. And that's what we'll be talking about. Disease modifying
therapy is not focused on the synapse. It's focused on the neuropathological lesions in the
hope of changing the course of the neurobiology of Alzheimer's disease.
So Alzheimer described two primary lesions, the plaques and the tangles. The plaques
are extra cellular deposits. We now know that they're made up of aggregated A-beta
peptide. So it's a fragment from a larger protein called APP or amyloid precursor protein.
The fragment aggregates into plaques that are very dense and sticky. And they deposit in
the brain substance, and in the walls of the blood vessels.
The second hallmark that Alzheimer described was the tangles. Tangles are also
insoluble deposits of protein. But they occur within the nerve cell body, rather than in the
vessel wall and the brain - extracellular brain tissue. And we know that the tangles are
composed of an abnormal form of a normal protein. The protein is tau. In the tangle tau
has been chemically changed, hyper-phosphorylated. Too many phosphate groups added.
The result is that the tau changes confirmation, changes shape, loses its function and
aggregates into the neurofibrillary tangle.
These processes are summarized in this cartoon that was developed by the Alzheimer's
Association. On the right side you see the healthy components with a healthy neuron up
on top, so a healthy neuron and here is a neuronal cell membrane. Within the neuronal
cell membrane there's a normal protein called the amyloid precursor protein. There are
enzymes in the cell membrane including gamma secretase and beta secretase and alpha
secretase. You see here drawings of healthy microtubules. Microtubules are the tubes
that are essential for transport of materials down long neuronal processes. The
mitochondrion is here, which is the energy producing organelle in the cell. So this is the
healthy cell.
In Alzheimer's disease we know that we get plaques which are made up of the amyloid
peptides, which is a piece of that normal amyloid precursor protein. There is a clipping of
the amyloid precursor protein to release this amyloidogenic fragment. Amyloidogenic,
meaning naturally sticky so that it aggregates into oligomers and then fibrils and plaques.
The aggregation of this peptide occurs at the same time as there are abnormalities in the
tau pathway with hyperphosphorylation.
Tau normally helps support the shape of microtubules. When tau is hyperphosphorylated
and the tangles form, the microtubules change shape and lose their function. We can also
see evidence of damaged energy metabolism, damaged mitochondria. So somehow we
go from this normal appearance to these abnormalities in the course of Alzheimer's
disease. Disease modifying therapy presumably needs to interrupt the pathways involved
in these changes. The last couple of slides that I'm going to show you is a formulation of
an approach to disease modifying therapy, that puts all of this into a single context.
I know Peter will have comments to make on this formulation, but many of us believe that
the process must start with the cleavage of the amyloid precursor protein, by beta and
gamma secretase to release that peptide. That all the other changes are downstream to
that event. We release the amyloid peptide. The peptide is toxic, probably in multiple
forms. Leading to neuron death, mitochondrial damage, hyperphosphorylation of tau and
tangle formation. Many of us believe the pivotal event is generation of the amyloid peptide.
The reason we think that is that it's the most concise explanation for all of the genetic
causes of Alzheimer's disease.
There are a number of genetic causes. Some families carry a mutation in one of three
proteins. Those proteins are APP, PS1 and PS2. We now know that all three are involved
in generation of the amyloid peptide. The mutations of the precursor protein and PS1 and
PS2 all increase amyloid peptide generation. The PS1 and PS2 have turned out to be
components of the gamma secretase complex. And the mutations associated with familial
AD increase generation of the amyloid peptide in its most toxic form, the 42 amino acid
form.
The other genetic cause of Alzheimer's disease, Down's Syndrome, where there's an extra
copy of chromosome 21. And the APP gene resides in chromosome 21. That extra copy
results in too much APP. Allowing excess generation of the amyloid peptide. We believe
that this formulation where the beginning of the process is, cleavage of APP to release Abeta,
is the most concise explanation for all the known genetic causes of Alzheimer's
disease. If that's the case, then if we want to develop disease modifying therapy, we
should fit our strategies into this slide.
I've broken down the current programs in this way. Theoretically strongest would be to be
on the left side of the slide and inhibit beta secretase or gamma secretase. It also might
be effective to attack directly the amyloid peptide with immunotherapy. Or with small
molecules that interact with A-beta and promote from the brain. Or, this encompasses all
of the anti-tau therapeutics which I also find interesting and encouraging. But we view that
as a method of reducing the toxicity of the amyloid peptide. So I've put all of the
approaches on a single slide that focuses on A-beta. I will stop here.
Catherine Arnold:
So we should then move on to Dr. Tanzi.
Rudy Tanzi:
Okay, well I think with that very succinct and elegant presentation from Dr. Aisen, it will
save me some words. So I would begin where Dr. Aisen left off. I think it - I'll reiterate and
reemphasize the point that all four established genes we know about in this disease (the
three early onset genes and the late onset genes apoE) all have in common that there's
excessive accumulation of A-beta aggregate in the brain. And this has led to originally
what was called the amyloid hypothesis, which I think is now sorely outdated. Only
because it - most of the emphasis was on plaques. And mature aggregates and fibrils of
amyloid. And I think, as I'll make a point on my presentation, the battle ground seems to
be more toward the synapse and more toward - which is what we originally heard from
Bob Terry decades ago. And we should have listened to him - and that it's the soluble
forms of A-beta, oligomers, namely dimers and trimers, that seem to be more problematic
than the fibrils.
And so I, in my own papers, I like to refer to this as the Synaptic A-beta Hypothesis. Since
I think amyloid hypothesis doesn't really do the mechanism justice. And in the first slide I
show the genes you heard about just now from Dr. Aisen. The APP gene discovered in
'87. We know 21 mutations in this gene now and duplications. And the average age of
onset there is 55.
Presenilin 1 is the real work horse. One hundred sixty five mutations, actually that's up to
about 170 now. And the average age of onset there is the earliest of the three, 44 years.
And that accounts for at least half of early onset familial AD. And then finally presenilin 2,
where there are 11 mutations and the average age of onset is closer to 55. The mutations
in these three genes account for about half of early onset FAD. But early onset FAD is
probably only about 5% or so of all AD. So these mutations in combination account for
just a tiny bit of AD. Maybe 1% or 2% at most.
But what - I want to make a point going beyond what Dr. Aisen pointed out about their
molecular consequences. And it's not - I think it's been - these mutations have not been
properly addressed by biotech and pharma in the past. Because first there was the
misinterpretation that they increase A-beta production. They do not. Maybe a couple of
them do. Most of these mutations don't increase A-beta production. And most of them
don't even increase A-beta 42. What most of these mutations have in common is they
increase the ratio of A-beta 42 to A-beta 40. About 90% of A-beta in the brain is 40. About
10% 42. And you also see some species anywhere from 37 to 43.
If you look at these mutations and what they do to A-beta, generally in any which way they
can they increase the ratio of 42/40. In some cases that means that 42 stays the same
and 40 goes down. Recent evidence also shows that if you flood a transgenic mouse
brain that has FAD mutations of these three genes with A-beta 40, just bringing that ratio
of 40/42 back (in other words, reversing that molecular phenotype) reversing the 42/40 by
actually flooding with 40, you can prevent amyloid. So it really is all about the
stoichiometry and the ratio. Not just the absolute amounts of A-beta and 42. I think when
you're looking in vivo in the brain. And that's what these mutations have taught us.
Now what's so bad about that ratio? The A-beta 42/40 ratio, as it increases, drives the
increased oligomerization of A-beta and aggregation of A-beta starting with the small
oligomers. And there are data now from a variety of groups that these small oligomers
play a role in regulating long term potentiation or neurotransmission of the synapse. I
think Roberto Malinow's data at Cold Spring Harbor are particularly compelling that this is
a normal event. You need A-beta. You need A-beta oligomers to act as a check and
balance on long term potentiation and excitatory synapses.
So I don't think you just want to turn A-beta off. But as you can imagine in any genetic
situation where you have a toxic function, in this case if you have too many oligomers, that
ratio gets too high. Then you're going to be impairing LTP. Rather than just, you know,
acting as a negative feedback. And then you get toxicana function. LTP is turned on too
much. And cognitive impairment ensues. So I think this is the landscape that's evolving
for the new revised version of what was called the Amyloid Hypothesis.
The next slide is on the late onset respect to apoE. And apoE4 is present in about 20% of
the population. In at least one copy. And that increases to over 50% in AD. Thus the
basis for the association with AD. One copy of E4 increases risks three fold. And two
copies increase risk greater than ten fold. This is a rare case of multiplicative transmission.
Not just dominate or additive. But you actually get a multiplicative affect where there are
two oligos versus one oligo. There's really no other example of a risk factor - a non
penetrent risk factor in genetics, never mind just AD. So apoE4 is quite unique.
And the meta analyses that Lindsay Farrow and also Debra Black in our group have
published, suggest that the strongest effect is between 60 and 70. But not limited to those
age groups. E4 works - you see it in early onset, and also in late onset. But the real effect
concentration is 60 to 70. And again apoE4 is associated with increased A-beta
aggregation in the brain.
Next slide - just a way of for people who are looking at - looking for Web sites to keep up
on the newest AD genes. This should be said that the three early onset genes - and
together with apoE, account for only about 30% of the genetic variance of AD. A full 70%
of the genetic variance is not accounted for by these genes. Remember apoE is in 50% or
60% of AD cases. But it is neither sufficient nor necessary to trigger the disease. Other
genes seem to be necessary to interact with it. And also of course lifestyle and
environmental factors.
This alzgenes site - this is something we maintain. This is alzgene.org. The Web site is
updated weekly. And what we do is we take - we keep track of every published genetic
paper in Alzheimer's from day one. There are actually about ten new papers per month.
And if a gene polymorphism or variance has been tested in at least four independent
samples. Either in a single study or a cross studies, we take the data and the meta
analyses. And say well, does this gene hold up?
And we think this is useful because otherwise you kind of hear about yeses and nos. You
hear so and so saw this gene was associated. Someone said it's not. You don't know
what to believe. So this a more of an objective, systematic collection.
And if you look on the top gene - the alzgene top results. Those are listed in order of effect.
So apoE is number one. Most people don't know that the second strongest effect late
onset gene is actually the nicotinic acetylcholine receptor beta 2 based on meta analyses.
So this is updated weekly. It's something to keep track of. Just for those of you who need
like the equivalent of ESPN Sports Center for AD genes.
So in the next slide, then this gets into the Amyloid Hypothesis. Basically mutations in
APP and in the presenilins - early onset mutations as I said earlier drive up the A-beta
42/40 ratio. And that ratio then going to the left side drives oligomerization of A-beta.
ApoE can play a role in this. And the metals zinc and copper - we found 15 years ago in
my lab are necessary to drive that aggregation process.
If you - you can add all the A-beta 42 and 40 you want into a test tube. If you leave out
metals, zinc and copper in particular, you will not get oligomerization and aggregation. So
this is a metal dependent process. And we had basically discovered this back in '92, '93
when we working on the discovery of the Wilson's disease gene, which is a copper binding
molecule. I saw similarities with the Wilson's protein and A-beta.
Going to the right side of the slide, the way the brain likes to deal with A-beta is that most
of the A-beta that's accumulating in the brain as monomer is rapidly exported out of the
brain into the plasma. David Holtzman's work and Beth Ozoklovich's work has shown that
on average an A-beta peptide is exported from the brain eight minutes after synthesis. It's
the second fasted peptide to be transported out of the brain if it stays monomeric. The
problem is if it becomes oligomeric. And again I think there's a normal role for these in the
synapse based on Malinow's work. It's tougher to get rid of the oligomers. They tend to
accumulate. And because of that the problem is that you start to see these oligomers, and
ultimately aggregates and fibrils.
The other way the brain deals with this is degradation. You can degrade A-beta in the
brain deliberately with enzymes such as neprilysin, insulin degrading enzyme, angiotensin
converting enzyme. Or the A-beta can be degraded in the periphery once it's exported out
of the brain. So based on this - and this is what we've learned from the four known genes
- production of A-beta, the ratio of 42/40, changes with the mutations. These ratio
changes drive aggregation together with metals. ApoE plays a role there. I should
mention that on the clearance pathway, apoE is also one of the chaperones that helps Abeta
get out of the brain and into the bloodstream.
Alpha 2 macroglobulin is another one of those chaperones. And the receptor used there is
probably the LDL receptor related protein LRP. Now based on this, and you've heard
some of this already from Dr. Aisen. In the left top text box there are beta and gamma
secretase inhibitors that are being developed by companies like Lilly, Merck and others. I
think there's more promise for gamma secretase modulators. Gamma secretase
modulators try to reverse that ratio of 42/40 that's going up to bring it back down. And I
think that the best way to treat a disease when you have genetic information is simply go
in and try to reverse the molecular phenotype. The molecular phenotype of those
mutations is increases in 42/40 that drives disease, so try to reverse that ratio. And that's
what the gamma secretase modulators that, for example Flurizan from Myriad. And also
much more potent direct gamma secretase modulators not based on NSAIDs as Flurizan
is being developed by Eiasi, Torrey Pines Therapeutics, Merck and others.
One could also try to enhance alpha secretase activity. This can be done with M1 agonists.
It can also be done with nicotinic agonists that are being developed as well. And
cholesterol drugs have also been implicated in reducing A-beta production. Statins have
not done very well in the clinic, as most of you know. But I think that ACAT inhibitors have
based on transgenic studies and preclinical studies. ACAT inhibitors hit much more
selectively into the correct part of the cholesterol pathway that affects A-beta production.
And those data from Dora Kovax are very encouraging.
Going to the other boxes - the vaccine which is being developed by several groups. I think
there are at least 15 companies in the vaccine space. We'll probably get into that.
Basically the peripheral sink hypothesis, which is the right side of the slide, maintains that
antibodies to A-beta, either introduced by passive immunization or by active immunization,
can sequester A-beta in the plasma and not let it back into the brain.
This is the hypothesis that was championed by Steve Paul at Lilly and (David Holsman).
Elan, namely Dale Shenk, favors the hypothesis that antibodies can sneak into the brain,
trigger microglia cells that then digest A-beta directly in the brain. It's possible that both of
these mechanisms of action are at work. And I think we'll probably get into more detail
about the nuances of vaccine in this - in the Q and A.
And finally the left side box next to oligomerization, anti fibrilogenics, Neurochem has
Alzemed, which is a gag based drug trying to block aggregation. And I think we're going
to discuss that as a discussion point.
Prana biotechnology, a company that I co-founded, is targeting the metals zinc and copper.
We have a compound called PBT2. Dosing on it in Phase II finishes actually today and the
results should be available sometime early next year. PBT2 had tremendous results
preclinically in preventing A-beta aggregation in the brain, clearing A-beta and also
rescuing the effects of A-beta on long term potentiation and neurotransmission. And this
drug has good bioavailability in brain. So it'll be interesting to see the Phase 2 results
given that the preclinical results provide for a lot of optimism. I want to make the point that
that drug is not a chelator. It does not actually chelate metal. It simply - it's more of an
ionifer. It exchanges metal from A-beta. And can deliver it to the cell. Or deliver it to other
enzymes like SOD that use metals. It is not anywhere near the affinity of a chelator. And
should not be considered a chelation therapy.
Finally Transition has a very interesting compound AZD-103. This is a scyllo inositol that
also seems to have an affect on both preventing oligomers from forming and protecting
from oligomer induced LTP deficits. And I should make the point that for both of those
compounds, the Prana compound and the Transition compound, they not only prevent
oligomerization, but to some extent, especially the Prana compound, they can dissolve
existing oligomers as long as they haven't been permanently cross linked.
So these are just an outline of some of the therapies that have been developed that are
most - many are in the clinic, many are preclinical. Based on the pathways that have been
elucidated from the four known Alzheimer's disease genes. And I'm sure we'll get further
into these therapies during the conference.
So I'll end my comments there. Thank you.
Catherine Arnold:
Okay, thank you, Rudy. To round out our third topic for our overview before we go into Q
and A, Dr. Peter Davies, who is going to talk about an opposing hypothesis of tau.
Obviously tau has a relationship to amyloid as well. But I've asked him to talk about tau,
the cell cycle series and otherwise. So thank you, Peter.
Peter Davies:
Okay you heard some of the introductory remarks about plaques. And some of the very
obvious causes for therapy here. But - and this is the most popular scheme - the Amyloid
Hypothesis in some version or another says that some mysterious form of amyloid occurs
in Alzheimer's disease and causes this cascade. I don't have time to (disprove) this
hypothesis. I could speak for the next hour and a half about why I think this hypothesis is
completely wrong. But nonetheless it's a hypothesis that is (worth) testing. And I don't
want to come across as someone who says we should not try amyloid therapeutics. This
is a hypothesis. It deserves to be tested fully and thoroughly. I think it will prove to be
wrong. But we'll see. The test of the hypothesis will tell us whether it's right.
Talking about the tau approach. Much, most, almost all the attention, and certainly in big
pharma and in most of the smaller biotech, is focused on development of kinase inhibitors.
Dr. Aisen mentioned the hyperphosphorylation of tau. And presumably if one can prevent
phosphorylation of tau, one can prevent tangle formation. The problem really here is
indentifying which kinase is the critical kinase of phosphorylating tau in Alzheimer's
disease. We don't actually know which kinase phosphorylates tau in the normal brain. So
it's a choice of a kinase deposit in Alzheimer's disease is quite difficult. GSK3 beta and
CDK5 are the major targets. There are a number of publications, particularly from
AstraZeneca on GSK3 inhibitors. We still have really very little idea about whether either
of these kinases has a role in tangle formation.
We don't actually know the phosphorylation of tau is a driving force for tangle formation.
It's quite possible that tangle formation could occur and phosphorylation be a later event
not driving tangle formation. We actually don't know at this point whether there is any role
for kinase inhibited therapy.
If the amyloid cascade hypothesis is not correct then it's obviously incumbent on us to
suggest an alternative hypothesis. And the way we look at Alzheimer's disease is
fundamentally different. The way I envision Alzheimer's disease is as a process that
begins and gives rise to amyloid deposits and tangles made of tau. Rather than having
either amyloid or tau at the top of the cascade, our approach basically sees amyloid
deposition and tangle formation as the result of a biochemical process. But what is that
process? The first clue really came for us from - well from Inez Vincent who showed that
antibodies that were very good at labeling tau were also very good markers for dividing
cells, like cancer cells. And a couple of the antibodies that have been made in my labs to
study tangle formation in Alzheimer's disease, are actually among the best reagents in the
world for looking at dividing cells. And they're used now widely in the cancer literature as
a way of marking dividing cells.
So what do dividing cells and Alzheimer's disease have in common? Well the fact is that
now over the last ten years a large number of markers typical of dividing cells have been
found in neurons in the brains of patients with Alzheimer's disease. There's a lot of
evidence. A lot of very consistent evidence that the cell cycle, the cell division mechanism,
is activated in neurons in Alzheimer's disease. Carl Harris and recently Paul Lorenz has
reported that there is full duplication of chromosomes, full replication of DNA in neurons in
the brains of patients with Alzheimer's disease. This does not occur normally. This is not
the normal process. And Lorenz in a publication just this year confirmed this with three
different very sophisticated mechanisms.
There's also an increasing body of evidence that turning on the cell cycle is actually a
response to a number of different insults in brain. As two papers in the last couple of
years showing that experimentally produced strokes in rat brains turn on the cell cycle.
And the cells that turn on the cell cycle are those that die following a stroke. This is again
very consistent data from some very reputable labs in this field. So turning on the cell
cycle may actually be a response to a number of insults.
We ask the question, well does it really matter? I mean when you see evidence of cell
cycle activation, does it matter to a cell? Is it just something that happens as a
consequence of the disease? And is really an epi phenomenon. Or is it a critical event in
cell death? So we created a mouse with (unintelligible) in collaboration with (unintelligible)
to deliberately turn on the cell cycle in neurons in the adult brain. And we made a mouse
in which we expressed an onkogene, a viral onkogene in adult neurons in the mouse. And
what did we see in these mice? Well the title actually kind of gives it away. It induces
Alzheimer like tau and amyloid pathology. There's evidence - very, very clear evidence of
cell cycle activation. There's very, very clear evidence of tau pathology in these mice. And
there's very compelling evidence of amyloid deposition in these mice. All the features -
that actually really surprised me. All the features that are typical of Alzheimer's disease
can be generated, at least in part, in a mouse just by turning on the cell cycle.
So we've developed a revised scheme that says cell cycle activation may be the what in
the flow scheme. And clearly it is possible to develop at least some amyloid deposition,
some tau pathology, some cell death simply by activating the cell cycle. And we know with
very strong evidence that cell cycle is in fact turned on in Alzheimer's disease. So
obviously we don't think it's a viral onkogene that's turning on the cell cycle in Alzheimer's
disease. But it's quite plausible that human onkergene is activated in response to a
number of stresses. A number of injuries to the brain can in fact be activated. Can
activate the cell cycle. And can lead to the pathology of the disease.
And this really opens up a new avenue for therapeutics. Because one can draw, in fact
very readily, on the vast experience in cancer research to develop therapies for turning off
the cell cycle. Or turning down the cell cycle in the brains of patients with Alzheimer's
disease. If the hypothesis - again it's a hypothesis that cell cycle activation leads to the
pathology of Alzheimer's disease. But maybe we can get upstream even of amyloid
formation. And block the process that leads to the amyloid end paths of Alzheimer's.
I'll stop there.
Catherine Arnold:
Okay, excellent.
We hope that background provides some fodder for our next part of the session, which is
I'd like to go through some of the key pipeline programs. And what I like about our panel
again is that they all have very different points of view. But what I want to do after we go
through them is see if we can reach any kind of consensus on what our view is. So I'm
going to go through each project and I'm going to ask one of our panelists to give us a
take on what the hypothesis for it working is and what might be the detriment to this
approach. Then I'd like all three to weigh in on the relative benefit prospect from a scale of
one to ten. So if you think not about symptomatic therapies, but think only about what you
know of the pipeline for disease modifying therapies. Ten having the most likelihood of
success. One having the least likelihood of success. If you could each give me your point
of view on how you think this program will pan out. We love to commit things in numbers
here on Wall Street so if you could just bear with us today, panelists, we'd be appreciative.
The first program, and I'm going to bring this up because it's the most recent data set that
was released from a late stage program. I'm going to ask Dr. Aisen to talk about, which
is the Neurochem program for Alzhemed where there was some disappointing study
results in August. And I'd love to know, Dr. Aisen, what your view is on what this
mechanism which relates to beta aggregation. And how promising, you know, after you
step back from that experience, how you're feeling about that mechanism.
Paul Aisen:
Okay so Alzhemed is tramiprosate. That's the chemical name. It's a small molecule. It
interferes with the interaction of the amyloid peptide with glycoseminoglycans in the brain.
And the result is that it blocks aggregation and deposition of amyloid. And we know that in
transgenic mouse models, which are the standard models for testing anti amyloid
therapies, that treatments with tramiprosate reduces the accumulation of amyloid in brain.
The program moved to humans where Phase II studies demonstrated that there was at
least some degree of penetration into the brain with oral tramiprosate. Some dose related
reduction in amyloid peptide, the 42 amino acid (unintelligible) in the cerebral spinal fluid.
And as a result the program moved into two large pivotal Phase III trials. Eighteen month
trials with 1000 subjects to see whether these anti amyloid affects would translate into
clinical benefit.
And the results that were announced in August were that while the trial results are very
difficult to analyze as a result of methodologic issues. Mainly in an 18 month, two country,
two language, 70-site trial there was a huge degree of variability in decline, in use of
concomitant medications, including cognitive enhancers that influenced the primary
outcomes in the trial. The analysis is very, very difficult. But it can be said the results of
the trial were not sufficient to support a clinical benefit.
And the company has decided to halt the second pivotal trial and move to try to develop a
form of tramiprosate that might give more consistent and robust brain penetration. Before
moving forward with additional clinical trial.
So what does that mean for the amyloid hypothesis, or more specifically, anti aggregation
therapies? I think what it means is that there was not a robust benefit with this - these
doses of this specific agent. I do not think that that has a significant effect, at least on the
way I look at anti aggregation approaches.
So the closest program in mechanism to tramiprosate would be the scyllo inositol program
that Rudy Tanzi mentioned. Scyllo inositol is chemically entirely different. But it's also
small molecule that interferes with aggregation and reduces amyloid accumulation in
transgenic mice. I do not think that the negative results of the first pivotal trial with
tramiprosate reduced enthusiasm for the scyllo inositol program. Because I think it's all
going to come down to the amount that can be tolerated. The amount that gets into the
brain. And the degree of anti aggregation and amyloid clearing affect that a specific
intervention brings us. So while discouraged specifically for tramiprosate, I would not say
that I'm discouraged for anti aggregation approaches or the broader anti amyloid
strategies.
Oh, you want a number. Okay so let's assume as Neurochem has suggested, they are
going to develop a form of tramiprosate with better brain penetration. Allowing a more
robust anti aggregation effect. Where would I put that on a scale of one to ten? A four.
Catherine Arnold:
Dr. Davies, would you comment on where - on a scale of one to ten. And if you have any
comments you'd like to make.
Paul Aisen:
You can go below one if you want.
Catherine Arnold:
For emphasis negative numbers are very positive numbers aren't they.
Peter Davies:
I won't go that low. But I would certainly be in the one or two range there. But I do agree
with what Paul said. That this is not an adequate test of the hypothesis. This is a failed
drug, or drug in this particular formulation. I don't think one can draw conclusions about
testing the hypothesis that amyloid aggregation therapies are, you know, useless or
helpful. I don't think one can reach that conclusion.
I should point out and it's terribly difficult research to do. Because the mouse models in
which these compounds are tested are very artificial models. In which amyloid deposition
is driven by over expression of the precursor of amyloid. And one can show in those
models that lots of things prevent amyloid deposition. Or even amyloid synthesis. But that
- in the mouse at least there's really virtually no downstream consequences of amyloid
aggregation. There's not tangle formation. There's no cell death. And so it's very hard to
get a read on these therapies. And one can say in the mouse that they do or do not
influence amyloid deposition that. But beyond that it's very difficult to say very much.
So the predictive value of the mouse models is quite low, I think. And doesn't help in
developing these therapeutic strategies. I'm not (unintelligible) yet, but…
Catherine Arnold:
Dr. Tanzi, your vote?
Rudy Tanzi:
Well I would echo those - both those comments. That I don't think Neurochem's Alzhemed
should be a litmus test for whether anti A-beta aggregation therapies will work. I would
also give it a very low score, a one or a two. Because I think that it's hitting too late. The
devil's always in the details. And it's hitting the - it's blocking the ability of A-beta to bind to
gags. And this is kind of a late stage event when A-beta oligomers and protofibers are
starting to deposit. And I agree with Peter on one thing. I don't think the plaques are
where it's at. And this seems to be hitting more late in the path way. If you think about
monomer going to oligomer, going to protofibril, going to fibril. This is hitting late in that
pathway toward deposition as a plaque or as a fibril. And I just think it's unfortunate. And
that's probably why the drug is failing in clinical trials so far. So I would give it probably a
one or a two as well.
But I think there's still great hope for the Transition AZD-103. And in my opinion even
greater hope even for the Prana PBT2 drug.
Catherine Arnold:
So while you're talking, Dr. Tanzi, who don't you give us a hypothesis for why PBT2 will
work when PBT1 was disappointing? And what your vote is there on a scale of one to ten.
Rudy Tanzi:
Okay, I'm not sure PBT1 was disappointing. But PBT1 was clioquinol. And this is based
on the fact that zinc and copper are necessary to drive aggregation as we discovered.
And if you just keep zinc and copper away from A-beta, it won't aggregate. I think Prana
got kind of a bum wrap because people were equating it with chelation. And it's not a
chelation therapy. These molecules are no where near chelators, which have very, very
strong (unintelligible) metals. And don't let metals go. These compounds take metals
from A-beta but then exchange them, or even deliver them to the cell.
In fact in the preclinical mouse studies with PBT1, clioquinol, copper levels went up in the
brain because it was being released from plaques where it was being sequestered. PBT2
- the problem with PBT1 and the disappointment came because its di-iodo contaminant
when it was being ramped up for a Phase IIb study. The Phase IIa studies were actually
encouraging, you know, not significantly positive. But encouraging enough to go to Phase
IIb. But the di-iodo contaminate killed clioquinol. And I think that was a good thing.
Because that drug had some baggage.
PBT2 is a non-iodine derivative. Like clioquinol or PBT1, it's a 8-hydroxy quinoline. It has
about 50 times the bioavailability of the brain. Much more soluble. And in the preclinical
studies just blows PBT1 away. And if you look across mouse studies with vaccine versus
gamma inhibitors versus PBT2, I would place it right up there at the very top. Which is
why I am optimistic.
And I actually think that if - in my - and Peter will like this. I think if PBT2 doesn't work in
humans given the bioavailability and what is done in mice, I will probably start questioning
the A-beta hypothesis myself as a therapy. So I would probably put PBT2 up at around a
nine. But I'm biased.
Catherine Arnold:
At least your cards are on the table.
Rudy Tanzi:
Yes.
Catherine Arnold:
Peter, what do you say?
Peter Davies:
I still can't get much enthusiasm for these compounds. I just don't - I think the amyloid
deposition in - many of the arguments I see based on these very rare genetic cases.
Drugs are being tested in patients who don't have mutations. Don't have mutations in APP.
Don't have mutations in presenilin. I think the amyloid part of the disease is downstream.
So I think you might get a little benefit. Because I don't think a head full of amyloid is good
for you. But - and you might get some benefit out of those somewhere along the lines of
the benefit we get out of Aricept, which might sell some drugs. But, you know, as a
disease modification mechanism, I still have to be down around the (unintelligible).
Catherine Arnold:
Okay. Dr. Aisen?
Paul Aisen:
Well I think that targeting the metals that contribute to amyloid aggregation is a very
intriguing and exciting idea. I am happy to see that there is a compound that can move
this strategy forward beyond clioquinol. And so I also am looking forward to seeing results
with PBT2. I would take strong exception only to Rudy's last comment, which is that the
failure of this program would dampen the whole hypothesis. I come back to, I believe
Rudy your own comments, that the details are essential here. And any specific drug and
dose and patient population, and all the details make all the difference. And no one
program, unless you can demonstrate that you have eliminated amyloid, which we're not
yet able to do. I don't think that any negative program is going to have a major impact on
the hypothesis. In a sense that's unfortunate. And if we make a hypothesis un testable, is
it an important hypothesis?
Well eventually I hope that we will demonstrate an effective anti amyloid approach that will
even convince Dr. Davies. I know that's a high bar to set. But that's what I think will
happen. As to what number I would put on the PBT2 program? I would give it a four.
Catherine Arnold:
Dr. Davies, I want to send the next program to you. And get you comments on the
Elan/Transition collaboration if you could talk about that.
Peter Davies:
Is this scyllo inositol?
Catherine Arnold:
Yes.
Peter Davies:
Again it's not - it's an anti amyloid aggregation approach that's been tested in the
transgenic mice. And I think it's incredibly difficult to know whether those mouse studies
are predictive. But assuming that the mouse studies were predictive, again, I would say
one might get some benefit if one can get the amyloid out of the brain. But I don't think it's
going to be big. I just don't see the evidence that the amyloid accumulation in brain is
really starting the problem. Or a participant in the driving of the disease process. I think
when we come to talk about the immunization therapies that will also be evident. But it
may already be that we have some data on testing it whether amyloid removal is a useful
therapy. And I just don't think it is. So I'm still down at two. You've got to find something
better to move me up.
Catherine Arnold:
We really have to beat you up a little bit about this comment “useful,” I think, because I
know it's going to be a pervasive theme, Peter. So in terms of - you mean useful as the
silver bullet that Alzheimer's goes away? Or do you mean useful in the symptomatic vane?
And I'm having fun with you. So I appreciate you responding accordingly.
Peter Davies:
I think if one looks at the Aricept experience. It's really instructive that, you know, if we look
at the ADAS-cog scale which is basically a 70 point scale, one sees a patient move from a
score of 35 to a score of 41 on Aricept. That's about the average result. I wouldn't be
surprised if removing all the amyloid from the brain produced an improvement of that
magnitude acutely. But I don't think it will modify the disease process. I mean maybe
that's enough to sell a lot of drug. It hasn't been bad. Aricept sales haven't been terrible.
But I just don't think there's a disease modification in that process.
Whatever the process in Alzheimer's disease that leads to all this amyloid deposition,
that's going to be going on anyway. Whether you remove - whether you suck out all the
amyloid or not - the process that leads to that amyloid deposition is still going to be there.
The disease process is still going to go on. So that's why I'm not enthusiastic.
Catherine Arnold:
Dr. Aisen, could you follow up with that?
Paul Aisen:
Well I again put much more capital in to the amyloid hypothesis than Dr. Davies does.
Again while the genetic form - genetically determined Alzheimer's disease - is rare, it looks
very much like sporadic Alzheimer's disease apart from its age of onset and inheritance.
And to me that is - I think, Peter, you'd have to say that it's quite a remarkable coincidence
that every single gene that produces Alzheimer's disease finds itself acting at the cleavage
of the amyloid precursor protein to release A-beta. I still don't see a way that that can be
accounted for unless this is driving the process.
So coming back to scyllo inositol and the Elan program, I found the preclinical evidence
quite strong. I'm not discouraged by the tramiprosate failure. Very interested in seeing
the results of clinical testing. And at the risk of being boring, I would give it a four.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes I would just add that it's not about sucking amyloid out of the brain. First of all I'll just
say my number. I would give the AZD-103 probably a seven. I think the preclinical data
with PBT2 are more compelling. I studied both of them in detail. So I give PBT2 a higher
rating. Just because I think the mouse studies are more compelling. Although they're not
published yet. And others can see that soon as they've been presented.
I think it's not about sucking amyloid out. Both of these drugs the AZD-103 and PBT2,
prevent the early oligomerization of A-beta. And I think we have to think about, you know,
in Alzheimer's as with AIDS, what is our culprit? HIV and AIDS - I think that the data
suggests that A-beta oligomers, long before protofibrils and fibrils and plaques, as their
formings are monomer, are causing the problem. If there are too many of them in a
synapse. A-beta 42/40 ratio increases these oligomers and they cause synaptic
dysfunction. And, you know, many great groups have shown this now. Including probably
the leading LTP guy in the country, Roberto Malinow. It's hard to doubt it. The oligomers
do this normally. And too much oligomer impairs LTP.
So if the battle ground - there's the synapse. And if the villain is the oligomer. And both of
these drugs AZD-103 and PBT2 prevent the oligomer, that's not about letting amyloid
deposit and sucking it back out, as Peter's simplifying it. It's - again the devil's in the
details. It's preventing the formation of the villain. The oligomer itself.
And I think that over the last five years there's been really strong data to suggest there's
been a paradigm shift toward these small dimers and trimers. And they don't only occur in
genetic early onset cases. Alex Roher showed ten years ago that AD brains, sporadic or
otherwise, are full of dimer and trimer. And these are the species that are exerting this
anti-synaptic effect in hippocampal slices and in vivo.
So I'm very optimistic about these approaches that hit A-beta oligomers. And particularly
the Transition and the Prana approach.
Catherine Arnold:
So I think we were anxious to move on to immunotherapies. And I would like to spend a
lot of time on the passive programs. But before we do I'd like to get your take on the
active programs. Obviously the first one being the Elan program where the Phase II
program was shut down due to meningoencephalitis. Could we just get your - we'll just
ask you for your call, on a scale of one to ten, of active program prospects. I think that's
probably the best way to go. And then let's move into the some of the passive programs.
So if I could ask Dr. Tanzi to start off?
Rudy Tanzi:
I'll give active a one. In no way would I want to open the flood gates and have my natural
antibodies producing anti A-beta immune responses. A-beta plays a normal role in the
brain. And I think you open up those flood gates. It's very dangerous. If you can control it
with passive immunization I think there's hope. But I would not want to have an active
immunization in any of my family or friends. So one. Dangerous.
Peter Davies:
I am a vigorous opponent of the active immunization process. I think it's dangerous. I
think it's being done without any reasonable degree of thought about how the immune
system works, without drawing on the vast experience in Multiple Sclerosis, and
Myasthenia Gravis, the classic auto immune disorders. I actually think this is a very
dangerous and silly thing to do. I am a vigorous opponent and will remain so. I'm a
vigorous opponent. In case anybody didn't hear that. Minus ten.
Rudy Tanzi:
Can I go minus ten?
Paul Aisen:
You couldn't think of a number less than minus ten?
Peter Davies:
Far away from ten.
Paul Aisen:
I first want to make the point that the original vaccine trial by Elan was historic. And it was
important. I think that it really strengthened the amyloid hypothesis in some ways.
Although not in others. But demonstrated that immunotherapy can actually clear fibrillar
amyloid. Now I would agree with both of my colleagues that fibrillar amyloid does not
seem most promising as a specific target. But it was very surprising to find that the
immune system could be induced to clear fibrillar amyloid from brain. And I think the post
mortem studies do demonstrate that. I think the clinical results of the trial are equivocal. I
think that they're encouraging. But certainly not highly compelling.
The safety is clearly present. I'm not sure I would have abandoned the program with 6%
meningo-encephalitis. Considering that most of those cases resolved without sequelae.
And that considering that this is a uniformly fatal, devastating, horrible disease. I might
have continued that program. I am certainly a big advocate of active immunization. I again
- if you favor the amyloid hypothesis, I think that the active immunization against amyloid
looks very promising. Because you have greater evidence of amyloid clearance with that
approach than with any other. Now clearing amyloid from mouse brain or even human
brain could be irrelevant. The fact that the plaques were phagocytised does not mean that
you helped people. But when we put the encouraging, if somewhat equivocal, clinical
results of the active vaccine together with that amyloid clearance. I think it supports an
active vaccination anti amyloid approach.
Is it risky? I do think it is. I think that we have some idea of why people got
meningoencephalitis. And I think the current active vaccination approaches which limit the
cellular immune response by restricting the epitope in the vaccination are likely to reduce
that risk. And the experience so far with bapineuzumab suggests that there's some risks.
But perhaps not of such serious toxicity we've seen in the original vaccine. But I am a
strong advocate of active amyloid vaccination. And I would give a number to active
vaccination? I would give it a 4.2.
Rudy Tanzi:
Oh, we can do points?
Catherine Arnold:
Very precise.
Paul Aisen:
I just wanted to be a little more optimistic even than with the others. In response to Peter's
vote.
Catherine Arnold:
Dr. Aisen, why don't we move to passive immunotherapy. And obviously bapineuzumab is
what's coming most here. As far as results in hand for Phase II by mid '08, starting a
Phase III program. Contrast your view, active versus passive, and what you think of
bapineuzumab's success.
Paul Aisen:
Well I'm also very interested in passive immunotherapy. Again if you believe the amyloid
hypothesis has merit, and you see the results of active vaccination, you want to build on
those results. And that includes passive immunotherapy with monoclonals, passive
immunotherapy with pulled human immunoglobulin, and active vaccination. And I'm
pleased that all of those programs are moving forward vigorously.
Specifically with bapineuzumab, it does mimic what is likely to have been the amyloid
clearing type of antibody produced in the active vaccination. And that's in it's favor. The
signal that was seen in Phase 1 was very interesting and intriguing. It supports the idea
that Rudy put forward, that amyloid is a synaptic toxin. And tying up amyloid will have an
acute beneficial effect.
My guess is that although the numbers are so tiny. It's certainly - one can't be confident.
But my guess is there was a real systematic improvement. That is being confirmed in
Phase II. And will be shown to be real in Phase III.
So of all the anti amyloid approaches, perhaps the strongest evidence of clinical benefit
today comes from bapineuzumab. So I think there are safety issues related to vaccination.
And the angioedema episodes, which fortunately don't seem to be so serious, but are of
some concern. But also good evidence of the possibility of a favorable clinical effect. So I
would give bapineuzumab a 4.3
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes - I think the preclinical data with passive immunization have been compelling. As you
know. And I have some safety concerns different than the active. I mean the active is
quite different situation. You're making antibodies A-beta for the rest of your life. But you
know titrating in monoclonals is a safer bet. There are still some concerns about
monoclonal antibodies accumulating on blood vessel walls together with A-beta and
causing microhemorrhages.
One of the reasons that as brought up by Pfizer is their antibodies are deglycosylated with
the hope that it would reduce inflammatory and microhemorrhage\-type potential adverse
events.
So putting safety aside. Just based on what we know about - and I agree with Paul that
the active vaccine experiment, even though it led to encephalitis and their safety concerns,
was historical. And did teach us something. If we then extrapolate that onto passive
immunization and look at in combination with the preclinical data there, I'm actually pretty
optimistic about the Elan monoclonal and other companies.
I think Neurimmune from - Roger Nitsch's company that Biogen just did a deal with.
Where they're actually using monoclonals based on the ones that are really being made in
the body upon active immunization is quite clever. So I would give the whole passive
immunization area, outside of IVIg, but the specific monoclonals against A-beta a seven.
I think there's optimism there. A seven but with an asterisk. Kind of like Barry Bonds,
because of safety issues. Maybe not as bad as steroids, but safety issues.
Catherine Arnold:
Patterns emerge. I was just looking back at the scores and, Dr. Aisen's highest score is a
4.3. Dr. Tanzi, you do have nine in there. I'm doing a relative check here as we go
through. Dr. Davies?
Peter Davies:
My highest score here would be a three or a four for passive immunization. And there's
really a fundamental difference between passive immunization and active immunization.
There are safety concerns with passive immunization. But if something starts to go wrong,
you can stop. With active immunization once you start the process, there is no stopping it.
I was immunized against Polio when I was 5. It's still effective. There isn't no way to turn
that off.
Passive immunization offers -- at least if the preclinical data is right -- a way to remove -
quite effective way to remove amyloid from the brain. I don't think the safety issues will be
anything like as bad as the active immunization. And it's a very intriguing way to test the
hypothesis. I think it's quite possible that one can disrupt oliogomer formation and
certainly reduce amyloid concentrations in brain.
This way the preclinical data is very compelling in that regard. And this is a good way to
test the amyloid hypothesis. I'm in favor of the trials. A three or a four. 3.5. How about
that?
Catherine Arnold:
Just for the record, that's your highest score as well.
Peter Davies:
Yes, that's my highest score so far. Okay.
Catherine Arnold:
Now could I ask just the group to comment on whether nor not your score would change
for a different location of epitope in terms of… Obviously the bapineuzumab program is at
the N terminus. You've got Pfizer's program is at C terminus, and Lilly is at mid region.
Would your number change based on where the activity is targeted?
Rudy Tanzi:
You mean for active immunization or passive?
Catherine Arnold:
I'm talking about passive.
Peter Davies:
No I don't think there's enough information to favor one over the other. And they all should
hit A-beta 40 and 42. So I see no advantage to one epitope over the other.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
I just like the Neuroimmune approach. Because regardless of epitope, they're taking
antibodies that were made in humans as part of active immunization. Saying okay what
are the anti oligomeric A-beta antibodies that are made when you immunize? And then
they're mimicking those in a reverse translation approach. So outside of epitope, I'd give a
little higher rank to that. But in terms of just choosing epitopes, I wouldn't change my
score.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
Well I think our big differences is with different epitopes. I’m just not sure which one is
going to be best. I think that an N terminus antibody is more like to induce the
phagocytosis in brain. Particularly if it's an intact antibody. I think that a mid sequence and
C terminus antibody are much less likely to produce clearance of fibrillar amyloid. I wish I
knew if that was an important target. But it's a - I think a significant difference in approach.
I think if the sequestration idea is prevails, as it well might, then we might have efficacy
with more safety with a non N terminus approach so that would encompass the Lilly
approach and the Pfizer approach. If in fact we want fibrillar clearance, we're not going to
want to deglycosylated the antibody. But maybe we don't need that. And so maybe the
Pfizer approach will prove to be safe and effective.
I'm glad that we're pursuing all of them. Until we beat this problem I think we should be
testing all promising avenues. And my scores - maybe you can guess this but - my scores
for each of those would be in the 4 to 4.3 range.
Catherine Arnold:
We - just for housekeeping purposes we are running a little bit behind. We have - we do
have another about 30, 35 minutes left before absolutely will lose our panelists. So I do
want to get through a couple more programs. And then open it up for questions. Another
passive approach that's a little less selective but obviously talked about quite a bit is IVIg.
I wonder, Dr. Aisen we'll start perhaps with you. What’s your - how do you think about that
program. That's the Baxter program.
Paul Aisen:
Well I am also pleased that this program is going forward. IVIg is pulled immunoglobulin. It
turns out that there are naturally occurring anti amyloid antibodies. Those antibodies are
relatively reduced in people with AD. You can replenish natural amyloid antibodies by
using infusions of IVIg. And biomarker evidence from small studies suggest that you can
have - you can influence the CSF A-beta and plasma A-beta levels by infusing IVIg. And I
think all that's encouraging.
But I'm particularly encouraged by the Phase 2 results. And the Phase 2 results have been
announced rather than fully presented. The announcement is just that there was a very
encouraging signal. So there is not a lot that we can say about that. But Norm Relkin,
who has developed this idea and has presented a number of small studies showing a
rapid and sustained improvement in people with AD with IVIg, has demonstrated that
although the titers of anti monomeric amyloid antibodies are quite low in IVIg, there are
high titers of anti oligomeric amyloid antibodies. And if the oligomers are the most toxic,
that might explain the clinical results he has reported. And it might increase the likelihood
that this approach will come to fruition and will be effective. And of course there's the
advantage of all of the history of IVIg so we know a lot about it. And it's available now. So
I'm pleased that there is a Phase 3 trial of IVIg starting up now. And I would probably give
that a 4.3.
Catherine Arnold:
Okay, Dr. Davies?
Peter Davies:
It's an intriguing approach. And I'm around a three I guess. Having seen the recent
presentation just a couple of months ago from Norm Relkin about the clinical benefit, I
must say I was not impressed that much was going on. A few patients seemed to do quite
nicely, but the aggregate effect did not seem terribly striking. It's an incredibly expensive
approach that is going to be I think limited at best in its availability. And, you know, it's
also one of those very troublesome things where it - if there is a small benefit, I don't think
we're going to know why.
IVIgG is used to treat a lot of conditions. And we have a great deal of experience with it. I
have a great deal of experience with it. We took my grandson who was treated for
Kawasaki disease with IVIgG. And now nobody really knows what the etiology or
pathophysiology of Kawasaki's disease is. But essentially 36 hours of infusion of IVIgG is
a complete cure. As it was in my grandson's case. So why does it work? Well we don't
know. What's the etiology of Kawasaki's disease? We don't know.
If we see a benefit in some Alzheimer's patients then I suspect it will be a small number of
patients. I don't think we're going to know why that happened. So it's kind of, from a basic
science perspective, a very frustrating approach.
Although since it's safe and relatively easy, let's test it. Hence my three.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes, I agree with Peter on this. I would give it also a three. I think that, you know, we
actually published a paper last year where we characterized the physiological auto
antibodies to A-beta in patients and controls. And in full agreement with what Dr. Aisen
pointed out. You see antibodies to oligomeric A-beta that are auto antibodies. And you
also see antibodies to monomeric A-beta. But in the patients versus controls, the only
decrease in auto antibodies was in the ones targeted against oligomeric A-beta. We saw
no difference in the levels and titers of auto antibodies of monomeric A-beta in patients
versus controls. There was reduction in auto antibodies through oligomeric A-beta. And
that reduction correlated with age of onset. So that's the pool you need to care about.
And it's possible that with the IVIg you're non selectively introducing and replenishing
antibodies against oligomeric A-beta in patients. But if you do that, why not just do it
directly? I mean that's why I'm not as enthusiastic about this. You know, make antibodies
to oligomeric A-beta the way Neurimmune is. Or other companies that are targeting
specific forms. Dig in and just do it selectively. I don't see why we have to do it in this
quick and dirty way. So I would give it around a - I guess around a four. I'll go above
Peter just to be different.
Catherine Arnold:
Dr. Tanzi since we're on you could we switch gears? I want talk about gamma secretase
modulators and inhibitors. And obviously the Myriad program is also a late stage program.
What are your thoughts on the promise of potential for Flurizan, and the risk. And what's
your score?
Rudy Tanzi:
Yes, well I'm very positive. I'm the most positive about gamma secretase modulators.
Because, you know, what we've learned that the 42/40 ratio is what matters. And that's
what genetics has taught us. And if you can reverse that ratio genetics predicts you'll treat
the disease effectively. Three different genes tell you that. We see more whether it's
genetic or not. You see more 42 in AD brain. It's the principal peptide driving aggregation.
And it's the principal peptide in plaque.
So I love gamma secretase modulators. But I think Flurizan was first out, but just not
potent enough. The Phase II trial showed no significant results. But except for this post
hoc assessment. There was a sub group in milder AD who took the higher dose. That
800 mg dose that had some improvement, and ADAS-cog during the open label extension,
12 to 18 month open label extension. So Flurizan is not a particularly potent gamma
secretase modulator. It's based on NSAIDs. Its bioavailability isn't optimal. But you know
it has a shot. I would say it has a 50/50 shot to make it out of the block. Because there
was some benefit with the high dose. If their Phase III results that we'll hear about in mid
'08 are positive, you know, they may be okay.
I'm much more optimistic about companies like Eisai, Torrey Pine Therapeutics, and
Merck who are going at it directly. Trying to specifically find new chemical entities that act
as gamma secretase modulators.
Some of the early reports of these compounds are that they are 1000 to 5000 times more
potent than Flurizan in reducing or reversing that ratio of 42/40. Eisai got off the block with
a Phase I that was stopped because of some tox issues. But there's lots of activity in this
area.
So I would give Flurizan a - probably a three. The prospect for future gamma secretase
modulators, a very high rating, eight or nine.
Peter Davies:
I'm sorry the potency in terms of Flurizan is just not there. If you go back to the basic
science literature on Flurizan, the doses that are being given to people are not going to
produce concentrations in brain that are adequate to modulate gamma secretase
(unintelligible). I don't think you could load people with enough Flurizan to have significant
effect on gamma secretase activity. The whole area for me - I'll go up to a five with the
gamma secretase program.
This to me is kind of the Amyloid Cascade Hypothesis putting its money where its mouth is.
You know, if you really believe that amyloid is the critical issue in Alzheimer's disease,
then block its production. You know, do the experiment the simple and direct way with the
gamma secretase inhibitors. And I've very enthusiastic about seeing these compounds go
forward. And I think it will take more than one. Paul was saying, the devil's in the details.
But somebody should come up with a good gamma secretase modulator. And we'll find
out whether the amyloid hypothesis is correct or not. I think this is where the testing is
really going to be vital to improving our understanding of Alzheimer's disease.
If these compounds really work, I'll eat my hat. And say I was wrong. But, you know, we
need to test these compounds to find out if this hypothesis is correct or not. I'm very
enthusiastic about the clinical testing.
Catherine Arnold:
I'm anxious for your number.
Peter Davies:
Five. There you go.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
I think we come closest to agreement among all three of us on Flurizan. I also think that
it's an interesting approach. And that this particular compound may well be not potent
enough, at least in the brain, to get the effect we want. I'm discouraged by the absence of
any biomarker evidence of a pharmacodynamic effect in humans. And I would agree with
Rudy's characterization of the Phase II result. Nonetheless we've got these two huge
pivotal trials. So we'll get a real answer before too long. I would give Flurizan itself a 3.8.
For the whole gamma secretase modulation program, well, I think there are some
theoretical reasons why it might be terrific. But practically speaking I'm not quite so
optimistic. I would give the whole program about a 4.1.
As to the test of the Amyloid Hypothesis, though, I would almost dare I say it? Almost
agree with Peter. But it's not a gamma secretase inhibitor in my view that's going to test
the hypothesis. It's going to be an effective beta secretase inhibitor. When we get an
effective beta secretase inhibitor that really eliminates - that is at least 75% reduction or
more, in the generation of the A-beta peptide. And I believe we're going to get there
before too long. I think we'll be in a position to test the Amyloid Hypothesis.
I don't think a gamma secretase inhibitor will do it. Because there are too many problems
with gamma secretase inhibition itself. In terms of a narrow therapeutic window. You can't
lower it enough without toxicity. I'm not yet convinced that modulation is going to do it. So
I would say the hypothesis will be tested with a beta secretase inhibitor.
Catherine Arnold:
I appreciate your patience. I know we're - we have so much to talk about. I'm just going to
bring up one more category. And then I want you all to sort of identify what's most
promising. And let's get some questions from the audience. Gamma secretase, Peter, if
you could talk about the hypothesis of potential positive and risks. And give us your
number on gamma secretase. I'm talking about the inhibitors, not the modulators now.
Lilly has a program in Phase III where they're among many other pharma companies.
Peter Davies:
Yes, again, I would not draw big distinctions between gamma secretase modulators and
gamma secretase inhibitors. I'm not - the data from the basic science is so complicated.
That I don't think it's really that easy to lump the compounds into two separate categories.
I am enthusiastic about the gamma secretase inhibitors. And about the beta secretase
inhibitors as tests of the hypothesis.
Beta secretase - I put them in the same category. I agree there may be some more
problems with gamma secretase inhibitors than with beta secretase inhibitors. I don't think
there's a huge difference there. And I think clever chemistry can get around a lot of the
potential side effects with both of those. I think these are the class of compounds that we
really - the field really needs to test. And I'm very encouraged that we are getting into
Phase III studies with these compounds.
We've got - the Amyloid Cascade Hypothesis is going to be tested in the clinic by these
kinds of compounds.
So I stick with my four or five - 4.5.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
Well I'm a little more optimistic about gamma secretase inhibitors than modulators.
Because we can really track the pharmacodynamic effect very clearly. And what most
impresses me about the Lilly secretase inhibitors program is the human pharmacodynamic
effect. There is the clear inhibition of A-beta generation as indicated by A-beta
measurement in blood in response to this particular gamma secretase inhibitor. If we were
to see that kind of clear effect with the modulator without toxicity, great. But the best
inhibition of A-beta generation that I've seen in humans is with the gamma secretase
inhibitor so far. So that I think is encouraging.
What's discouraging, again, is the fact we know that you have very little room to move.
And too much gamma secretase inhibition is going to give you serious gastrointestinal
toxicity and immune problems. And whether you can thread that needle, I don't know. So
I would probably still be in the same range for the Lilly program. I'd probably be about 4.1.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes, just simply based on safety I'm very negative on gamma secretase inhibitors. As
positive as I am on gamma secretase modulators, I do think there's a huge difference.
With gamma inhibitors, you're hitting notch. You're hitting two dozen other substrates.
And the recent findings that have been published now by (Washy Shu) and Phil Wong and
others. It's also oncogenic. You block gamma secretase and you cause cancer. So I
think this is a doomed technology to inhibit. It's a very different to modulate. It's a different
mechanism of action, different type of molecule, different program. So I would give
gamma secretase inhibitors themselves a one or a two. Maybe - I would say a one. Just
because I think they're too dangerous. You can't just hit gamma secretase with a sledge
hammer. Even if you try to take tiny little whacks. It's too dangerous. But gamma
modulators are safe.
Catherine Arnold:
Dr. Tanzi, now the companies, they all talk about this issue.
But some have more
confidence than others that they've gotten around it. Would you think that's warranted as
you look across the program?
Rudy Tanzi:
I think that eventually, with people who inevitably have different responses to drugs, that
there will be those who will have serious adverse effects - adverse reactions to pure
gamma secretase inhibitors. There are too many substrates. It's way too important and
essential an enzyme. It's removing the C terminal stubs of dozens of proteins after their
extra cellular domains are shed. I don't think you fool around with that. That's very, very
different than allowing gamma secretase activity to occur as with modulators. Simply
trying to target the production of 42 specifically on APP. And there's been great progress
in trying to achieve that selectivity. I'd rather wait for that.
Catherine Arnold:
What did we not talk about that you are most enthusiastic about? Obviously we didn't talk
about beta secretase. And if there's a program there, and there's some mechanisms as
well. So would you each just tell us what you - what you're enthusiastic about that we did
not speak about? Dr. Tanzi?
Rudy Tanzi:
Yes I mean beta secretase inhibitors have potentially greater safety than gamma because
there are fewer substrates. But beta secretase has its substrates, too. (Nuregulin) which
is essential for myelin formation is a beta secretase substrate. The sodium channel, beta
2 subunit which controls sodium channel activity by modulating the alpha channel
expression is substrate. And there are a couple of other substrates. So beta, you know,
many say, well beta secretase is clean because you don't have the substrates like gamma.
You have some pretty important substrates for beta secretase. So it's a safer bet. It's a
nice target. But I think you might also have issues there.
There are a handful of companies - I think Wyeth has at least one small molecule inhibitor
- and a few others. It's been an incredibly tough target to get at. So I think there's going to
be real challenges trying to get an effective BASE inhibitor. But even once we get one,
we're not out of the woods in terms of safety. There's a growing list of essential targets for
beta secretase as well. So I'll put them higher than gamma inhibitors. But I would still only
be at about a four or a five. A 4.5 let's say, for beta secretase inhibitors. Because of
safety issues and the rapidly emerging list of substrates that we didn't know about before.
Paul Aisen:
So as I said before I would put beta secretase inhibitors at the top of my list. I think that,
yes, they are in the body for a purpose. But as far as we can tell, that purpose is current
development and not adult life. The fact you have in the beta secretase (unintelligible) in
mouse. It took a long time to find anything wrong with those mice. I believe this is going to
be quite a safe target. I think there are a number of promising compounds moving forward.
And I believe that in the next couple of years there will be small molecule, a brain
penetrating, highly potent, and safe beta secretase inhibitors of moving into the clinic. So
that would certainly be theoretically, and I hope without too much optimism, at the top of
my list. And that - and I would put beta secretase inhibitors at five or so.
Practically speaking I just wanted to mention one other program. On theoretical grounds,
BASE inhibition is best in my view. But on practical grounds - that is empirical grounds -
looking at actual clinical results, nothing can touch the Medivation result with Dimebon. I
think it was an excellent trial in Russia. I think that the data is simply powerful. And I
believe that you will see replication and proof in subsequent trails. I don't know how that
drug works. Theoretically I have no idea. But as far as empirical data, that's the best that
I've seen.
Catherine Arnold:
Dr. Davies?
Peter Davies:
We all as scientists have to be humble and recognize that many drugs are discovered by
accident. And there's a 1 in 1000 shot with almost any compound you pull off the shelf.
And I - just by the odds - I would give Dimebon in the improving a successful therapy for
Alzheimer's disease.
I think Paul and I agree, though. That our scores are all below fives and below. Because
we don't think we're quite there yet. In terms of understanding the mechanisms of the
disease. At least in terms of what's in Phase I and Phase II right now. None of these
compounds make it above five in terms of enthusiasm. I think we've got more work to do.
You know, I think the understanding the pathways of the disease. And developing
scientifically rational targets based on that knowledge of the disease pathways, is what we
need to do. That's what - where the action needs to be. That's why nothing gets above a
five in my book.
Catherine Arnold:
At last we have a chance to take some questions from the audience here at Credit Suisse.
To see if you all have any questions. If you do, there are two microphones. If you could
step up to the microphone and as a question.
Audience Member:
Bapineuzumab, do we know what human dose concentrations are achieved versus the
mouse model, first one. Secondly given the transgenic mice you, get the plaque formation
but not neuro death. How is that explained? And then, using your mice where you have
oncogenic activation, could you use those mice to test all these drugs? To see if then, you
know, you actually get the alleviation of the plaques and (unintelligible)?
Peter Davies:
I think the mice are terribly difficult. The predictive value of the transgenic mice is very
poor. Because the mechanism - you drive amyloid deposition by over expression of the
amyloid precursor protein. That's not the mechanism that happens in the human disease.
And that's the whole key I think. It's an artificially generated amyloid deposit that doesn't
do very much to the brain of the mouse. So it's very difficult to use these (unintelligible).
You can use them to show that the antibodies, for example, get in and or clear the amyloid
deposition in some way. The mechanism isn't clear. Can we use the mice that we have?
We're already into second generation mice here. So hopefully we'll define the pathways
for it.
Obviously I have my favorite target that I'm not going to talk about in public. It's not public
information. So yes, one - you know, I really think if we tested an anti-amyloid drug in our
mice, we would see small improvements in our mice. Because it's not good to have all this
junk deposited in your brain. But I don't think you're ever going to see more with the
amyloid therapies. And that's, you know, we could talk for two hours about - why the
Amyloid Hypothesis doesn't work. And it's not just my opinion.
Audience Member:
(Unintelligible).
Peter Davies:
No. That's not true. The amyloid - in the mice, as in the human brain. The association
between the amyloid deposit and cell death is very foreign. So the amyloid's over here,
the cell death's over here. I don't think the amyloid deposition has anything to do with cell
death.
Audience Member:
(Unintelligible).
Peter Davies:
Yes, but in the mice in general. I mean, you can remove the amyloid and you don't prevent
cell death. Because there isn't any cell death to begin with.
Rudy Tanzi:
Sure, because it's not the plaques.
Paul Aisen:
Just going along with your question. There is a triple transgenic mouse model that
includes both amyloid deposition and tau hyperphosphorylation. It is also not a perfect
model and it doesn't have much neurodegeneration. But it does have markers of change
that are somewhat reminiscent of tau pathology. And, in that mouse model, if you give an
antibody, you reduce not only the amyloid accumulation, but also the tau pathology. It's
not a complete answer. It's - I'm sure Peter's going to say “Well yes, but you don't have
cell death” which you don't. None of these models are great. But that's been used as
evidence for a connection of the cascade, of the amyloid to the tau cascade.
Catherine Arnold:
Dr. Tanzi, you were going to add something?
Rudy Tanzi:
Well I was just saying in response to Peter. That it's not about the plaques or the deposits.
I mean that's where we were led astray for years. It's about dimers, trimers and soluble
oligomers. These are doing the damage. And removing plaques doesn't necessarily
remove the oligomers. You really need to target what drives oligomerization, which is
the A-beta 42/40 ratio. Or try to prevent those oligomers from forming or dissolve them.
So that's a very limited number of therapeutic targets. Hit 42/40 or try to prevent the
oligomers. But in terms of hitting plaques, it could be even dangerous to dissolve a plaque
because you might turn it back into oligomers.
So I think we have to get a way from this idea that the amyloid therapy - the success of
amyloid therapy depends on whether plaques matter. I think the overwhelming evidence
is that we have to move upstream of the plaques toward - and even upstream of
protofibrills - toward soluble oligomers. And we haven't discussed in detail those data
from excellent groups. But it's overwhelming at this point.
Catherine Arnold:
And I just want to throw something out that I've discussed with each of you. I think that it's
an important point. Is that if you step back and some of these actually are effective. And
you actually know the patients at risk earlier, then the debate may be different. If I think I
understood some past conversations that I've had at least with maybe two or three of you.
As we think about what happening, what's changing is that you are testing the patient with
the plaques already developed. But if you're doing PET scans early on and identifying
patients at risk, some of what we're talking about may be a different conclusion. I - tell me
if that's a wrong statement or…
Paul Aisen:
Well I would certainly say that we are headed towards identifying the genetic carriers of
PS1, PS2, APP mutation carriers, Down’s Syndrome individuals, as well as people at high
risk for sporadic AD on the basis of amyloid accumulation before any symptoms. And I
think all of these groups are going to end up being effectively treated with anti-amyloid
compounds.
Rudy Tanzi:
is that recent data from Christian Haas show that gamma
secretase modulators - the NSAID based gamma secretase modulators - cannot treat
patients with presenilin mutations. Presenilin mutations lock up the active side of gamma
secretase into such a rigid confirmation that NSAIDs, including Flurizan, cannot unlock it.
So there's been a lot of talk of lets use early onset mutation patients in trials. You got to
be careful there because now we know that at least NSAID based gamma secretase
modulators can do very little against most presenilin mutations that lock up the active site.
But again it's all about the details. I mean, all of these therapies and patient selection for
clinical trials. You have to look at mechanism of action very carefully before moving
forward.
Peter Davies:
And disease progression, it's going to be incredibly important to accurately diagnose
patients early. Because you don't want to slow or stop disease progression in a patient
who's already well down the hill. You're really then prolonging the agony. When we do
develop effective disease modification, we're going to have to put huge emphasis on early
and accurate diagnosis so we can treat people really in the MCI stage. Rather than after
they already have a diagnosis of moderate Alzheimer's disease.
Catherine Arnold:
We have time for one more question. I appreciate the warrior like attendance here.
Operator, are there any questions on the line?
Operator:
There are no questions at this time.
Catherine Arnold:
Okay, well we are really butting up against 12 o'clock. This is a very dense and important
topic. I guess as I step back… If I could try to summarize, and you'll tell me if you disagree.
I know that.
That we have a lot to learn about the disease. We're probably more likely to make baby
steps in even some of what we learn. The relevance may be different based upon the
continuum of the disease. And all of that is encouraging. But we still have to wait and see.
Next year will bring us some really interesting insights. I think you both - you three -
sounded optimistic on beta amyloid but not, we ought - the devil will be in the details. I
think we've said that a couple of times.
Any last words from my three panelists?
Peter Davies:
Thank you.
Rudy Tanzi:
One thing we didn't mention is alpha secretase enhancers. Trying to preclude A-beta
production by enhancing alpha secretase. And there you have the nicotinic and muscular
agonists. I think they shouldn't be discounted in terms of keeping track of future trials as
well.
Paul Aisen:
I wanted to make the point that while we gave scores from one to ten, I was thinking in
terms of percentage likelihood to succeed. The last point I would make is while we don't
think that anything is more than even odds, I think that all of them together are extremely
likely to produce at least one winner. So I'm optimistic because of the breadth of the field.
Catherine Arnold:
Well and I have to add that, Dr. Aisen, each of you in individual calls before today,
mentioned the promise of combination therapy amongst even a lot of these mechanisms
that we don't have yet, could be really exciting. Dr. Tanzi, we talked about that yesterday,
you had a very strong point of view there.
Rudy Tanzi:
Yes, I mean my own guess is that since - I guess it's obvious now - I believe that the Abeta
oligomers are to Alzheimer's as what HIV is to AIDS. I think based on the newest
data, that if you hit the production of oligomers with a gamma modulator and you try to
stop them from forming or dissolving them with one of the anti-oligomer strategies. That
that combination will be very powerful. Perhaps together with more powerful symptomatic
treatments. And there I think keeping track of nicotinic agonists, and muscularinic
agonists, which may also have some effect on APP. That might be a third component. So
I doubt any one therapy is going to take care of it. But my hope is that if you hit A-beta
toxicity at the synapse, that hopefully downstream events will also be averted.
Catherine Arnold:
Well I would like to thank our panelists and all of our attendees. I think it's been a really
provocative and fun discussion. So have a good afternoon and a nice weekend.
Companies Mentioned (Price as of 02 Jan 08)
AstraZeneca (AZN.L, 2175.00 p, NEUTRAL, TP 2375.00 p, MARKET WEIGHT)
Baxter International (BAX, $58.05, NEUTRAL, TP $63.00, MARKET WEIGHT)
Eisai (4523, ¥4,400, NEUTRAL, TP ¥4,800, MARKET WEIGHT)
Elan (ELN, $22.09)
Eli Lilly (LLY, $52.55, OUTPERFORM, TP $65.00, OVERWEIGHT)
Forest Laboratories Inc. (FRX, $37.16, OUTPERFORM, TP $60.00, OVERWEIGHT)
Johnson & Johnson (JNJ, $65.91, NEUTRAL, TP $65.00, OVERWEIGHT)
Medivation (mdvn, $14.47)
Merck & Co. (MRK, $57.37, NEUTRAL, TP $57.00, OVERWEIGHT)
Myriad Genetics Inc. (MYGN, $46.14)
Neurochem Inc. (NRMX, $2.47)
Novartis (NOVN.VX, SFr62.10, UNDERPERFORM, TP SFr64.00, MARKET WEIGHT)
Pfizer (PFE, $22.91, NEUTRAL, TP $26.00, OVERWEIGHT)
Prana Biotechnology Ltd. (PRAN, $4.12)
Transition Therapeutics Inc. (TTHI, $11.25)
Wyeth (WYE, $43.89, OUTPERFORM, TP $54.00, OVERWEIGHT)
This is a 32-page Report from Credit Suisse discussing all of the approaches, and even numerically handicapping the various clinical trials, by a Panel of 3 experts. One thing I have learned is that you can't always trust those knock-out mice.
As this may be the longest post ever placed on iHub, and the formatting may not work out, you may prefer to email me at my yahoo address (not hard to work out), put ALZ Panel in the subject line (to distinguish from the spam) and I will send the PDF in response.
Date: January '08
Alzheimer's Disease - Relative Optimism from Diversified Experts
*
Three weeks ago we assembled a panel of cutting-edge researchers in
Alzheimer's Disease (AD). Our purpose was to provide background on
disease pathology and identify interests in new therapeutic prospects from
the mid to late stage pipeline. This group had relative optimism on a number
of new mechanisms and saw several as potentially playing a lead role in
future therapy (e.g. WYE/ELN's bapineuzumab), but did not see any one
option as the means for eradicating the disease. The transcript from the
panel is included in this note.
*
The amyloid cascade is the point of attack for all late stage therapies. Two
panelists were optimistic that this cascade is critical to the disease course so
key drugs taking this approach should be meaningful in some way, implying
commercial promise. Even the less positive panelist felt this approach may
yield acute cognitive effects that would be embraced by the AD community.
*
Our panelists believed that therapies that reach the marketplace in the next
3-5 years should provide incremental benefits. Given the prevalence and
severity of AD and current lack of disease modifying drugs, therapies that
provide even incremental benefits (along with reasonable safety) could be
potential billion dollar blockbusters. Using these agents in combination with
each other could help to truly modify the course of the disease.
*
Our panelists expressed the greatest optimism for a passive immunotherapy
approach with antibodies directly targeting amyloid beta (Aβ). The leaders
in this approach are WYE/ELN’s bapineuzumab (recently began phase III),
LLY’s m266 (recently finished pII dosing) and PFE’s 4360366 (in pII).
*
Our current valuation of WYE/ELN's bapineuzumab suggests a $2.5Bn
opportunity (using a 50% probability of success adjustment along with 20%
market penetration and $25,000 price/pt estimates). We see this as
conservative and our panelists directionally agreed that there is at least a
50% chance that bapineuzumab (and other passive immunotherapies)
proves to be an advance in the treatment of AD. Current WYE share price
provides a free call option on bapineuzumab, with pII data in mid-late ‘08
likely to be positive and, if so, providing more positive support to the stock.
*
Our panelists were also optimistic about the potential for gamma secretase
modulators and other drugs that alter the Aβ 42/40 ratio, but felt that Myriad
Genetic’s Flurizan may not be potent enough to have a significant benefit.
Each panelist expressed some concern about the safety of therapies that
completely inhibit enzymes involved in generating Aβ (e.g. gamma and beta
secretase inhibitors) because these enzymes have other important functions
in the body, but they anxiously await more data.
Transcript from our Expert Panel Discussion
Catherine Arnold:
We're going to go over the neuropathophysiology of the disease. We're going to talk about
the amyloid hypothesis. And we're going to talk about the tau hypothesis to set the stage
for the balance of our time together which will be looking at the drug pipeline and after that
we'll go into full Q and A.
Now as far as background, I would bet everyone in this room has been affected by
Alzheimer's disease in one way or another. With that experience you're undoubtedly
aware that no approved options are available that which actually modify the course of the
disease. The roughly three billion in drug sales in this area come only from symptomatic
treatments. With an aging population, the Alzheimer's market will grow faster over the next
several decades. And estimates now are about 26 people million worldwide and growing
to about 100 million people by the year 2050. Couple this with the fact that we've got
several drugs in late stage development, which hope to actually modify the disease and
we see this as a critical investment theme for 2008 forward.
I'd like to introduce you more specifically to our panelists. I want to comment, though,
that we really worked to put together a panel of individuals with very different perspectives.
You'll hear some very strong positives and strong criticisms on a variety of topics, but we
will work through them to try to reach closure and see if we can identify a consensus.
Now immediately beside me for those of you here is Dr. Paul Aisen, who until November
was running the Memory Disorders program at Georgetown University. For those of you
who don't know, he's transitioning. He's going to be a professor in the Department of
Neurosciences at the University of California San Diego. He has a wide breadth of
experience. He's studied almost every drug that has been studied - symptomatic and
disease modifying. He should have a very broad view for us. He's collaborated and
designed a number of large NIH funded trials. He's worked with the pharmaceutical
industry.
I'd next like to introduce you to Peter Davies. Dr. Davies is a Professor of Pathology and
Neuroscience at the Albert Einstein College of Medicine. He's a chair of the Judith and
Burton Resnick on Alzheimer's disease and the Scientific Director of the Litwin-Zucker
Center for Research on Alzheimer's disease. Now he's been working in the field for
obviously a very long time. His bio says he published his first paper on Alzheimer's
disease 31 years ago. And that led to the development of Aricept and Exelon. A lot of
colleagues think of him as being an expert on tau. Although as I've gotten to know him I've
come to appreciate that he's actually very involved in other aspects of the neuropathology.
So I think you'll find him to be very insightful.
Our last, but not least, panelist is Dr. Rudy Tanzi. He's currently a Professor of Neurology
and Neuroscience at The Harvard University. He's the Director of Genetics and Aging
Research, at the Mass General Institute for Neurodegenerative diseases at Mass General
Hospital. Going back to 1987, Dr. Tanzi focused his studies on Alzheimer's disease, and
isolated the APP, which we associate with Familial Alzheimer's disease. In 1995 he was
with the identification of presenilin 2, which is obviously important to the process. He
has also been involved intimately in the discovery that the metals zinc and copper are
necessary for the formation of neurotoxic assemblies. He's been very involved obviously
therefore in the amyloid hypothesis and we're really pleased to have him here.
I've asked Dr. Aisen to start us off with a high level overview of Alzheimer's disease
neuropathology. And talk about current treatment options.
Paul Aisen:
Thanks very much.
Okay so I can summarize current treatment of Alzheimer's disease on this slide. Dr.
Alzheimer described the disease just over 100 years ago. And then we did little about it for
3/4 of a century. Dr. Davies and others in the late '70’s developed the cholinergic
hypothesis that first suggested the possibilities for treatment. This led to the first small
positive study in 1985. At that time the pharmaceutical industry awoke to the possibility
that AD was not a degenerative disease about which nothing could be done, but rather a
specific disease that could be effectively treated. This led to the approval of tacrine eight
years later and three other cholinesterase inhibitors. Finally memantine was improved in
2003. So a total of five drugs have been approved in this country for AD treatment.
Okay so I want to start by just saying for two minutes a little bit more about what our
standard therapy for AD is today. It's all focused on the synapse. On neurotransmitters
that allow communication from one brain cell to another. The most important
neurotransmitter for cognition is acetylcholine. Each of the first four drugs approved act on
the cholinergic pathways - specifically they inhibit acetylcholinesterase and thereby
increase the activity of acetylcholine in the synapse. So there's a cholinergic deficit in
Alzheimer's disease. That deficit can be partially compensated by reducing the
breakdown and therefore increasing the activity of the acetylcholine that's left in the
synapse.
Of the four cholinesterase inhibitors - again they are tacrine, donepezil which is Aricept,
rivastigmine which is Exelon, and galantamine which is Razadyne - they are all approved
for the treatment of mild to moderate Alzheimer's disease. Donepezil has recently
extended its indication to severe Alzheimer's disease. Rivastigmine has recently extended
its indication to cognitive impairment in the studying of Parkinson's disease.
Memantine is the first drug of a different class for the treatment of Alzheimer's disease. It
is also acting as a synapse but not directly on the cholinergic pathway. Rather,
memantine is acting on the glutamate pathway - a specific glutamate receptor called the
NMDA receptor. Memantine is a partial NMDA receptor antagonist. It seems to cause
some hyperactivity at glutamate's interaction with the NMDA receptor, and the net result is
improved cognition. Based on its mechanism, there was hope that it might reduce
neurotoxicity. That is cell damage. There is not clinical evidence that there's any disease
modifying effect with memantine. But clinical evidence suggests that its action on the
glutamate pathway augments cholinergic activity and provides a symptomatic boost
generally similar to what we see with cholinesterase inhibitors. But the two work well
together. So a cholinesterase inhibitor plus memantine provides an additive symptomatic
medicine.
So this slide summarizes where we are today at the end of 2007. Standard therapy for
Alzheimer's disease is symptomatic by and large, including a cholinesterase inhibitor. One,
but not more than one, of the drugs I listed. Memantine is particularly effective at
moderate to severe stages of Alzheimer's disease. It's only approved for moderate to
severe. The results in mild disease show a very small - an effect, but a very small effect.
Practice varies from one clinician to another, but a standard evidence based approach
would be to start a cholinesterase inhibitor as soon as possible after diagnosis. And then
add memantine to it at the moderate stage of Alzheimer's disease.
I put Vitamin E here as well. We used to use Vitamin E to treat everybody with
Alzheimer's disease on the basis of oxidative stress. In one positive treatment trial, that
component of the regimen has been reduced recently as we've learned about some of the
long toxicity that may be associated with Vitamin E. So Vitamin E is really no longer a
standard part of the treatment regime. Rather it's a cholinesterase inhibitor with the
addition of memantine later on.
And I would just point out that we have a huge problem in the area of Mild Cognitive
Impairment, which is the symptomatic prodrome to Alzheimer's disease. And we don't yet
have any effective therapy for the millions of people who have MCI - Mild Cognitive
Impairment.
Now just in the last few minutes I want to set the stage for the discussion of drugs in
development today. I don't think we have any disease modifying therapy for AD. But I do
believe that we're getting close. And that's what we'll be talking about. Disease modifying
therapy is not focused on the synapse. It's focused on the neuropathological lesions in the
hope of changing the course of the neurobiology of Alzheimer's disease.
So Alzheimer described two primary lesions, the plaques and the tangles. The plaques
are extra cellular deposits. We now know that they're made up of aggregated A-beta
peptide. So it's a fragment from a larger protein called APP or amyloid precursor protein.
The fragment aggregates into plaques that are very dense and sticky. And they deposit in
the brain substance, and in the walls of the blood vessels.
The second hallmark that Alzheimer described was the tangles. Tangles are also
insoluble deposits of protein. But they occur within the nerve cell body, rather than in the
vessel wall and the brain - extracellular brain tissue. And we know that the tangles are
composed of an abnormal form of a normal protein. The protein is tau. In the tangle tau
has been chemically changed, hyper-phosphorylated. Too many phosphate groups added.
The result is that the tau changes confirmation, changes shape, loses its function and
aggregates into the neurofibrillary tangle.
These processes are summarized in this cartoon that was developed by the Alzheimer's
Association. On the right side you see the healthy components with a healthy neuron up
on top, so a healthy neuron and here is a neuronal cell membrane. Within the neuronal
cell membrane there's a normal protein called the amyloid precursor protein. There are
enzymes in the cell membrane including gamma secretase and beta secretase and alpha
secretase. You see here drawings of healthy microtubules. Microtubules are the tubes
that are essential for transport of materials down long neuronal processes. The
mitochondrion is here, which is the energy producing organelle in the cell. So this is the
healthy cell.
In Alzheimer's disease we know that we get plaques which are made up of the amyloid
peptides, which is a piece of that normal amyloid precursor protein. There is a clipping of
the amyloid precursor protein to release this amyloidogenic fragment. Amyloidogenic,
meaning naturally sticky so that it aggregates into oligomers and then fibrils and plaques.
The aggregation of this peptide occurs at the same time as there are abnormalities in the
tau pathway with hyperphosphorylation.
Tau normally helps support the shape of microtubules. When tau is hyperphosphorylated
and the tangles form, the microtubules change shape and lose their function. We can also
see evidence of damaged energy metabolism, damaged mitochondria. So somehow we
go from this normal appearance to these abnormalities in the course of Alzheimer's
disease. Disease modifying therapy presumably needs to interrupt the pathways involved
in these changes. The last couple of slides that I'm going to show you is a formulation of
an approach to disease modifying therapy, that puts all of this into a single context.
I know Peter will have comments to make on this formulation, but many of us believe that
the process must start with the cleavage of the amyloid precursor protein, by beta and
gamma secretase to release that peptide. That all the other changes are downstream to
that event. We release the amyloid peptide. The peptide is toxic, probably in multiple
forms. Leading to neuron death, mitochondrial damage, hyperphosphorylation of tau and
tangle formation. Many of us believe the pivotal event is generation of the amyloid peptide.
The reason we think that is that it's the most concise explanation for all of the genetic
causes of Alzheimer's disease.
There are a number of genetic causes. Some families carry a mutation in one of three
proteins. Those proteins are APP, PS1 and PS2. We now know that all three are involved
in generation of the amyloid peptide. The mutations of the precursor protein and PS1 and
PS2 all increase amyloid peptide generation. The PS1 and PS2 have turned out to be
components of the gamma secretase complex. And the mutations associated with familial
AD increase generation of the amyloid peptide in its most toxic form, the 42 amino acid
form.
The other genetic cause of Alzheimer's disease, Down's Syndrome, where there's an extra
copy of chromosome 21. And the APP gene resides in chromosome 21. That extra copy
results in too much APP. Allowing excess generation of the amyloid peptide. We believe
that this formulation where the beginning of the process is, cleavage of APP to release Abeta,
is the most concise explanation for all the known genetic causes of Alzheimer's
disease. If that's the case, then if we want to develop disease modifying therapy, we
should fit our strategies into this slide.
I've broken down the current programs in this way. Theoretically strongest would be to be
on the left side of the slide and inhibit beta secretase or gamma secretase. It also might
be effective to attack directly the amyloid peptide with immunotherapy. Or with small
molecules that interact with A-beta and promote from the brain. Or, this encompasses all
of the anti-tau therapeutics which I also find interesting and encouraging. But we view that
as a method of reducing the toxicity of the amyloid peptide. So I've put all of the
approaches on a single slide that focuses on A-beta. I will stop here.
Catherine Arnold:
So we should then move on to Dr. Tanzi.
Rudy Tanzi:
Okay, well I think with that very succinct and elegant presentation from Dr. Aisen, it will
save me some words. So I would begin where Dr. Aisen left off. I think it - I'll reiterate and
reemphasize the point that all four established genes we know about in this disease (the
three early onset genes and the late onset genes apoE) all have in common that there's
excessive accumulation of A-beta aggregate in the brain. And this has led to originally
what was called the amyloid hypothesis, which I think is now sorely outdated. Only
because it - most of the emphasis was on plaques. And mature aggregates and fibrils of
amyloid. And I think, as I'll make a point on my presentation, the battle ground seems to
be more toward the synapse and more toward - which is what we originally heard from
Bob Terry decades ago. And we should have listened to him - and that it's the soluble
forms of A-beta, oligomers, namely dimers and trimers, that seem to be more problematic
than the fibrils.
And so I, in my own papers, I like to refer to this as the Synaptic A-beta Hypothesis. Since
I think amyloid hypothesis doesn't really do the mechanism justice. And in the first slide I
show the genes you heard about just now from Dr. Aisen. The APP gene discovered in
'87. We know 21 mutations in this gene now and duplications. And the average age of
onset there is 55.
Presenilin 1 is the real work horse. One hundred sixty five mutations, actually that's up to
about 170 now. And the average age of onset there is the earliest of the three, 44 years.
And that accounts for at least half of early onset familial AD. And then finally presenilin 2,
where there are 11 mutations and the average age of onset is closer to 55. The mutations
in these three genes account for about half of early onset FAD. But early onset FAD is
probably only about 5% or so of all AD. So these mutations in combination account for
just a tiny bit of AD. Maybe 1% or 2% at most.
But what - I want to make a point going beyond what Dr. Aisen pointed out about their
molecular consequences. And it's not - I think it's been - these mutations have not been
properly addressed by biotech and pharma in the past. Because first there was the
misinterpretation that they increase A-beta production. They do not. Maybe a couple of
them do. Most of these mutations don't increase A-beta production. And most of them
don't even increase A-beta 42. What most of these mutations have in common is they
increase the ratio of A-beta 42 to A-beta 40. About 90% of A-beta in the brain is 40. About
10% 42. And you also see some species anywhere from 37 to 43.
If you look at these mutations and what they do to A-beta, generally in any which way they
can they increase the ratio of 42/40. In some cases that means that 42 stays the same
and 40 goes down. Recent evidence also shows that if you flood a transgenic mouse
brain that has FAD mutations of these three genes with A-beta 40, just bringing that ratio
of 40/42 back (in other words, reversing that molecular phenotype) reversing the 42/40 by
actually flooding with 40, you can prevent amyloid. So it really is all about the
stoichiometry and the ratio. Not just the absolute amounts of A-beta and 42. I think when
you're looking in vivo in the brain. And that's what these mutations have taught us.
Now what's so bad about that ratio? The A-beta 42/40 ratio, as it increases, drives the
increased oligomerization of A-beta and aggregation of A-beta starting with the small
oligomers. And there are data now from a variety of groups that these small oligomers
play a role in regulating long term potentiation or neurotransmission of the synapse. I
think Roberto Malinow's data at Cold Spring Harbor are particularly compelling that this is
a normal event. You need A-beta. You need A-beta oligomers to act as a check and
balance on long term potentiation and excitatory synapses.
So I don't think you just want to turn A-beta off. But as you can imagine in any genetic
situation where you have a toxic function, in this case if you have too many oligomers, that
ratio gets too high. Then you're going to be impairing LTP. Rather than just, you know,
acting as a negative feedback. And then you get toxicana function. LTP is turned on too
much. And cognitive impairment ensues. So I think this is the landscape that's evolving
for the new revised version of what was called the Amyloid Hypothesis.
The next slide is on the late onset respect to apoE. And apoE4 is present in about 20% of
the population. In at least one copy. And that increases to over 50% in AD. Thus the
basis for the association with AD. One copy of E4 increases risks three fold. And two
copies increase risk greater than ten fold. This is a rare case of multiplicative transmission.
Not just dominate or additive. But you actually get a multiplicative affect where there are
two oligos versus one oligo. There's really no other example of a risk factor - a non
penetrent risk factor in genetics, never mind just AD. So apoE4 is quite unique.
And the meta analyses that Lindsay Farrow and also Debra Black in our group have
published, suggest that the strongest effect is between 60 and 70. But not limited to those
age groups. E4 works - you see it in early onset, and also in late onset. But the real effect
concentration is 60 to 70. And again apoE4 is associated with increased A-beta
aggregation in the brain.
Next slide - just a way of for people who are looking at - looking for Web sites to keep up
on the newest AD genes. This should be said that the three early onset genes - and
together with apoE, account for only about 30% of the genetic variance of AD. A full 70%
of the genetic variance is not accounted for by these genes. Remember apoE is in 50% or
60% of AD cases. But it is neither sufficient nor necessary to trigger the disease. Other
genes seem to be necessary to interact with it. And also of course lifestyle and
environmental factors.
This alzgenes site - this is something we maintain. This is alzgene.org. The Web site is
updated weekly. And what we do is we take - we keep track of every published genetic
paper in Alzheimer's from day one. There are actually about ten new papers per month.
And if a gene polymorphism or variance has been tested in at least four independent
samples. Either in a single study or a cross studies, we take the data and the meta
analyses. And say well, does this gene hold up?
And we think this is useful because otherwise you kind of hear about yeses and nos. You
hear so and so saw this gene was associated. Someone said it's not. You don't know
what to believe. So this a more of an objective, systematic collection.
And if you look on the top gene - the alzgene top results. Those are listed in order of effect.
So apoE is number one. Most people don't know that the second strongest effect late
onset gene is actually the nicotinic acetylcholine receptor beta 2 based on meta analyses.
So this is updated weekly. It's something to keep track of. Just for those of you who need
like the equivalent of ESPN Sports Center for AD genes.
So in the next slide, then this gets into the Amyloid Hypothesis. Basically mutations in
APP and in the presenilins - early onset mutations as I said earlier drive up the A-beta
42/40 ratio. And that ratio then going to the left side drives oligomerization of A-beta.
ApoE can play a role in this. And the metals zinc and copper - we found 15 years ago in
my lab are necessary to drive that aggregation process.
If you - you can add all the A-beta 42 and 40 you want into a test tube. If you leave out
metals, zinc and copper in particular, you will not get oligomerization and aggregation. So
this is a metal dependent process. And we had basically discovered this back in '92, '93
when we working on the discovery of the Wilson's disease gene, which is a copper binding
molecule. I saw similarities with the Wilson's protein and A-beta.
Going to the right side of the slide, the way the brain likes to deal with A-beta is that most
of the A-beta that's accumulating in the brain as monomer is rapidly exported out of the
brain into the plasma. David Holtzman's work and Beth Ozoklovich's work has shown that
on average an A-beta peptide is exported from the brain eight minutes after synthesis. It's
the second fasted peptide to be transported out of the brain if it stays monomeric. The
problem is if it becomes oligomeric. And again I think there's a normal role for these in the
synapse based on Malinow's work. It's tougher to get rid of the oligomers. They tend to
accumulate. And because of that the problem is that you start to see these oligomers, and
ultimately aggregates and fibrils.
The other way the brain deals with this is degradation. You can degrade A-beta in the
brain deliberately with enzymes such as neprilysin, insulin degrading enzyme, angiotensin
converting enzyme. Or the A-beta can be degraded in the periphery once it's exported out
of the brain. So based on this - and this is what we've learned from the four known genes
- production of A-beta, the ratio of 42/40, changes with the mutations. These ratio
changes drive aggregation together with metals. ApoE plays a role there. I should
mention that on the clearance pathway, apoE is also one of the chaperones that helps Abeta
get out of the brain and into the bloodstream.
Alpha 2 macroglobulin is another one of those chaperones. And the receptor used there is
probably the LDL receptor related protein LRP. Now based on this, and you've heard
some of this already from Dr. Aisen. In the left top text box there are beta and gamma
secretase inhibitors that are being developed by companies like Lilly, Merck and others. I
think there's more promise for gamma secretase modulators. Gamma secretase
modulators try to reverse that ratio of 42/40 that's going up to bring it back down. And I
think that the best way to treat a disease when you have genetic information is simply go
in and try to reverse the molecular phenotype. The molecular phenotype of those
mutations is increases in 42/40 that drives disease, so try to reverse that ratio. And that's
what the gamma secretase modulators that, for example Flurizan from Myriad. And also
much more potent direct gamma secretase modulators not based on NSAIDs as Flurizan
is being developed by Eiasi, Torrey Pines Therapeutics, Merck and others.
One could also try to enhance alpha secretase activity. This can be done with M1 agonists.
It can also be done with nicotinic agonists that are being developed as well. And
cholesterol drugs have also been implicated in reducing A-beta production. Statins have
not done very well in the clinic, as most of you know. But I think that ACAT inhibitors have
based on transgenic studies and preclinical studies. ACAT inhibitors hit much more
selectively into the correct part of the cholesterol pathway that affects A-beta production.
And those data from Dora Kovax are very encouraging.
Going to the other boxes - the vaccine which is being developed by several groups. I think
there are at least 15 companies in the vaccine space. We'll probably get into that.
Basically the peripheral sink hypothesis, which is the right side of the slide, maintains that
antibodies to A-beta, either introduced by passive immunization or by active immunization,
can sequester A-beta in the plasma and not let it back into the brain.
This is the hypothesis that was championed by Steve Paul at Lilly and (David Holsman).
Elan, namely Dale Shenk, favors the hypothesis that antibodies can sneak into the brain,
trigger microglia cells that then digest A-beta directly in the brain. It's possible that both of
these mechanisms of action are at work. And I think we'll probably get into more detail
about the nuances of vaccine in this - in the Q and A.
And finally the left side box next to oligomerization, anti fibrilogenics, Neurochem has
Alzemed, which is a gag based drug trying to block aggregation. And I think we're going
to discuss that as a discussion point.
Prana biotechnology, a company that I co-founded, is targeting the metals zinc and copper.
We have a compound called PBT2. Dosing on it in Phase II finishes actually today and the
results should be available sometime early next year. PBT2 had tremendous results
preclinically in preventing A-beta aggregation in the brain, clearing A-beta and also
rescuing the effects of A-beta on long term potentiation and neurotransmission. And this
drug has good bioavailability in brain. So it'll be interesting to see the Phase 2 results
given that the preclinical results provide for a lot of optimism. I want to make the point that
that drug is not a chelator. It does not actually chelate metal. It simply - it's more of an
ionifer. It exchanges metal from A-beta. And can deliver it to the cell. Or deliver it to other
enzymes like SOD that use metals. It is not anywhere near the affinity of a chelator. And
should not be considered a chelation therapy.
Finally Transition has a very interesting compound AZD-103. This is a scyllo inositol that
also seems to have an affect on both preventing oligomers from forming and protecting
from oligomer induced LTP deficits. And I should make the point that for both of those
compounds, the Prana compound and the Transition compound, they not only prevent
oligomerization, but to some extent, especially the Prana compound, they can dissolve
existing oligomers as long as they haven't been permanently cross linked.
So these are just an outline of some of the therapies that have been developed that are
most - many are in the clinic, many are preclinical. Based on the pathways that have been
elucidated from the four known Alzheimer's disease genes. And I'm sure we'll get further
into these therapies during the conference.
So I'll end my comments there. Thank you.
Catherine Arnold:
Okay, thank you, Rudy. To round out our third topic for our overview before we go into Q
and A, Dr. Peter Davies, who is going to talk about an opposing hypothesis of tau.
Obviously tau has a relationship to amyloid as well. But I've asked him to talk about tau,
the cell cycle series and otherwise. So thank you, Peter.
Peter Davies:
Okay you heard some of the introductory remarks about plaques. And some of the very
obvious causes for therapy here. But - and this is the most popular scheme - the Amyloid
Hypothesis in some version or another says that some mysterious form of amyloid occurs
in Alzheimer's disease and causes this cascade. I don't have time to (disprove) this
hypothesis. I could speak for the next hour and a half about why I think this hypothesis is
completely wrong. But nonetheless it's a hypothesis that is (worth) testing. And I don't
want to come across as someone who says we should not try amyloid therapeutics. This
is a hypothesis. It deserves to be tested fully and thoroughly. I think it will prove to be
wrong. But we'll see. The test of the hypothesis will tell us whether it's right.
Talking about the tau approach. Much, most, almost all the attention, and certainly in big
pharma and in most of the smaller biotech, is focused on development of kinase inhibitors.
Dr. Aisen mentioned the hyperphosphorylation of tau. And presumably if one can prevent
phosphorylation of tau, one can prevent tangle formation. The problem really here is
indentifying which kinase is the critical kinase of phosphorylating tau in Alzheimer's
disease. We don't actually know which kinase phosphorylates tau in the normal brain. So
it's a choice of a kinase deposit in Alzheimer's disease is quite difficult. GSK3 beta and
CDK5 are the major targets. There are a number of publications, particularly from
AstraZeneca on GSK3 inhibitors. We still have really very little idea about whether either
of these kinases has a role in tangle formation.
We don't actually know the phosphorylation of tau is a driving force for tangle formation.
It's quite possible that tangle formation could occur and phosphorylation be a later event
not driving tangle formation. We actually don't know at this point whether there is any role
for kinase inhibited therapy.
If the amyloid cascade hypothesis is not correct then it's obviously incumbent on us to
suggest an alternative hypothesis. And the way we look at Alzheimer's disease is
fundamentally different. The way I envision Alzheimer's disease is as a process that
begins and gives rise to amyloid deposits and tangles made of tau. Rather than having
either amyloid or tau at the top of the cascade, our approach basically sees amyloid
deposition and tangle formation as the result of a biochemical process. But what is that
process? The first clue really came for us from - well from Inez Vincent who showed that
antibodies that were very good at labeling tau were also very good markers for dividing
cells, like cancer cells. And a couple of the antibodies that have been made in my labs to
study tangle formation in Alzheimer's disease, are actually among the best reagents in the
world for looking at dividing cells. And they're used now widely in the cancer literature as
a way of marking dividing cells.
So what do dividing cells and Alzheimer's disease have in common? Well the fact is that
now over the last ten years a large number of markers typical of dividing cells have been
found in neurons in the brains of patients with Alzheimer's disease. There's a lot of
evidence. A lot of very consistent evidence that the cell cycle, the cell division mechanism,
is activated in neurons in Alzheimer's disease. Carl Harris and recently Paul Lorenz has
reported that there is full duplication of chromosomes, full replication of DNA in neurons in
the brains of patients with Alzheimer's disease. This does not occur normally. This is not
the normal process. And Lorenz in a publication just this year confirmed this with three
different very sophisticated mechanisms.
There's also an increasing body of evidence that turning on the cell cycle is actually a
response to a number of different insults in brain. As two papers in the last couple of
years showing that experimentally produced strokes in rat brains turn on the cell cycle.
And the cells that turn on the cell cycle are those that die following a stroke. This is again
very consistent data from some very reputable labs in this field. So turning on the cell
cycle may actually be a response to a number of insults.
We ask the question, well does it really matter? I mean when you see evidence of cell
cycle activation, does it matter to a cell? Is it just something that happens as a
consequence of the disease? And is really an epi phenomenon. Or is it a critical event in
cell death? So we created a mouse with (unintelligible) in collaboration with (unintelligible)
to deliberately turn on the cell cycle in neurons in the adult brain. And we made a mouse
in which we expressed an onkogene, a viral onkogene in adult neurons in the mouse. And
what did we see in these mice? Well the title actually kind of gives it away. It induces
Alzheimer like tau and amyloid pathology. There's evidence - very, very clear evidence of
cell cycle activation. There's very, very clear evidence of tau pathology in these mice. And
there's very compelling evidence of amyloid deposition in these mice. All the features -
that actually really surprised me. All the features that are typical of Alzheimer's disease
can be generated, at least in part, in a mouse just by turning on the cell cycle.
So we've developed a revised scheme that says cell cycle activation may be the what in
the flow scheme. And clearly it is possible to develop at least some amyloid deposition,
some tau pathology, some cell death simply by activating the cell cycle. And we know with
very strong evidence that cell cycle is in fact turned on in Alzheimer's disease. So
obviously we don't think it's a viral onkogene that's turning on the cell cycle in Alzheimer's
disease. But it's quite plausible that human onkergene is activated in response to a
number of stresses. A number of injuries to the brain can in fact be activated. Can
activate the cell cycle. And can lead to the pathology of the disease.
And this really opens up a new avenue for therapeutics. Because one can draw, in fact
very readily, on the vast experience in cancer research to develop therapies for turning off
the cell cycle. Or turning down the cell cycle in the brains of patients with Alzheimer's
disease. If the hypothesis - again it's a hypothesis that cell cycle activation leads to the
pathology of Alzheimer's disease. But maybe we can get upstream even of amyloid
formation. And block the process that leads to the amyloid end paths of Alzheimer's.
I'll stop there.
Catherine Arnold:
Okay, excellent.
We hope that background provides some fodder for our next part of the session, which is
I'd like to go through some of the key pipeline programs. And what I like about our panel
again is that they all have very different points of view. But what I want to do after we go
through them is see if we can reach any kind of consensus on what our view is. So I'm
going to go through each project and I'm going to ask one of our panelists to give us a
take on what the hypothesis for it working is and what might be the detriment to this
approach. Then I'd like all three to weigh in on the relative benefit prospect from a scale of
one to ten. So if you think not about symptomatic therapies, but think only about what you
know of the pipeline for disease modifying therapies. Ten having the most likelihood of
success. One having the least likelihood of success. If you could each give me your point
of view on how you think this program will pan out. We love to commit things in numbers
here on Wall Street so if you could just bear with us today, panelists, we'd be appreciative.
The first program, and I'm going to bring this up because it's the most recent data set that
was released from a late stage program. I'm going to ask Dr. Aisen to talk about, which
is the Neurochem program for Alzhemed where there was some disappointing study
results in August. And I'd love to know, Dr. Aisen, what your view is on what this
mechanism which relates to beta aggregation. And how promising, you know, after you
step back from that experience, how you're feeling about that mechanism.
Paul Aisen:
Okay so Alzhemed is tramiprosate. That's the chemical name. It's a small molecule. It
interferes with the interaction of the amyloid peptide with glycoseminoglycans in the brain.
And the result is that it blocks aggregation and deposition of amyloid. And we know that in
transgenic mouse models, which are the standard models for testing anti amyloid
therapies, that treatments with tramiprosate reduces the accumulation of amyloid in brain.
The program moved to humans where Phase II studies demonstrated that there was at
least some degree of penetration into the brain with oral tramiprosate. Some dose related
reduction in amyloid peptide, the 42 amino acid (unintelligible) in the cerebral spinal fluid.
And as a result the program moved into two large pivotal Phase III trials. Eighteen month
trials with 1000 subjects to see whether these anti amyloid affects would translate into
clinical benefit.
And the results that were announced in August were that while the trial results are very
difficult to analyze as a result of methodologic issues. Mainly in an 18 month, two country,
two language, 70-site trial there was a huge degree of variability in decline, in use of
concomitant medications, including cognitive enhancers that influenced the primary
outcomes in the trial. The analysis is very, very difficult. But it can be said the results of
the trial were not sufficient to support a clinical benefit.
And the company has decided to halt the second pivotal trial and move to try to develop a
form of tramiprosate that might give more consistent and robust brain penetration. Before
moving forward with additional clinical trial.
So what does that mean for the amyloid hypothesis, or more specifically, anti aggregation
therapies? I think what it means is that there was not a robust benefit with this - these
doses of this specific agent. I do not think that that has a significant effect, at least on the
way I look at anti aggregation approaches.
So the closest program in mechanism to tramiprosate would be the scyllo inositol program
that Rudy Tanzi mentioned. Scyllo inositol is chemically entirely different. But it's also
small molecule that interferes with aggregation and reduces amyloid accumulation in
transgenic mice. I do not think that the negative results of the first pivotal trial with
tramiprosate reduced enthusiasm for the scyllo inositol program. Because I think it's all
going to come down to the amount that can be tolerated. The amount that gets into the
brain. And the degree of anti aggregation and amyloid clearing affect that a specific
intervention brings us. So while discouraged specifically for tramiprosate, I would not say
that I'm discouraged for anti aggregation approaches or the broader anti amyloid
strategies.
Oh, you want a number. Okay so let's assume as Neurochem has suggested, they are
going to develop a form of tramiprosate with better brain penetration. Allowing a more
robust anti aggregation effect. Where would I put that on a scale of one to ten? A four.
Catherine Arnold:
Dr. Davies, would you comment on where - on a scale of one to ten. And if you have any
comments you'd like to make.
Paul Aisen:
You can go below one if you want.
Catherine Arnold:
For emphasis negative numbers are very positive numbers aren't they.
Peter Davies:
I won't go that low. But I would certainly be in the one or two range there. But I do agree
with what Paul said. That this is not an adequate test of the hypothesis. This is a failed
drug, or drug in this particular formulation. I don't think one can draw conclusions about
testing the hypothesis that amyloid aggregation therapies are, you know, useless or
helpful. I don't think one can reach that conclusion.
I should point out and it's terribly difficult research to do. Because the mouse models in
which these compounds are tested are very artificial models. In which amyloid deposition
is driven by over expression of the precursor of amyloid. And one can show in those
models that lots of things prevent amyloid deposition. Or even amyloid synthesis. But that
- in the mouse at least there's really virtually no downstream consequences of amyloid
aggregation. There's not tangle formation. There's no cell death. And so it's very hard to
get a read on these therapies. And one can say in the mouse that they do or do not
influence amyloid deposition that. But beyond that it's very difficult to say very much.
So the predictive value of the mouse models is quite low, I think. And doesn't help in
developing these therapeutic strategies. I'm not (unintelligible) yet, but…
Catherine Arnold:
Dr. Tanzi, your vote?
Rudy Tanzi:
Well I would echo those - both those comments. That I don't think Neurochem's Alzhemed
should be a litmus test for whether anti A-beta aggregation therapies will work. I would
also give it a very low score, a one or a two. Because I think that it's hitting too late. The
devil's always in the details. And it's hitting the - it's blocking the ability of A-beta to bind to
gags. And this is kind of a late stage event when A-beta oligomers and protofibers are
starting to deposit. And I agree with Peter on one thing. I don't think the plaques are
where it's at. And this seems to be hitting more late in the path way. If you think about
monomer going to oligomer, going to protofibril, going to fibril. This is hitting late in that
pathway toward deposition as a plaque or as a fibril. And I just think it's unfortunate. And
that's probably why the drug is failing in clinical trials so far. So I would give it probably a
one or a two as well.
But I think there's still great hope for the Transition AZD-103. And in my opinion even
greater hope even for the Prana PBT2 drug.
Catherine Arnold:
So while you're talking, Dr. Tanzi, who don't you give us a hypothesis for why PBT2 will
work when PBT1 was disappointing? And what your vote is there on a scale of one to ten.
Rudy Tanzi:
Okay, I'm not sure PBT1 was disappointing. But PBT1 was clioquinol. And this is based
on the fact that zinc and copper are necessary to drive aggregation as we discovered.
And if you just keep zinc and copper away from A-beta, it won't aggregate. I think Prana
got kind of a bum wrap because people were equating it with chelation. And it's not a
chelation therapy. These molecules are no where near chelators, which have very, very
strong (unintelligible) metals. And don't let metals go. These compounds take metals
from A-beta but then exchange them, or even deliver them to the cell.
In fact in the preclinical mouse studies with PBT1, clioquinol, copper levels went up in the
brain because it was being released from plaques where it was being sequestered. PBT2
- the problem with PBT1 and the disappointment came because its di-iodo contaminant
when it was being ramped up for a Phase IIb study. The Phase IIa studies were actually
encouraging, you know, not significantly positive. But encouraging enough to go to Phase
IIb. But the di-iodo contaminate killed clioquinol. And I think that was a good thing.
Because that drug had some baggage.
PBT2 is a non-iodine derivative. Like clioquinol or PBT1, it's a 8-hydroxy quinoline. It has
about 50 times the bioavailability of the brain. Much more soluble. And in the preclinical
studies just blows PBT1 away. And if you look across mouse studies with vaccine versus
gamma inhibitors versus PBT2, I would place it right up there at the very top. Which is
why I am optimistic.
And I actually think that if - in my - and Peter will like this. I think if PBT2 doesn't work in
humans given the bioavailability and what is done in mice, I will probably start questioning
the A-beta hypothesis myself as a therapy. So I would probably put PBT2 up at around a
nine. But I'm biased.
Catherine Arnold:
At least your cards are on the table.
Rudy Tanzi:
Yes.
Catherine Arnold:
Peter, what do you say?
Peter Davies:
I still can't get much enthusiasm for these compounds. I just don't - I think the amyloid
deposition in - many of the arguments I see based on these very rare genetic cases.
Drugs are being tested in patients who don't have mutations. Don't have mutations in APP.
Don't have mutations in presenilin. I think the amyloid part of the disease is downstream.
So I think you might get a little benefit. Because I don't think a head full of amyloid is good
for you. But - and you might get some benefit out of those somewhere along the lines of
the benefit we get out of Aricept, which might sell some drugs. But, you know, as a
disease modification mechanism, I still have to be down around the (unintelligible).
Catherine Arnold:
Okay. Dr. Aisen?
Paul Aisen:
Well I think that targeting the metals that contribute to amyloid aggregation is a very
intriguing and exciting idea. I am happy to see that there is a compound that can move
this strategy forward beyond clioquinol. And so I also am looking forward to seeing results
with PBT2. I would take strong exception only to Rudy's last comment, which is that the
failure of this program would dampen the whole hypothesis. I come back to, I believe
Rudy your own comments, that the details are essential here. And any specific drug and
dose and patient population, and all the details make all the difference. And no one
program, unless you can demonstrate that you have eliminated amyloid, which we're not
yet able to do. I don't think that any negative program is going to have a major impact on
the hypothesis. In a sense that's unfortunate. And if we make a hypothesis un testable, is
it an important hypothesis?
Well eventually I hope that we will demonstrate an effective anti amyloid approach that will
even convince Dr. Davies. I know that's a high bar to set. But that's what I think will
happen. As to what number I would put on the PBT2 program? I would give it a four.
Catherine Arnold:
Dr. Davies, I want to send the next program to you. And get you comments on the
Elan/Transition collaboration if you could talk about that.
Peter Davies:
Is this scyllo inositol?
Catherine Arnold:
Yes.
Peter Davies:
Again it's not - it's an anti amyloid aggregation approach that's been tested in the
transgenic mice. And I think it's incredibly difficult to know whether those mouse studies
are predictive. But assuming that the mouse studies were predictive, again, I would say
one might get some benefit if one can get the amyloid out of the brain. But I don't think it's
going to be big. I just don't see the evidence that the amyloid accumulation in brain is
really starting the problem. Or a participant in the driving of the disease process. I think
when we come to talk about the immunization therapies that will also be evident. But it
may already be that we have some data on testing it whether amyloid removal is a useful
therapy. And I just don't think it is. So I'm still down at two. You've got to find something
better to move me up.
Catherine Arnold:
We really have to beat you up a little bit about this comment “useful,” I think, because I
know it's going to be a pervasive theme, Peter. So in terms of - you mean useful as the
silver bullet that Alzheimer's goes away? Or do you mean useful in the symptomatic vane?
And I'm having fun with you. So I appreciate you responding accordingly.
Peter Davies:
I think if one looks at the Aricept experience. It's really instructive that, you know, if we look
at the ADAS-cog scale which is basically a 70 point scale, one sees a patient move from a
score of 35 to a score of 41 on Aricept. That's about the average result. I wouldn't be
surprised if removing all the amyloid from the brain produced an improvement of that
magnitude acutely. But I don't think it will modify the disease process. I mean maybe
that's enough to sell a lot of drug. It hasn't been bad. Aricept sales haven't been terrible.
But I just don't think there's a disease modification in that process.
Whatever the process in Alzheimer's disease that leads to all this amyloid deposition,
that's going to be going on anyway. Whether you remove - whether you suck out all the
amyloid or not - the process that leads to that amyloid deposition is still going to be there.
The disease process is still going to go on. So that's why I'm not enthusiastic.
Catherine Arnold:
Dr. Aisen, could you follow up with that?
Paul Aisen:
Well I again put much more capital in to the amyloid hypothesis than Dr. Davies does.
Again while the genetic form - genetically determined Alzheimer's disease - is rare, it looks
very much like sporadic Alzheimer's disease apart from its age of onset and inheritance.
And to me that is - I think, Peter, you'd have to say that it's quite a remarkable coincidence
that every single gene that produces Alzheimer's disease finds itself acting at the cleavage
of the amyloid precursor protein to release A-beta. I still don't see a way that that can be
accounted for unless this is driving the process.
So coming back to scyllo inositol and the Elan program, I found the preclinical evidence
quite strong. I'm not discouraged by the tramiprosate failure. Very interested in seeing
the results of clinical testing. And at the risk of being boring, I would give it a four.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes I would just add that it's not about sucking amyloid out of the brain. First of all I'll just
say my number. I would give the AZD-103 probably a seven. I think the preclinical data
with PBT2 are more compelling. I studied both of them in detail. So I give PBT2 a higher
rating. Just because I think the mouse studies are more compelling. Although they're not
published yet. And others can see that soon as they've been presented.
I think it's not about sucking amyloid out. Both of these drugs the AZD-103 and PBT2,
prevent the early oligomerization of A-beta. And I think we have to think about, you know,
in Alzheimer's as with AIDS, what is our culprit? HIV and AIDS - I think that the data
suggests that A-beta oligomers, long before protofibrils and fibrils and plaques, as their
formings are monomer, are causing the problem. If there are too many of them in a
synapse. A-beta 42/40 ratio increases these oligomers and they cause synaptic
dysfunction. And, you know, many great groups have shown this now. Including probably
the leading LTP guy in the country, Roberto Malinow. It's hard to doubt it. The oligomers
do this normally. And too much oligomer impairs LTP.
So if the battle ground - there's the synapse. And if the villain is the oligomer. And both of
these drugs AZD-103 and PBT2 prevent the oligomer, that's not about letting amyloid
deposit and sucking it back out, as Peter's simplifying it. It's - again the devil's in the
details. It's preventing the formation of the villain. The oligomer itself.
And I think that over the last five years there's been really strong data to suggest there's
been a paradigm shift toward these small dimers and trimers. And they don't only occur in
genetic early onset cases. Alex Roher showed ten years ago that AD brains, sporadic or
otherwise, are full of dimer and trimer. And these are the species that are exerting this
anti-synaptic effect in hippocampal slices and in vivo.
So I'm very optimistic about these approaches that hit A-beta oligomers. And particularly
the Transition and the Prana approach.
Catherine Arnold:
So I think we were anxious to move on to immunotherapies. And I would like to spend a
lot of time on the passive programs. But before we do I'd like to get your take on the
active programs. Obviously the first one being the Elan program where the Phase II
program was shut down due to meningoencephalitis. Could we just get your - we'll just
ask you for your call, on a scale of one to ten, of active program prospects. I think that's
probably the best way to go. And then let's move into the some of the passive programs.
So if I could ask Dr. Tanzi to start off?
Rudy Tanzi:
I'll give active a one. In no way would I want to open the flood gates and have my natural
antibodies producing anti A-beta immune responses. A-beta plays a normal role in the
brain. And I think you open up those flood gates. It's very dangerous. If you can control it
with passive immunization I think there's hope. But I would not want to have an active
immunization in any of my family or friends. So one. Dangerous.
Peter Davies:
I am a vigorous opponent of the active immunization process. I think it's dangerous. I
think it's being done without any reasonable degree of thought about how the immune
system works, without drawing on the vast experience in Multiple Sclerosis, and
Myasthenia Gravis, the classic auto immune disorders. I actually think this is a very
dangerous and silly thing to do. I am a vigorous opponent and will remain so. I'm a
vigorous opponent. In case anybody didn't hear that. Minus ten.
Rudy Tanzi:
Can I go minus ten?
Paul Aisen:
You couldn't think of a number less than minus ten?
Peter Davies:
Far away from ten.
Paul Aisen:
I first want to make the point that the original vaccine trial by Elan was historic. And it was
important. I think that it really strengthened the amyloid hypothesis in some ways.
Although not in others. But demonstrated that immunotherapy can actually clear fibrillar
amyloid. Now I would agree with both of my colleagues that fibrillar amyloid does not
seem most promising as a specific target. But it was very surprising to find that the
immune system could be induced to clear fibrillar amyloid from brain. And I think the post
mortem studies do demonstrate that. I think the clinical results of the trial are equivocal. I
think that they're encouraging. But certainly not highly compelling.
The safety is clearly present. I'm not sure I would have abandoned the program with 6%
meningo-encephalitis. Considering that most of those cases resolved without sequelae.
And that considering that this is a uniformly fatal, devastating, horrible disease. I might
have continued that program. I am certainly a big advocate of active immunization. I again
- if you favor the amyloid hypothesis, I think that the active immunization against amyloid
looks very promising. Because you have greater evidence of amyloid clearance with that
approach than with any other. Now clearing amyloid from mouse brain or even human
brain could be irrelevant. The fact that the plaques were phagocytised does not mean that
you helped people. But when we put the encouraging, if somewhat equivocal, clinical
results of the active vaccine together with that amyloid clearance. I think it supports an
active vaccination anti amyloid approach.
Is it risky? I do think it is. I think that we have some idea of why people got
meningoencephalitis. And I think the current active vaccination approaches which limit the
cellular immune response by restricting the epitope in the vaccination are likely to reduce
that risk. And the experience so far with bapineuzumab suggests that there's some risks.
But perhaps not of such serious toxicity we've seen in the original vaccine. But I am a
strong advocate of active amyloid vaccination. And I would give a number to active
vaccination? I would give it a 4.2.
Rudy Tanzi:
Oh, we can do points?
Catherine Arnold:
Very precise.
Paul Aisen:
I just wanted to be a little more optimistic even than with the others. In response to Peter's
vote.
Catherine Arnold:
Dr. Aisen, why don't we move to passive immunotherapy. And obviously bapineuzumab is
what's coming most here. As far as results in hand for Phase II by mid '08, starting a
Phase III program. Contrast your view, active versus passive, and what you think of
bapineuzumab's success.
Paul Aisen:
Well I'm also very interested in passive immunotherapy. Again if you believe the amyloid
hypothesis has merit, and you see the results of active vaccination, you want to build on
those results. And that includes passive immunotherapy with monoclonals, passive
immunotherapy with pulled human immunoglobulin, and active vaccination. And I'm
pleased that all of those programs are moving forward vigorously.
Specifically with bapineuzumab, it does mimic what is likely to have been the amyloid
clearing type of antibody produced in the active vaccination. And that's in it's favor. The
signal that was seen in Phase 1 was very interesting and intriguing. It supports the idea
that Rudy put forward, that amyloid is a synaptic toxin. And tying up amyloid will have an
acute beneficial effect.
My guess is that although the numbers are so tiny. It's certainly - one can't be confident.
But my guess is there was a real systematic improvement. That is being confirmed in
Phase II. And will be shown to be real in Phase III.
So of all the anti amyloid approaches, perhaps the strongest evidence of clinical benefit
today comes from bapineuzumab. So I think there are safety issues related to vaccination.
And the angioedema episodes, which fortunately don't seem to be so serious, but are of
some concern. But also good evidence of the possibility of a favorable clinical effect. So I
would give bapineuzumab a 4.3
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes - I think the preclinical data with passive immunization have been compelling. As you
know. And I have some safety concerns different than the active. I mean the active is
quite different situation. You're making antibodies A-beta for the rest of your life. But you
know titrating in monoclonals is a safer bet. There are still some concerns about
monoclonal antibodies accumulating on blood vessel walls together with A-beta and
causing microhemorrhages.
One of the reasons that as brought up by Pfizer is their antibodies are deglycosylated with
the hope that it would reduce inflammatory and microhemorrhage\-type potential adverse
events.
So putting safety aside. Just based on what we know about - and I agree with Paul that
the active vaccine experiment, even though it led to encephalitis and their safety concerns,
was historical. And did teach us something. If we then extrapolate that onto passive
immunization and look at in combination with the preclinical data there, I'm actually pretty
optimistic about the Elan monoclonal and other companies.
I think Neurimmune from - Roger Nitsch's company that Biogen just did a deal with.
Where they're actually using monoclonals based on the ones that are really being made in
the body upon active immunization is quite clever. So I would give the whole passive
immunization area, outside of IVIg, but the specific monoclonals against A-beta a seven.
I think there's optimism there. A seven but with an asterisk. Kind of like Barry Bonds,
because of safety issues. Maybe not as bad as steroids, but safety issues.
Catherine Arnold:
Patterns emerge. I was just looking back at the scores and, Dr. Aisen's highest score is a
4.3. Dr. Tanzi, you do have nine in there. I'm doing a relative check here as we go
through. Dr. Davies?
Peter Davies:
My highest score here would be a three or a four for passive immunization. And there's
really a fundamental difference between passive immunization and active immunization.
There are safety concerns with passive immunization. But if something starts to go wrong,
you can stop. With active immunization once you start the process, there is no stopping it.
I was immunized against Polio when I was 5. It's still effective. There isn't no way to turn
that off.
Passive immunization offers -- at least if the preclinical data is right -- a way to remove -
quite effective way to remove amyloid from the brain. I don't think the safety issues will be
anything like as bad as the active immunization. And it's a very intriguing way to test the
hypothesis. I think it's quite possible that one can disrupt oliogomer formation and
certainly reduce amyloid concentrations in brain.
This way the preclinical data is very compelling in that regard. And this is a good way to
test the amyloid hypothesis. I'm in favor of the trials. A three or a four. 3.5. How about
that?
Catherine Arnold:
Just for the record, that's your highest score as well.
Peter Davies:
Yes, that's my highest score so far. Okay.
Catherine Arnold:
Now could I ask just the group to comment on whether nor not your score would change
for a different location of epitope in terms of… Obviously the bapineuzumab program is at
the N terminus. You've got Pfizer's program is at C terminus, and Lilly is at mid region.
Would your number change based on where the activity is targeted?
Rudy Tanzi:
You mean for active immunization or passive?
Catherine Arnold:
I'm talking about passive.
Peter Davies:
No I don't think there's enough information to favor one over the other. And they all should
hit A-beta 40 and 42. So I see no advantage to one epitope over the other.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
I just like the Neuroimmune approach. Because regardless of epitope, they're taking
antibodies that were made in humans as part of active immunization. Saying okay what
are the anti oligomeric A-beta antibodies that are made when you immunize? And then
they're mimicking those in a reverse translation approach. So outside of epitope, I'd give a
little higher rank to that. But in terms of just choosing epitopes, I wouldn't change my
score.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
Well I think our big differences is with different epitopes. I’m just not sure which one is
going to be best. I think that an N terminus antibody is more like to induce the
phagocytosis in brain. Particularly if it's an intact antibody. I think that a mid sequence and
C terminus antibody are much less likely to produce clearance of fibrillar amyloid. I wish I
knew if that was an important target. But it's a - I think a significant difference in approach.
I think if the sequestration idea is prevails, as it well might, then we might have efficacy
with more safety with a non N terminus approach so that would encompass the Lilly
approach and the Pfizer approach. If in fact we want fibrillar clearance, we're not going to
want to deglycosylated the antibody. But maybe we don't need that. And so maybe the
Pfizer approach will prove to be safe and effective.
I'm glad that we're pursuing all of them. Until we beat this problem I think we should be
testing all promising avenues. And my scores - maybe you can guess this but - my scores
for each of those would be in the 4 to 4.3 range.
Catherine Arnold:
We - just for housekeeping purposes we are running a little bit behind. We have - we do
have another about 30, 35 minutes left before absolutely will lose our panelists. So I do
want to get through a couple more programs. And then open it up for questions. Another
passive approach that's a little less selective but obviously talked about quite a bit is IVIg.
I wonder, Dr. Aisen we'll start perhaps with you. What’s your - how do you think about that
program. That's the Baxter program.
Paul Aisen:
Well I am also pleased that this program is going forward. IVIg is pulled immunoglobulin. It
turns out that there are naturally occurring anti amyloid antibodies. Those antibodies are
relatively reduced in people with AD. You can replenish natural amyloid antibodies by
using infusions of IVIg. And biomarker evidence from small studies suggest that you can
have - you can influence the CSF A-beta and plasma A-beta levels by infusing IVIg. And I
think all that's encouraging.
But I'm particularly encouraged by the Phase 2 results. And the Phase 2 results have been
announced rather than fully presented. The announcement is just that there was a very
encouraging signal. So there is not a lot that we can say about that. But Norm Relkin,
who has developed this idea and has presented a number of small studies showing a
rapid and sustained improvement in people with AD with IVIg, has demonstrated that
although the titers of anti monomeric amyloid antibodies are quite low in IVIg, there are
high titers of anti oligomeric amyloid antibodies. And if the oligomers are the most toxic,
that might explain the clinical results he has reported. And it might increase the likelihood
that this approach will come to fruition and will be effective. And of course there's the
advantage of all of the history of IVIg so we know a lot about it. And it's available now. So
I'm pleased that there is a Phase 3 trial of IVIg starting up now. And I would probably give
that a 4.3.
Catherine Arnold:
Okay, Dr. Davies?
Peter Davies:
It's an intriguing approach. And I'm around a three I guess. Having seen the recent
presentation just a couple of months ago from Norm Relkin about the clinical benefit, I
must say I was not impressed that much was going on. A few patients seemed to do quite
nicely, but the aggregate effect did not seem terribly striking. It's an incredibly expensive
approach that is going to be I think limited at best in its availability. And, you know, it's
also one of those very troublesome things where it - if there is a small benefit, I don't think
we're going to know why.
IVIgG is used to treat a lot of conditions. And we have a great deal of experience with it. I
have a great deal of experience with it. We took my grandson who was treated for
Kawasaki disease with IVIgG. And now nobody really knows what the etiology or
pathophysiology of Kawasaki's disease is. But essentially 36 hours of infusion of IVIgG is
a complete cure. As it was in my grandson's case. So why does it work? Well we don't
know. What's the etiology of Kawasaki's disease? We don't know.
If we see a benefit in some Alzheimer's patients then I suspect it will be a small number of
patients. I don't think we're going to know why that happened. So it's kind of, from a basic
science perspective, a very frustrating approach.
Although since it's safe and relatively easy, let's test it. Hence my three.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes, I agree with Peter on this. I would give it also a three. I think that, you know, we
actually published a paper last year where we characterized the physiological auto
antibodies to A-beta in patients and controls. And in full agreement with what Dr. Aisen
pointed out. You see antibodies to oligomeric A-beta that are auto antibodies. And you
also see antibodies to monomeric A-beta. But in the patients versus controls, the only
decrease in auto antibodies was in the ones targeted against oligomeric A-beta. We saw
no difference in the levels and titers of auto antibodies of monomeric A-beta in patients
versus controls. There was reduction in auto antibodies through oligomeric A-beta. And
that reduction correlated with age of onset. So that's the pool you need to care about.
And it's possible that with the IVIg you're non selectively introducing and replenishing
antibodies against oligomeric A-beta in patients. But if you do that, why not just do it
directly? I mean that's why I'm not as enthusiastic about this. You know, make antibodies
to oligomeric A-beta the way Neurimmune is. Or other companies that are targeting
specific forms. Dig in and just do it selectively. I don't see why we have to do it in this
quick and dirty way. So I would give it around a - I guess around a four. I'll go above
Peter just to be different.
Catherine Arnold:
Dr. Tanzi since we're on you could we switch gears? I want talk about gamma secretase
modulators and inhibitors. And obviously the Myriad program is also a late stage program.
What are your thoughts on the promise of potential for Flurizan, and the risk. And what's
your score?
Rudy Tanzi:
Yes, well I'm very positive. I'm the most positive about gamma secretase modulators.
Because, you know, what we've learned that the 42/40 ratio is what matters. And that's
what genetics has taught us. And if you can reverse that ratio genetics predicts you'll treat
the disease effectively. Three different genes tell you that. We see more whether it's
genetic or not. You see more 42 in AD brain. It's the principal peptide driving aggregation.
And it's the principal peptide in plaque.
So I love gamma secretase modulators. But I think Flurizan was first out, but just not
potent enough. The Phase II trial showed no significant results. But except for this post
hoc assessment. There was a sub group in milder AD who took the higher dose. That
800 mg dose that had some improvement, and ADAS-cog during the open label extension,
12 to 18 month open label extension. So Flurizan is not a particularly potent gamma
secretase modulator. It's based on NSAIDs. Its bioavailability isn't optimal. But you know
it has a shot. I would say it has a 50/50 shot to make it out of the block. Because there
was some benefit with the high dose. If their Phase III results that we'll hear about in mid
'08 are positive, you know, they may be okay.
I'm much more optimistic about companies like Eisai, Torrey Pine Therapeutics, and
Merck who are going at it directly. Trying to specifically find new chemical entities that act
as gamma secretase modulators.
Some of the early reports of these compounds are that they are 1000 to 5000 times more
potent than Flurizan in reducing or reversing that ratio of 42/40. Eisai got off the block with
a Phase I that was stopped because of some tox issues. But there's lots of activity in this
area.
So I would give Flurizan a - probably a three. The prospect for future gamma secretase
modulators, a very high rating, eight or nine.
Peter Davies:
I'm sorry the potency in terms of Flurizan is just not there. If you go back to the basic
science literature on Flurizan, the doses that are being given to people are not going to
produce concentrations in brain that are adequate to modulate gamma secretase
(unintelligible). I don't think you could load people with enough Flurizan to have significant
effect on gamma secretase activity. The whole area for me - I'll go up to a five with the
gamma secretase program.
This to me is kind of the Amyloid Cascade Hypothesis putting its money where its mouth is.
You know, if you really believe that amyloid is the critical issue in Alzheimer's disease,
then block its production. You know, do the experiment the simple and direct way with the
gamma secretase inhibitors. And I've very enthusiastic about seeing these compounds go
forward. And I think it will take more than one. Paul was saying, the devil's in the details.
But somebody should come up with a good gamma secretase modulator. And we'll find
out whether the amyloid hypothesis is correct or not. I think this is where the testing is
really going to be vital to improving our understanding of Alzheimer's disease.
If these compounds really work, I'll eat my hat. And say I was wrong. But, you know, we
need to test these compounds to find out if this hypothesis is correct or not. I'm very
enthusiastic about the clinical testing.
Catherine Arnold:
I'm anxious for your number.
Peter Davies:
Five. There you go.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
I think we come closest to agreement among all three of us on Flurizan. I also think that
it's an interesting approach. And that this particular compound may well be not potent
enough, at least in the brain, to get the effect we want. I'm discouraged by the absence of
any biomarker evidence of a pharmacodynamic effect in humans. And I would agree with
Rudy's characterization of the Phase II result. Nonetheless we've got these two huge
pivotal trials. So we'll get a real answer before too long. I would give Flurizan itself a 3.8.
For the whole gamma secretase modulation program, well, I think there are some
theoretical reasons why it might be terrific. But practically speaking I'm not quite so
optimistic. I would give the whole program about a 4.1.
As to the test of the Amyloid Hypothesis, though, I would almost dare I say it? Almost
agree with Peter. But it's not a gamma secretase inhibitor in my view that's going to test
the hypothesis. It's going to be an effective beta secretase inhibitor. When we get an
effective beta secretase inhibitor that really eliminates - that is at least 75% reduction or
more, in the generation of the A-beta peptide. And I believe we're going to get there
before too long. I think we'll be in a position to test the Amyloid Hypothesis.
I don't think a gamma secretase inhibitor will do it. Because there are too many problems
with gamma secretase inhibition itself. In terms of a narrow therapeutic window. You can't
lower it enough without toxicity. I'm not yet convinced that modulation is going to do it. So
I would say the hypothesis will be tested with a beta secretase inhibitor.
Catherine Arnold:
I appreciate your patience. I know we're - we have so much to talk about. I'm just going to
bring up one more category. And then I want you all to sort of identify what's most
promising. And let's get some questions from the audience. Gamma secretase, Peter, if
you could talk about the hypothesis of potential positive and risks. And give us your
number on gamma secretase. I'm talking about the inhibitors, not the modulators now.
Lilly has a program in Phase III where they're among many other pharma companies.
Peter Davies:
Yes, again, I would not draw big distinctions between gamma secretase modulators and
gamma secretase inhibitors. I'm not - the data from the basic science is so complicated.
That I don't think it's really that easy to lump the compounds into two separate categories.
I am enthusiastic about the gamma secretase inhibitors. And about the beta secretase
inhibitors as tests of the hypothesis.
Beta secretase - I put them in the same category. I agree there may be some more
problems with gamma secretase inhibitors than with beta secretase inhibitors. I don't think
there's a huge difference there. And I think clever chemistry can get around a lot of the
potential side effects with both of those. I think these are the class of compounds that we
really - the field really needs to test. And I'm very encouraged that we are getting into
Phase III studies with these compounds.
We've got - the Amyloid Cascade Hypothesis is going to be tested in the clinic by these
kinds of compounds.
So I stick with my four or five - 4.5.
Catherine Arnold:
Dr. Aisen?
Paul Aisen:
Well I'm a little more optimistic about gamma secretase inhibitors than modulators.
Because we can really track the pharmacodynamic effect very clearly. And what most
impresses me about the Lilly secretase inhibitors program is the human pharmacodynamic
effect. There is the clear inhibition of A-beta generation as indicated by A-beta
measurement in blood in response to this particular gamma secretase inhibitor. If we were
to see that kind of clear effect with the modulator without toxicity, great. But the best
inhibition of A-beta generation that I've seen in humans is with the gamma secretase
inhibitor so far. So that I think is encouraging.
What's discouraging, again, is the fact we know that you have very little room to move.
And too much gamma secretase inhibition is going to give you serious gastrointestinal
toxicity and immune problems. And whether you can thread that needle, I don't know. So
I would probably still be in the same range for the Lilly program. I'd probably be about 4.1.
Catherine Arnold:
Dr. Tanzi?
Rudy Tanzi:
Yes, just simply based on safety I'm very negative on gamma secretase inhibitors. As
positive as I am on gamma secretase modulators, I do think there's a huge difference.
With gamma inhibitors, you're hitting notch. You're hitting two dozen other substrates.
And the recent findings that have been published now by (Washy Shu) and Phil Wong and
others. It's also oncogenic. You block gamma secretase and you cause cancer. So I
think this is a doomed technology to inhibit. It's a very different to modulate. It's a different
mechanism of action, different type of molecule, different program. So I would give
gamma secretase inhibitors themselves a one or a two. Maybe - I would say a one. Just
because I think they're too dangerous. You can't just hit gamma secretase with a sledge
hammer. Even if you try to take tiny little whacks. It's too dangerous. But gamma
modulators are safe.
Catherine Arnold:
Dr. Tanzi, now the companies, they all talk about this issue.
But some have more
confidence than others that they've gotten around it. Would you think that's warranted as
you look across the program?
Rudy Tanzi:
I think that eventually, with people who inevitably have different responses to drugs, that
there will be those who will have serious adverse effects - adverse reactions to pure
gamma secretase inhibitors. There are too many substrates. It's way too important and
essential an enzyme. It's removing the C terminal stubs of dozens of proteins after their
extra cellular domains are shed. I don't think you fool around with that. That's very, very
different than allowing gamma secretase activity to occur as with modulators. Simply
trying to target the production of 42 specifically on APP. And there's been great progress
in trying to achieve that selectivity. I'd rather wait for that.
Catherine Arnold:
What did we not talk about that you are most enthusiastic about? Obviously we didn't talk
about beta secretase. And if there's a program there, and there's some mechanisms as
well. So would you each just tell us what you - what you're enthusiastic about that we did
not speak about? Dr. Tanzi?
Rudy Tanzi:
Yes I mean beta secretase inhibitors have potentially greater safety than gamma because
there are fewer substrates. But beta secretase has its substrates, too. (Nuregulin) which
is essential for myelin formation is a beta secretase substrate. The sodium channel, beta
2 subunit which controls sodium channel activity by modulating the alpha channel
expression is substrate. And there are a couple of other substrates. So beta, you know,
many say, well beta secretase is clean because you don't have the substrates like gamma.
You have some pretty important substrates for beta secretase. So it's a safer bet. It's a
nice target. But I think you might also have issues there.
There are a handful of companies - I think Wyeth has at least one small molecule inhibitor
- and a few others. It's been an incredibly tough target to get at. So I think there's going to
be real challenges trying to get an effective BASE inhibitor. But even once we get one,
we're not out of the woods in terms of safety. There's a growing list of essential targets for
beta secretase as well. So I'll put them higher than gamma inhibitors. But I would still only
be at about a four or a five. A 4.5 let's say, for beta secretase inhibitors. Because of
safety issues and the rapidly emerging list of substrates that we didn't know about before.
Paul Aisen:
So as I said before I would put beta secretase inhibitors at the top of my list. I think that,
yes, they are in the body for a purpose. But as far as we can tell, that purpose is current
development and not adult life. The fact you have in the beta secretase (unintelligible) in
mouse. It took a long time to find anything wrong with those mice. I believe this is going to
be quite a safe target. I think there are a number of promising compounds moving forward.
And I believe that in the next couple of years there will be small molecule, a brain
penetrating, highly potent, and safe beta secretase inhibitors of moving into the clinic. So
that would certainly be theoretically, and I hope without too much optimism, at the top of
my list. And that - and I would put beta secretase inhibitors at five or so.
Practically speaking I just wanted to mention one other program. On theoretical grounds,
BASE inhibition is best in my view. But on practical grounds - that is empirical grounds -
looking at actual clinical results, nothing can touch the Medivation result with Dimebon. I
think it was an excellent trial in Russia. I think that the data is simply powerful. And I
believe that you will see replication and proof in subsequent trails. I don't know how that
drug works. Theoretically I have no idea. But as far as empirical data, that's the best that
I've seen.
Catherine Arnold:
Dr. Davies?
Peter Davies:
We all as scientists have to be humble and recognize that many drugs are discovered by
accident. And there's a 1 in 1000 shot with almost any compound you pull off the shelf.
And I - just by the odds - I would give Dimebon in the improving a successful therapy for
Alzheimer's disease.
I think Paul and I agree, though. That our scores are all below fives and below. Because
we don't think we're quite there yet. In terms of understanding the mechanisms of the
disease. At least in terms of what's in Phase I and Phase II right now. None of these
compounds make it above five in terms of enthusiasm. I think we've got more work to do.
You know, I think the understanding the pathways of the disease. And developing
scientifically rational targets based on that knowledge of the disease pathways, is what we
need to do. That's what - where the action needs to be. That's why nothing gets above a
five in my book.
Catherine Arnold:
At last we have a chance to take some questions from the audience here at Credit Suisse.
To see if you all have any questions. If you do, there are two microphones. If you could
step up to the microphone and as a question.
Audience Member:
Bapineuzumab, do we know what human dose concentrations are achieved versus the
mouse model, first one. Secondly given the transgenic mice you, get the plaque formation
but not neuro death. How is that explained? And then, using your mice where you have
oncogenic activation, could you use those mice to test all these drugs? To see if then, you
know, you actually get the alleviation of the plaques and (unintelligible)?
Peter Davies:
I think the mice are terribly difficult. The predictive value of the transgenic mice is very
poor. Because the mechanism - you drive amyloid deposition by over expression of the
amyloid precursor protein. That's not the mechanism that happens in the human disease.
And that's the whole key I think. It's an artificially generated amyloid deposit that doesn't
do very much to the brain of the mouse. So it's very difficult to use these (unintelligible).
You can use them to show that the antibodies, for example, get in and or clear the amyloid
deposition in some way. The mechanism isn't clear. Can we use the mice that we have?
We're already into second generation mice here. So hopefully we'll define the pathways
for it.
Obviously I have my favorite target that I'm not going to talk about in public. It's not public
information. So yes, one - you know, I really think if we tested an anti-amyloid drug in our
mice, we would see small improvements in our mice. Because it's not good to have all this
junk deposited in your brain. But I don't think you're ever going to see more with the
amyloid therapies. And that's, you know, we could talk for two hours about - why the
Amyloid Hypothesis doesn't work. And it's not just my opinion.
Audience Member:
(Unintelligible).
Peter Davies:
No. That's not true. The amyloid - in the mice, as in the human brain. The association
between the amyloid deposit and cell death is very foreign. So the amyloid's over here,
the cell death's over here. I don't think the amyloid deposition has anything to do with cell
death.
Audience Member:
(Unintelligible).
Peter Davies:
Yes, but in the mice in general. I mean, you can remove the amyloid and you don't prevent
cell death. Because there isn't any cell death to begin with.
Rudy Tanzi:
Sure, because it's not the plaques.
Paul Aisen:
Just going along with your question. There is a triple transgenic mouse model that
includes both amyloid deposition and tau hyperphosphorylation. It is also not a perfect
model and it doesn't have much neurodegeneration. But it does have markers of change
that are somewhat reminiscent of tau pathology. And, in that mouse model, if you give an
antibody, you reduce not only the amyloid accumulation, but also the tau pathology. It's
not a complete answer. It's - I'm sure Peter's going to say “Well yes, but you don't have
cell death” which you don't. None of these models are great. But that's been used as
evidence for a connection of the cascade, of the amyloid to the tau cascade.
Catherine Arnold:
Dr. Tanzi, you were going to add something?
Rudy Tanzi:
Well I was just saying in response to Peter. That it's not about the plaques or the deposits.
I mean that's where we were led astray for years. It's about dimers, trimers and soluble
oligomers. These are doing the damage. And removing plaques doesn't necessarily
remove the oligomers. You really need to target what drives oligomerization, which is
the A-beta 42/40 ratio. Or try to prevent those oligomers from forming or dissolve them.
So that's a very limited number of therapeutic targets. Hit 42/40 or try to prevent the
oligomers. But in terms of hitting plaques, it could be even dangerous to dissolve a plaque
because you might turn it back into oligomers.
So I think we have to get a way from this idea that the amyloid therapy - the success of
amyloid therapy depends on whether plaques matter. I think the overwhelming evidence
is that we have to move upstream of the plaques toward - and even upstream of
protofibrills - toward soluble oligomers. And we haven't discussed in detail those data
from excellent groups. But it's overwhelming at this point.
Catherine Arnold:
And I just want to throw something out that I've discussed with each of you. I think that it's
an important point. Is that if you step back and some of these actually are effective. And
you actually know the patients at risk earlier, then the debate may be different. If I think I
understood some past conversations that I've had at least with maybe two or three of you.
As we think about what happening, what's changing is that you are testing the patient with
the plaques already developed. But if you're doing PET scans early on and identifying
patients at risk, some of what we're talking about may be a different conclusion. I - tell me
if that's a wrong statement or…
Paul Aisen:
Well I would certainly say that we are headed towards identifying the genetic carriers of
PS1, PS2, APP mutation carriers, Down’s Syndrome individuals, as well as people at high
risk for sporadic AD on the basis of amyloid accumulation before any symptoms. And I
think all of these groups are going to end up being effectively treated with anti-amyloid
compounds.
Rudy Tanzi:
is that recent data from Christian Haas show that gamma
secretase modulators - the NSAID based gamma secretase modulators - cannot treat
patients with presenilin mutations. Presenilin mutations lock up the active side of gamma
secretase into such a rigid confirmation that NSAIDs, including Flurizan, cannot unlock it.
So there's been a lot of talk of lets use early onset mutation patients in trials. You got to
be careful there because now we know that at least NSAID based gamma secretase
modulators can do very little against most presenilin mutations that lock up the active site.
But again it's all about the details. I mean, all of these therapies and patient selection for
clinical trials. You have to look at mechanism of action very carefully before moving
forward.
Peter Davies:
And disease progression, it's going to be incredibly important to accurately diagnose
patients early. Because you don't want to slow or stop disease progression in a patient
who's already well down the hill. You're really then prolonging the agony. When we do
develop effective disease modification, we're going to have to put huge emphasis on early
and accurate diagnosis so we can treat people really in the MCI stage. Rather than after
they already have a diagnosis of moderate Alzheimer's disease.
Catherine Arnold:
We have time for one more question. I appreciate the warrior like attendance here.
Operator, are there any questions on the line?
Operator:
There are no questions at this time.
Catherine Arnold:
Okay, well we are really butting up against 12 o'clock. This is a very dense and important
topic. I guess as I step back… If I could try to summarize, and you'll tell me if you disagree.
I know that.
That we have a lot to learn about the disease. We're probably more likely to make baby
steps in even some of what we learn. The relevance may be different based upon the
continuum of the disease. And all of that is encouraging. But we still have to wait and see.
Next year will bring us some really interesting insights. I think you both - you three -
sounded optimistic on beta amyloid but not, we ought - the devil will be in the details. I
think we've said that a couple of times.
Any last words from my three panelists?
Peter Davies:
Thank you.
Rudy Tanzi:
One thing we didn't mention is alpha secretase enhancers. Trying to preclude A-beta
production by enhancing alpha secretase. And there you have the nicotinic and muscular
agonists. I think they shouldn't be discounted in terms of keeping track of future trials as
well.
Paul Aisen:
I wanted to make the point that while we gave scores from one to ten, I was thinking in
terms of percentage likelihood to succeed. The last point I would make is while we don't
think that anything is more than even odds, I think that all of them together are extremely
likely to produce at least one winner. So I'm optimistic because of the breadth of the field.
Catherine Arnold:
Well and I have to add that, Dr. Aisen, each of you in individual calls before today,
mentioned the promise of combination therapy amongst even a lot of these mechanisms
that we don't have yet, could be really exciting. Dr. Tanzi, we talked about that yesterday,
you had a very strong point of view there.
Rudy Tanzi:
Yes, I mean my own guess is that since - I guess it's obvious now - I believe that the Abeta
oligomers are to Alzheimer's as what HIV is to AIDS. I think based on the newest
data, that if you hit the production of oligomers with a gamma modulator and you try to
stop them from forming or dissolving them with one of the anti-oligomer strategies. That
that combination will be very powerful. Perhaps together with more powerful symptomatic
treatments. And there I think keeping track of nicotinic agonists, and muscularinic
agonists, which may also have some effect on APP. That might be a third component. So
I doubt any one therapy is going to take care of it. But my hope is that if you hit A-beta
toxicity at the synapse, that hopefully downstream events will also be averted.
Catherine Arnold:
Well I would like to thank our panelists and all of our attendees. I think it's been a really
provocative and fun discussion. So have a good afternoon and a nice weekend.
Companies Mentioned (Price as of 02 Jan 08)
AstraZeneca (AZN.L, 2175.00 p, NEUTRAL, TP 2375.00 p, MARKET WEIGHT)
Baxter International (BAX, $58.05, NEUTRAL, TP $63.00, MARKET WEIGHT)
Eisai (4523, ¥4,400, NEUTRAL, TP ¥4,800, MARKET WEIGHT)
Elan (ELN, $22.09)
Eli Lilly (LLY, $52.55, OUTPERFORM, TP $65.00, OVERWEIGHT)
Forest Laboratories Inc. (FRX, $37.16, OUTPERFORM, TP $60.00, OVERWEIGHT)
Johnson & Johnson (JNJ, $65.91, NEUTRAL, TP $65.00, OVERWEIGHT)
Medivation (mdvn, $14.47)
Merck & Co. (MRK, $57.37, NEUTRAL, TP $57.00, OVERWEIGHT)
Myriad Genetics Inc. (MYGN, $46.14)
Neurochem Inc. (NRMX, $2.47)
Novartis (NOVN.VX, SFr62.10, UNDERPERFORM, TP SFr64.00, MARKET WEIGHT)
Pfizer (PFE, $22.91, NEUTRAL, TP $26.00, OVERWEIGHT)
Prana Biotechnology Ltd. (PRAN, $4.12)
Transition Therapeutics Inc. (TTHI, $11.25)
Wyeth (WYE, $43.89, OUTPERFORM, TP $54.00, OVERWEIGHT)
"....on the biotech battle-field, you need some élan...."
FEATURED Cannabix Technologies to Deliver Breath Logix Alcohol Screening Device to Australia • Apr 30, 2024 8:53 AM
Hydromer, Inc. Reports Preliminary Unaudited Financial Results for First Quarter 2024 • HYDI • Apr 29, 2024 9:10 AM
Avant Technologies to Implement AI-Empowered, Zero Trust Architecture in Its Data Centers • AVAI • Apr 29, 2024 8:00 AM
Bantec Reports an Over 50 Percent Increase in Sales and Profits in Q1 2024 from Q1 2023 • BANT • Apr 25, 2024 10:00 AM
Cannabix's Breath Logix Alcohol Device Delivers Positive Impact to Private Monitoring Agency in Montana, USA • BLO • Apr 25, 2024 8:52 AM
Kona Gold Beverages, Inc. Announces Name Change to NuVibe, Inc. and Initiation of Ticker Symbol Application Process • KGKG • Apr 25, 2024 8:30 AM
Start posting your company's news
Only $200 per official company press release!
