Celebration Thread! Biogen is going to the FDA with Aducan.

Alzheimer's, cardiovascular, and other chronic diseases; biomarkers, lifestyle, supplements, drugs, and health care.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

I have been trying to think about how we could get to a no.
Is non-approval of aducan plausible?
How would that make sense?

I have tried speaking to the mirror and telling the tens of millions of those coping with Alzheimer's dementia and the hundreds of millions (billions?) who likely could benefit from microdosing aducan should not be approved. I couldn't do it. The evidence as was reported in the FDA's own Briefing Document with the lead-in quotes "The FDA's Position" ".... aducan shows extremely persuasive evidence of efficacy ..." etc. etc. (see earlier in the thread somewhere for further details) has convinced me. If that's their position, then from what I understand now I agree with it.

That was then (i.e., ~early November, 2020).
This is now.

From my current I understanding the conversation has shifted since the adcomm. Simply reiterating the talking points from the adcomm is not convincing.

Now we have the dona result and tau selection is on the table. It is no longer realistic to stay with the narrative as it existed way back 5 months ago. It would be helpful if any grownups out there (if they exist) could address this issue.

It is helpful to have open dialogue to arrive at higher levels of insight. Though the nays seem to be at an inherent disadvantage because the story is moving forward with new AD research findings. Admittedly it really wasn't that easy to strongly rebut the statistical appendix directly because it largely was about looking at subgroups some of which were very small. When you have a strong top line result it is not easy to argue against it.


The dona tau results are impressive.
Selecting for the low tercile in the intermediate third greatly amplified effect size. iADRS moved from 32% to 50% benefit making this selection. This is still likely not optimal. Other covariates might further enhance the result. Trailblazer II could give us resolving power to deciles of terciles of thirds which potentially could give us microscopic insight into the optimal tau level. This is telling us that even after selecting for amyloid, aducan still under-reported the benefit for responders because there were a great many with too much tau to benefit. Perhaps this could help to explain the problem of rapid-responders.

Since the timeline was ~January 11 dona initial results published and then ~January 20 FDA seeks more information and extends PDUFA to June one certainly wonders whether there might be a connection. I have been thinking that perhaps one might pre-specify post-facto the aducan results based on the tau levels. Specifically, I considered how the differences in rate of placebo and treatment decline between the low and high terciles ( ~ 8 and 4 and ~12 and 12 iADRS, for placebo/treatment against low/high)
might let one to take the dataset and just extract the high tercile non-responders in aducan. It probably wouldn't work that well with the middle group because the placebo was ~8 similar to the low tercile. Doing this could further amplify the aducan top-line results. Basically, you might be able to have a good educated guess of which of the patients were high Tau even without actually knowing their tau levels. This might not be super accurate at the individual level, though much stronger at the group level. Roughly equal to saying I don't know what someone with a 120 IQ will achieve in their life, but it is easy to predict that on average a roomful of 120 IQ will achieve quite a bit.

It is true that by selecting patients in this way you are removing patients that did not respond which doesn't seem fair to the null hypothesis, yet the dona trial clearly showed that there quite a few patients that did not respond based on tau levels. The high tau tercile showed 0 benefit from treatment, and there appeared to be minimal overlap between the low and high tercile groups so one might simply remove a proportion of the placebo and treatment that progressed the most.

aducan chose the wrong patients (i.e, no tau selection), dosed at the wrong dose for the wrong time, measured with the wrong cognitive instrument, and stopped the trial due to futility with a wrong analysis, and yet remarkably still achieved statistical significance on every primary, secondary and tertiary measure on the high dose 302 arm. The lingering question here is how much additional benefit would have been captured if there hadn't been so much headwind? How do we get a no from that? Recent research has moved us even further away from a no with tau selection. The current aducan extension (with 2400 projected patients) includes tau imaging.
Last edited by J11 on Sat Apr 10, 2021 2:52 pm, edited 10 times in total.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

302 Demographics 1.GIF
You do not have the required permissions to view the files attached to this post.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

I read a forum thread that talked about the role of testosterone and tau. Apparently females have more tau accumulation and potentially this would result in a different treatment response in a clinical trial. The above figure seems to substantiate this view. The high dose 302 female patients had a non-statistically significant 12% reduction in CDR-sb progression versus 33% in men. Women with a more tau-heavy neuropathology would be expected to do less well, in a tau non-selected trial. This also potentially highlights the importance of treating women earlier. Possibly testosterone levels could also explain the LARGE Asian response.

Now that we have reached the age of tau these questions could all be answered economically and rapidly by tau selecting. If only ~150 are needed then perhaps a boutique style of clinical trials could be run. There could be many of these smallish trials conducted that confirmed efficacy in a range of different groups. Such trials could be small, short and cheap. This could be the way of future of AD trials. Even a 150 person trial would have the statistical power to provide very strong evidence. It would avoid the problem of the naive assumption that people are people (despite such contradictory evidence as the presence of males and females).
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

Members of the adcomm have published an article in JAMA reiterating their position as stated in November, 2020. There is no mention in their response of the latest tau findings. They only refer to the state of knowledge as it existed at the time of their decision.

They state that since the 302 low dose did not achieve statistical significance, no secondary measures should be included even for the high dose (which was significant) as specified in the statistical analysis plan. The problem here is that as we have seen on this thread in the correlation plots posted ad nauseum that even if the 302 low dose had achieved statistical significance then such a result would have had to be rejected because it would need to be too far off the regression. Basically, there is a p-value of 100% in favor of rejecting the aducan secondaries. If the effect size were too low --> reject, if the effect size were too high --> reject.

The regression line defines the relationship between SUVR and cognitive benefit. There was an almost improbably high correlation reported for this relationship of close to 100% with a p-value < 0.0001. Perhaps if it could be justified that this p-value is the real p-value, then it could be used to unlock the secondaries. It is then not so much did the low dose achieve statistical significance[?] which the regression line suggests it would have not been capable of, but did the low dose fall where it should on the regression[?] (the answer is yes both low doses were very close to where they should have been).

The argument that there is insufficient aducan replication is also problematic. The dona phase 2 trial strongly replicated the aducan 302 high dose results. Surprisingly dona reported 23% benefit on CDR-sb versus 22% for aducan. Also the dona full result showed very strong reduction in decline of ~70% for CDR-sb at the 36 week mark and p=0.002. What you start to imagine with these results is that the brain appears to be waiting for the next line of treatment for its repair, though this never arrives on the scene. Removing amyloid is clearly not sufficient for a massive sustained treatment effect, though adding in perhaps methylene blue and others known to this forum could do the trick. Stop the fire and then start the clean-up. The empty reductionism that the brain needs one thing (amyloid clearance) and one thing only to repair itself should be seen as self-evidently false. Complex diseases in complex organs often need complex cocktails of therapy. Approving aducan would allow this complexifying to begin.

As the regression line is telling us that many of these anti-amyloids are essentially me-toos, then the collective evidence should become cumulative and possibly one anti-amyloid can largely be thought of as a substitute for another.

One of the items that I am puzzled about in the article was the 2.01% adjustment due to having one of the two hitting ~1% significance. What has me uncertain is that the 2.01% is only calculated after the fact. In the Statistical appendix of the Briefing Document they use ~10% because 5% is the pre-specified statistical level. Without knowing the outcome, you would say, well, ~10% of time at least one of the 2 studies will read out at 5% even when this is just by chance. It is only after we see the top line do we know that it actually made 1% on the top line on one study that we can restate this as ~2% chance to make 1% in at last 1 of 2. i am not sure what adjustment should be made after the fact to account for changing the goal posts.



Discussing all of the unsuccessful trials overlooks that several of the anti-amyloids now appear to have been successful and further that they all belong on the same regression line. The probability of this collinearity happening by chance appears to below the threshold of calculation by online apps.

Discordance of results might now be explainable by tau burdens. This might especially apply to the rapid progressors who had very large leverage on the mean of the results. Tau research was described in the dona trial and aducan now has a large extension that is measuring tau in possibly thousands of patients, when only ~50 per treatment arm could provide a statistically significant readout.


Regarding safety, selecting on tau would mean that those who had minimal chance of responding would not be drug exposed this enhances the overall treatment safety for those treated. The broader issue of safety relates to the fact that the aducan trials really should probably not be thought of as the endgame of Alzheimer clinical treatment. Once approved clinical treatment will migrate to earlier stages of disease (and pre-disease) where aggressive 78 week dosing with a needed cognitive benefit will no longer be centrally important. In the pre-disease stage people would be happy to remove a known neuro-toxin from the brains. Remarkably, some of the patients in the anti-amyloid trials had had AD for 20 years(!). With aducan on the market people would not need to watch and wait as they neurodegenerated. AD is a life-long process with subtle deficits that emerge decades before a diagnosis; approving aducan will people to be treated even for these early symptoms. We know that aducan is safe and effective against AD at super low dosing over a lifetime because those with naturally occurring aducan do not develop AD and have no obvious side effects from their nano-dosing.

What has reduced the safety of anti-amyloids for millions of future patients has been the delay of approval of the treatment. As those millions with amyloid burdens have aged they have not had the opportunity to pro-actively manage their disease trajectory.
Ergo, instead of micro-dosing anti-amyloid with 100 fold less than an effective plaque clearing dose (which has no side effects such as ARIA in mice) mega dosing of ~1 gram was tested. Once the current cohort of symptomatic patients has been treated then less aggressive clinical management can be considered. Those in the current generation might still want to choose less dramatic dosing to reduce ARIA.

The adcomm's current voting opinion on aducan could have been helpfully included, though I suppose the answer is implicit in the sentiment expressed in the article.
Last edited by J11 on Sat Apr 03, 2021 8:04 pm, edited 4 times in total.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

NF52, might you be able to provide some guidance?
I think that I might have identified our very rare AD variant.

I have contacted others on the genotyping platforms who potentially share the variant.
Now that we have full genome sequencing on board it will be easier to clearly relate genotype to phenotype.
Question is what next?

If we can sort all of this out, then perhaps the extended family with this genotype could fund some basic science to see
how our genotype relates to dementia and to Alzheimer's clinical research. It would clearly be disheartening if we were to find that
we had a non-amyloid dementing illness that would not benefit from current strategies. There are many rare dementing disorders that have no reach into the research budgets that the near universal prevalence of amyloid pathology has people age affords Alzheimer's.
What might such research entail? Mouse models CRISPRed to our variant? Are mice really that closely related to humans that almost any variant could be edited into a mouse?
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by NF52 »

J11 wrote:NF52, might you be able to provide some guidance?
I think that I might have identified our very rare AD variant.

I have contacted others on the genotyping platforms who potentially share the variant.
Now that we have full genome sequencing on board it will be easier to clearly relate genotype to phenotype.
Question is what next?...Mouse models CRISPRed to our variant? Are mice really that closely related to humans that almost any variant could be edited into a mouse?
Hi J11, I wish I had Jeff Bezos or Elon Musk on speed dial!

FWIW, I'll offer some second-hand information about your mice and CRISPR questions I've gleaned from talking with NIH-funded researchers at various meetings. Turns out there are multiple mouse models of AD, including "humanized" and transgenic ApoE4 and TREM models. They are used to expand the understanding of risk and progression as well as for early study of treatments, but all have limitations. Some researchers are using induced pluripotent stem cell lines (Ipsc) developed from skin or blood cells of a small number of healthy 70+ year individuals with ApoE 4/4 and ApoE 4/4 indidivuals with AD. Those are useful for studying thousands of biological or chemical possibilities for therapy among other studies. My understanding is that in both cases, the need is for uniformity of a model, not the study of exceedingly rare variants. Joseph Jebelli, a neuroscientist at Oxford whose 2017 book In Pursuit of Memory combines a history and discussion of AD research with his grandfather's experience with the disease, predicted that genes like ApoE4, TREM and APP would be the spires of the tallest cathedrals in Oxford, and the rare variants would be the diverse smaller rooftops--not nothing, but not adding greatly to the "picture" or risk of disease. It will be interesting to see if he is proven correct.

You may be interested in this open access article, just published Friday in Alzheimer's & Dementia.Whole‐genome sequencing reveals new Alzheimer's disease–associated rare variants in loci related to synaptic function and neuronal development

It describe a novel and presumably expensive method of finding rare variant alleles seen in < 1% of a non-Finnish European database. The scope of recruitment and sample size seems daunting.
We performed single‐variant and spatial clustering–based testing on rare variants (minor allele frequency [MAF] ≤1%) in a family‐based WGS‐based association study of 2247 subjects from 605 multiplex AD families, followed by replication in 1669 unrelated individuals....
The identification of genetic determinants underlying polygenic AD has been the aim of more than 1000 genetic association studies, including more than 75 genome‐wide association studies (GWAS) on AD and related traits as outcomes...The largest AD GWAS to date was conducted on over 600,000 individuals...and highlighted a total of 29 independent genome‐wide significant (P < 5×10–8) AD risk loci. Another recent GWAS by Kunkle et al. found 25 loci in their analyses of clinically diagnosed LOAD in over 90,000 individuals. Essentially, these and other AD GWAS focused on common (typically with a minor allele frequency [MAF] ≥1%) variants either directly assayed or imputed using high‐density reference panels. The few exceptions to these common‐variant studies utilized either microarray‐based or next‐generation sequencing (NGS)–based genotyping limited to exonic variants and identified rare (MAF <1%) missense variants either increasing (TREM2, PLCG2, ABI3, ADAM10) or decreasing (APP, CD33) risk for AD.
This explanation of the effort and scope of finding rare variants suggests to me that researchers like Dr. Rudy Tanzi at Mass General, the corresponding author of this study, are not looking for small-fish family sets. Dr. Tanzi's email is in the article author information (he's listed last) so you could reach out to him and share what you think is your rare variant and see if he can confirm its risk, or is interested in knowing more.

Good luck!
4/4 and still an optimist!
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

NF52, thank you for responding!
I was not sure how the basic science would unfold.

My concern is that AD might now start to degenerate into a million types of Alzheimer's.
Cognitive decline even with amyloid (and tau) present might be more heterogeneoius than we might
imagine as of now. Even the new genetic findings that you cited, speak of AD moving in a range of new
directions. The brain is highly complex; what if only one path is through synaptic dysfunction? What of
all the rest?

One observation that I made while contacting one of the potential AD affected extended family was how
well connected they appeared to be into the scientific community. I thought Bingo! We all have different
gifts and the trick is to find those with the right gifts and then, well-- it's their baby. If the variant truly
is AD dominant causing then their would be substantial motivation.

Also of interest is that yes I am familiar with the researchers you noted. We sent them cheques to help the cause.
The thing that I have noticed with a fair amount of charitable giving is that it can be somewhat disturbing. People
want to help out others, yet sometimes it devolves down to monetizing other people's misery. I have seen this first
hand and it was fairly shocking.

What I can see with AD and other charitable endeavors is a way out of that type of charity. Connecting people up
with clear objectives could help drive a more engaged form of philanthropy. Setting up a unique strain of AD mice
and seeing how they responded to amyloid treatment would certainly enhance the connection between research
and those affected. Perhaps this will be more how people understand research in the future, especially if it turns
out that every type of AD is its own snowflake.l
NF52
Support Team
Support Team
Posts: 2772
Joined: Tue Oct 25, 2016 9:41 am
Location: Eastern U.S.

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by NF52 »

J11 wrote:Setting up a unique strain of AD mice and seeing how they responded to amyloid treatment would certainly enhance the connection between research and those affected. Perhaps this will be more how people understand research in the future, especially if it turns out that every type of AD is its own snowflake.
Serendipitously, I got a link today to this virtual conference coming up April 13 from 3:00-4:00 PM Eastern time:
FROM BENCH TO BEDSIDE: GROWTH AND INNOVATION
The mouse model is a key platform for drug development. Join JAX to learn more about how the Laboratory is continuously innovating in the field of mouse model development. Leaders will share how JAX® Mice, Clinical and Research Services are developing new solutions in the preclinical path, helping scientists around the world drive their research forward, and, ultimately, helping patients across the globe.
My understanding is that the Jackson Lab in Bar Harbor, Maine is one a leading developer of transgenic and humanized mouse models for AD and a variety of specific pathological processes (tau models, amyloid models, APP etc.) which can then be purchased by researchers designing preclinical trials to discover the heterogeneity you mention. I watched a presentation last year by them and came away impressed by the "figure it out" attitude.

As for a million AD snowflakes, I haven't heard that prediction, but Dr. Karlawish at Penn does quote a researcher as saying "The rarest type of Alzheimer's disease is Alzheimer's disease" --meaning that especially over age 80, many people diagnosed clinically (without PET or MRI scans) have biomarkers of TDP-43 protein misfoldng (possibly more men than women), vascular dementia and Lewy body dementia. People who clearly have biomarkers of amyloid beta and/or tau have it in different areas of the brain, for reasons that are unclear, with differing effects on functioning and cognition.

What I have heard researchers say is that they anticipate a "cocktail" of treatments being used for some people: maybe early anti-amyloid drugs in ApoE 4/4 followed by a low maintenance dose; others having a synapse-preserving drug (like CT-1812 which is into Phase 2 trials in MCI and mild AD); others needing a tau-reduction or prevention drug. And ALL of them are in favor of lifestyle and environmental interventions to build and maintain brain resistance and resilience!

Here's one more recent article that may give you a cognitive resilience gene [Chromosome 8 rs12056505 CC or CT] to check when you get your results: Coping with brain amyloid: genetic heterogeneity and cognitive resilience to Alzheimer’s pathophysiology
This study leveraged a large, population-based sample of amyloid PET positive older adults who had genome-wide SNP data and serial longitudinal cognitive assessments to understand genetic factors that contribute to cognitive resilience. Through this design, we discovered a novel association with cognitive resilience to amyloidosis for a locus on chromosome 8 which was specific to the amyloid positive setting and displayed replication in an independent cohort. We also identified biological pathways with enrichment of association, including a preponderance related to immune system activation. Our data support the hypothesis that genetic heterogeneity is one of the factors that explains differential cognitive resilience to brain amyloidosis
4/4 and still an optimist!
User avatar
SusanJ
Senior Contributor
Senior Contributor
Posts: 3058
Joined: Wed Oct 30, 2013 7:33 am
Location: Western Colorado

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by SusanJ »

J11 wrote:Perhaps this will be more how people understand research in the future, especially if it turns out that every type of AD is its own snowflake.
I think you probably are closer than anyone here, at least in terms of the genetics, in understanding how each of us is unique enough to require different approaches to avoiding AD, and more generally, our health.

I studied pharmacy many years ago (not my profession though), and even back then, felt strongly how we needed to treat people individually. Why should a petite woman of my size need the same amount of a particular drug as a 6 foot, 200 pound man (and back then almost all drug testing was on men!)? The same thing is true when we look at genetics, and how people metabolize drugs, and even vitamins and supplements. And what about food intake, and the conversion of those building blocks into useable metabolites? So many places where things can go off the rails in our health (and why many of us are thankful for Bredesen and his attempts to get doctors and researchers to think holistically in treating AD.)

I truly hope that some day we can get to truly individualized medicine. Perhaps it won't happen in my lifetime, but it sure is a good goal.
J11
Contributor
Contributor
Posts: 3351
Joined: Sat May 17, 2014 4:04 pm

Re: Celebration Thread! Biogen is going to the FDA with Aducan.

Post by J11 »

Thank you for replying SusanJ.

Yes, I think that we now need to be aware that genetic uplift is starting.
Life will never be the same.

Natural reproduction is no longer plausible.
Might not be that long before it is illegal.
The idea of simply spinning the genetic roulette wheel is already absurd.

Human traits are polygentically determined and using embryo selection will rapidly and dramatically change the nature of humanity.
The first eugenic generation is probably already emerging and will likely be quite appalled with us unselecteds.

There has been a great deal of blurring of distinctions in clinical trials.
I was surprised to read on this forum that women have higher tau levels than men.
This might explain the apparent difference in response seen with anti-amyloids.
Now that we have tau on board and the potential for micro-scale clinical trials more of the demographic individuality might
be reported. I am anxious to see tau selection trials for men versus women, epsilon 44 vs placebo etc..
This would show us that there can be fairly large effects with treatment.

I suppose ironically the big point that is illustrated on this forum is that a broad genetic narrative is possible in medicine.
Yet, APOE epsilon 4 is almost alone in this respect. There are few if any other similar common genotypes that have such a
detrimental large effect. The genetics community had been all prepared to find hundreds of epsilon 4 like genotypes to explain
human health: All they found was epsilon 4. The hindsight logic is that if something is detrimental than it is selected against through time. The strange part of APOE 4 is that it is the ancestral allele! The ancestral allele increased AD risk. I would expect that no other
illness would be like that. It is only with dementia that people have been able to greatly increase its prevalence by increasing life expectancy,

The other counter-point is that the long-term will probably move more to human medicine than medicine for the person. Human being understood as genetically uniform and specified. This probably will be a tough one for people to cope with, though it is the logical expectation given economic efficiency. When you buy a car you don't buy a car that has been built specifically for you. They make cars in a factory and all of them are the same. This creates overwhelming efficiencies and greatly improved safety. With people who knows? Everyone is a unique snowflake. There are probably tens of thousands of different genetic variants that cause human illness. There are thousands and thousands of medically recognized diseases. The future will not be like that. There will be a highly limited number of medical illnesses. There will probably be few if any genetically caused illnesses.

It must be so fascinating to be a doctor today in an ER. You literally have no idea what might walk through the door. You have no idea what their biology might be like, how they might respond. This was true with our loved: went in for a simple lithotripsy operation; they injected the contrasting agent; medical crisis ensues. We are entering the time when this will not happen anymore.
Life will be genetically specified. Communities will be genetically specified.

My expectation is that life will begin to be much more highly functioning as genetic uplift occurs. We will know its happening simply by there being less drama. What will be especially interesting is the cognitive uplift. People are going to become smarter and smarter and smarter. This is going to be fascinating to watch. This is the best time ever. We have a front row seat to watch IQ dramatically increase. It is very well understood that this will dramatically change society. With AD you can see the reverse process unfold for decade after decade. With uplift, completely new behaviors will begin to emerge. This might last for decades.
Post Reply