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 »

The principal investigator must submit the complete trial protocol, identify the relevant CMS research questions that will be addressed and cite the location of the detailed analysis plan for those questions in the protocol, plus provide a statement addressing how the trial satisfies each of the standards of scientific integrity (a. through m. listed above), as well as the investigator’s contact information, to the email address below. The information will be reviewed, and approved trials will be identified on the CMS website.



Email address for protocol submissions: clinicalstudynotification@cms.hhs.gov Email subject line: "CED Monoclonal Antibodies for the Treatment of Alzheimer’s Disease [name of sponsor/primary investigator]"


D. Nationally Non-Covered Indications
Monoclonal antibodies directed against amyloid for the treatment of AD provided outside of the CMS approved randomized controlled trials and trials supported by the NIH are nationally non-covered.
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. Proposed Decision

All of the above report is then reiterated.

II. Background

I found it of interest in the Background section note is made of the NDSS.
The only other mention of the NDSS is that they made a comment during the first CMS comment period for the coverage determination. What struck me as odd is that no consideration appears to be given to those with Down's syndrome. Down's syndrome can roughly be thought of as Alzheimer's in children. Why was this not more fully considered in the CMS' decision? What anti-amyloid coverage does the CMS envision for these children?

It is probably for the best not to bring this overly to the forefront in the highly politicized environment that exists for this determination, though some sort of comment would seem warranted. Interestingly, given the somewhat rarity of Down's syndrome insurance and other coverage seems to be more easily attainable within budgets. Admittedly, difficult questions regarding informed consent, risk tolerance etc. would need to be contemplated carefully (probably preferably away from highly charged rhetorical environments).

NDSS – National Down Syndrome Society
Last edited by J11 on Sat Jan 22, 2022 11:28 am, edited 1 time 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 »

Epidemiology
Clinical presentation and progression
Etiology and diagnosis
Treatment
III. History of Medicare Coverage
A. Current Request
B. Benefit Category

IV. Timeline of Recent Activities
V. Food and Drug Administration (FDA) Status
VI. General Methodological Principles

VII. Evidence
A. Introduction
B. Discussion of Evidence
1. Evidence Question
2. External Technology Assessments
3. Internal Technology Assessment


I wanted to provide the organizational outline so I could go back and comment on any sections that I found of particular interest. I can't see that much of interest on commenting above.

However, I did find it odd that they decided to take a pubmed approach. Why read about it in the library when you could contact the FDA and hear their insights after months of intensive interaction with the dataset. The FDA's modeling that was reported was not that clear to me, though it appears that they were able to develop much greater understanding then what could be obtained by merely reading typical scientific literature on anti-amyloids. Even something as simple as a waterfall plot would be of great interest. There were many patients treated on high dose that had 40% reductions in cognitive decline. Would be interesting to see the comparison with those on placebo. This would be the power of open science; one could ask these questions and then have the dataset available to answer them.



4. Medicare Evidence Development & Coverage Advisory Committee (MEDCAC)


Table 1 contains summary descriptive information about individual Phase 3 RCTs.
Table 2 contains summary results from Phase 3 RCTs.




These evidence tables are clearly problematic.
Table 1 includes the phase 3 trials for Bapineuzumab, Solaneuzumab, Gantenerumab, Aducanumab and
Crenzumab. All of these phase 3 trials failed except Aducanumab's Emerge.
Table 2 then provides detailed analysis of these trials.
All these tables display are the treatments that failed (except Aducanuamb).
Where is the value in that?
The prior expectation would be that the later stage treatments would have failed because there hasn't been
a treatment breakthrough until recently and most of these have reported in phase 2.
It is essentially tautological: Treatments fail until they don't and then they succeed.
The treatments in the tables are basically all the wrong turns until there was success.

Using this approach to finding evidence of anti-amyloid efficacy does not seem advisable.
I do not have the best grasp of the finer details of which specific species of amyloid each of the anti-amyloid treatments targeted, though Bapineuzumab seems to have targeted monomers which is now clearly understood to be a wrong turn. Solaneuzumab also did not seem to have effect on what is thought of now as the optimal target (oligomers).

My idea to look at the CDR-sb versus SUVR correlation plot for the various anti-amyloids can be seen in retrospect to be brilliant. Yeah J11! With the scatter plot you can see right away which of the anti-amyloids are working to plan.
It really is not necessary to go through all the decades of failure with the anti-amyloids: in the scatter plots the successful treatments jump right out of the figure. Clearly the 4 most recent anti-amyloids are exactly where they should be. With the others it is not so clear. Solaneuzumab might actually have some efficacy, though it has a different mechanism of action than the others.

Gantenerumab is now recognized as being on target. It was simply a dosing question. The failed phase 3 trials were only dosing with 105 and 225 mg; new phase 3 dosing 1020 mg. It was more that patients were not receiving enough treatment. These Gantenerumab phase 3s fits right in line with expectation on the correlation plots on this thread.

Aducanumab is well known to us.

I am not sure about Crenzumab; some of the earlier reports did seem to show some efficacy, though it is not clear whether this was simply a problem of not updosing.

With science the game plan is to keep failing until success occurs.
It is a brute force approach that is not compatible with the theory that everyone deserves a trophy in life for participating. Until Aducanumab there arguably had been a 100% failure rate that lasted for over a century.
Science is not for wimps.

Yet, fail fail and then fail more is the goto game plan for all of science.
If you take an historical point of view of anything in science that is what you'll always see.
Honestly, I could do science and probably greatly enjoy it, but at a certain point after another
century of failure and another ride on public transit, I probably would call it quits.
An eternity of non-success would just be too much for me.
Yet, the funny thing is that eternity is over-- the dawn has arrived; yet for whatever reason
a whole bunch of people have not read the memo.

And that is what you see with anti-amyloids-- a lot of tissues.
Hindsight bias with science will always be failure. Table 1 is merely showing us all the treatments that did not report success (except Aducanumab).


But when you do start to see success as with Aducanumab, then science usually has turned the corner.
If you want to bring this success into focus the strategy is not to start with Dr. Alzheimer in Germany in 1905,
but fast forward to the leading edge of change for today. You want to read online (and tabloid) headlines more
than antediluvian accounts on papyri-- basically reverse history: once success happens start from today and work back. When you think of it in that way, you no longer see a century of failure but about 2 years of uninterrupted success. ALL of the most recent Alzheimer's trials have reported success.

If it were me making these charts I would actually have included the trials that have shown success, though these
are largely not phase 3s. Starting from today and working back, we have Lecanemab (phase 2), Donanemab (phase 2), Aducanumab (phase 3 Emerge), Aducanumab (phase 1 Prime). All of these trials reported statistically significant results.

One additional consideration is the gamma entrainment trials including FLICKER, OVERTURE etc.
The phase 2 Overture actually posted statistically significant results with quite very few patients in the trial.
"...treatment group exhibited a significant 84% slowing of functional decline in ADCS-ADL scores as well as a significant 83% slowing in memory and cognitive decline as measured by MMSE scores compared to sham."
The trial measured amyloid so the relationship to cognition could be posted over the nearish term.
It is almost embarrassing (unbelievable) when you can report an amyloid success with only 19 in the placebo arm.
The contention that anti-amyloids are ineffective then becomes untenable.]

There is also emerging clinical evidence for an anti-dementing effect for curcumin. I am not clear whether SUVR
amyloid results have been published, though it is widely recognized that India (heavy consumer of curcumin in curry) has low dementia rates. Interestingly, curcumin is used as an amyloid tracer as it binds amyloid and then it glows on imaging. (It then also has the ability to disrupt amyloid accumulations).


One other aspect that probably should be included in the evidence survey are the in progress clinical trials for anti-amyloids. The CMS might contact the pharmas and ask if they might provide the latest futility analysis. Biogen seems to do nearly continuous futility testing. For example with the Lecanemab phase 2 I seem to vaguely recall that they were surprisingly doing futility analyses every month! It would seem highly relevant to me to know what the latest futility p-value is in order for the CMS to be fully informed of the current state of anti-amyloid knowledge. For purposes of example, I would be surprised if the Clarity phase 3 would even have a p-value as high as 0.10 at his point in the trial. If such a p-value were present, then the statistical plan would probably call the trial as futile (though I really hope that they don't call another trial as futile might be too much for me). What I am rambling towards here is that if the Clarity trial has a p-value at most 0.10, and the Graduate 1 trial has a p-value at most 0.10, and the Graduate 2 has a p-value at most 0.10, and the Donanemab Trailblazer-2 phase 3 has a p-value at most 0.10 and they are all independent trials then the combined p-value is 10^-4. Roughly, we already can make statistical inferences about anti-amyloid efficacy with the existing unreported data.

From the patient point of view perhaps the CMS might also ask the pharmas about the betas. As has been noted previously on thread the FDA and to a large extent the CMS are in the not be wrong on alpha business. Don't make an alpha error! Don't approve treatments that don't work! The two phase 3 rule at at least 5% p-value gives you an alpha of 0.25% error rate. Obsessing about alpha values that can be lower than 0.25% no longer represents a patient centric viewpoint. It is allowing an external and alien objective to dominate the conversation away from what patients are focused on. It is difficult to imagine that patients would find it important to argue where alpha were 0.1% or 0.01%. For patients the focus would seem to then be more centered on beta. Don't make a beta error! Don't not approve an effective treatment. This can be a somewhat confusing distinction. Don't approve an ineffective treatment is different from do not not approve an effective treatment (i.e., approve an effective treatment). Beta has been almost entirely absent from the discussion to date. From best that I could make out beta was 60% for high dose Emerge. I take that to mean that not approving Aducanumab would be the wrong answer with 60% probability. The betas for the other trials would be expected to increase this percentage.


Don't be wrong (alpha wrong)! With the information from the anti-amyloid trials we have:

Emerge High Dose top-line primary CDR-sb p = 0.012
Donanemab phase 2 top-line primary iADRS p = 0.042
Lecanemab phase 2 wiggle ADCOMS p = 0.023

Grand total p = 0.000012


(The wiggle here was that they did not actually hit the top-line primary though this was because they moved the goal posts along with other irregularities. Since the top-line was missed they called all subsequent comparisons as nominal. To start "success" was redefined as 12 month superiority of treatment over placebo by at least 25%. This was considered clinically superior. "Success" will be understood here to be superiority versus placebo as this was the definition used in both the Emerge and Dona phase 2 trials and this removes the qualifier of "nominal" in the analysis. The highest dose in the leca phase 2 only had a 64% probability at 12 months of being clinically superior - this was the topline primary result which was reported as negative. Yet, at 18 months there was a 76% probability of reaching clinical superiority (success was defined as 80%). An irregularity of note is that the APOE epsilon 4s were not enrolled into the trial once the European Health regulator considered this potentially unsafe. Yet, these are the patients that achieved the highest responses. The 2.3% p-value that I included above was the apples to apples comparison of 18 month superiority against straight placebo. This was reported as 97.7% in favor of Lecanemab.

Considering all of these trials as independent, there is 12 chances in a million that there is an alpha error.
Probably should correct for the Engage miss as this could then be accused of cherry picking.
There is also the criticism that simply multiplying the p-values together like that would tend to favor
merely having more trials that might or might not actually be that effective. Yet, with the above 3 trials
we can clearly see that the evidence did support relevant anti-AD effects.

Of course the even more powerful approach is to consider the scatter approach that has been applied consistently on this thread. Using that approach everything lines up almost perfectly except Engage high dose. Everything else is highly aligned with the regression. This includes Emerge High Dose, Emerge Low Dose, Engage Low Dose,
Ban2401 High Dose, Ban2401 low dose, Donanemab phase 2 top-line, Gantenerumab High Dose Scarlet Road, Gantenerumab Low Dose Scarlet Road (was slightly off, though in the right general region), and the placebo origin (might not be considered an independent point). Additional Ban2401 dosing arms were also fit with strong alignment with the scatter plot. With all of this information 12 per million would likely be a highly conservative estimate.

The gamma entrainment result might also be included, though I suspect that it will not actually fit the regression line that well. With gamma entrainment there does seem to be an amyloid response, though other therapeutic features also seem to at play. From the evidence to date gamma entrainment appears to be an over-achiever in terms of the CDR-sb SUVR relationship. )

Last edited by J11 on Sat Jan 22, 2022 4:42 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 »

I do not feel that it is meaningful to debate the efficacy of anti-amyloids.
It is essentially abundantly clear that anti-amyloids have anti-dementing effects.
The evidence that has already accumulated is conclusive; much more is now on the medium term time horizon.
The majority view of the FDA pharmacology group was for full approval of Aducanumab (i.e., without the accelerated approval limitation). Those who dispute Aducanumab's efficacy do not appear to realize that it was only a minority view that it would even need any further supporting clinical evidence.

VIII. Public Comment
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 »

IX. CMS Analysis

The FDA reports that Biogen’s post-hoc, secondary analysis also demonstrated that Study 302 [EMERGE] met statistical significance for other, secondary outcomes assessing cognition and function: ADAS-Cog-13 and ADCS-ADL-MCI.


I am unsure about the claim that it was a post-hoc analysis. Emerge high dose reported out cleanly on the topline CDR-sb result with high statistical significance (p=0.012). The secondaries appear to be properly formulated pre-hoc and they also read out cleanly. The FDA Briefing Document suggestion that the order of readout for Emerge would be
High Dose primary (CDR-sb); Low Dose primary (CDR-sb); then High Dose MMSE --> ADAS-cog --> ADCS-ADL-MCI
does not seem believable-- The sensible idea that should be considered is High Dose primary --> High Dose Secondaries. This was thoroughly discussed in the second round of FDA Briefing Documents. Even the placement of the MMSE as the first round selector for the secondaries seems somewhat odd. The MMSE is not a very sensitive measure in early stage AD patients.

What is also of interest is that even CDR-sb does not appear to have been the optimal primary measure for the Aducanuamb trials. We have seen in other trials that ADCOMS, and iADRS (ADAS-cog + ADCS-ADL-MCI) often are more accurate indicators of AD cognition. For example, donanemab's phase 2 did not readout positive on CDR-sb, though on the iADRS primary it did achieve significance. This also was seen in other trials including with Aducanumab.

For Emerge high dose , ADCOMS reported 22% less decline with p= 0.0049; iADRS with 34% less decline p= 0.0002;
ADAS-cog 27% less decline p=0.0097; and ADCS-ADL-MCI 40% less decline p=0.0006.
Even Engage high dose did much better on iADRS with 15% less decline with p= 0.1025.

On these arguably more sensitive measures of cognition, there is no longer any question of Aducanumab's efficacy.
Emerge's iADRS p-value is highly impressive.

It should also be noted that the Aducanumab phase 3 trials had the unusual feature of a non-stationary primary endpoint. The trials changed midway through and patients then received more treatment. This is very unusual.
Almost all clinical trials would be expected to have unbiased estimates of the effect size through the entire duration of the trial. This was not true for either Emerge or Engage. The primary results for both of these trials were clearly
moving towards higher clinical benefit from the time of the first futile analysis to the final readout. Such bias makes it unclear how to understand the 301 high dose result.
Last edited by J11 on Sat Jan 22, 2022 4:45 pm, edited 6 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 »

One suggestion that could improve patient safety that I have noted before is that perhaps a central location could be used for interpretation. With high end artificial intelligence processing, a very high quality analysis could be done that might greatly reduce the risk of misreadings of the ARIA scans. This problem apparently was involved in the fatality that was reported recently for Aducanumab. Such infrastructure would then be of great benefit to those remote centers that might not have the available expertise to properly read the scans. The CMS could show leadership on this issue by initiating the development of this infrastructure.
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 »

Extra 2.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 have added in two more anti-amyloid results in the figure above.

First there is the orange dot on the pink line to the bottom left with an orange line to the right.
This is the CDR-sb result for the phase 2 Overture gamma entrainment study. The reported result was -0.43, though this was not significant (p=0.485) as there was a small sample size involved. Other readouts including ADCS-ADL and MMSE which were strongly positive p= 0.0009 and p=0.0127, respectively. No amyloid results have been posted yet for this study though as there are various other potential mechanisms of action unlike the more monotherapeutic effect of the anti-amyloid mabs, I drew an orange line to the right of the on regression result to suggest how the trial might have overachieved its pure amyloid lowering effect. So, for example, the "true" amyloid result (i.e., without other treatment effects) might be closer to -.30 SUVR with a largish overachievement deviation down to -0.43 CDR-sb. This overachievement gap could be quite largish, though it is unclear without further evidence how large. I anxiously await the amyloid readout on this gamma entrainment trial. It is particularly interesting to note that with the amyloid readout and then any deviation from the regression, we could have quite a good idea of the magnitude of other potential AD mechanisms of action (e.g., anti-inflammatory, etc.). The regression line now seems so solid (and it should become even more solid as yet more clinical trials report out), so the other AD treatment effects could be known with impressive precision.

What is also of note is that as a rough guess ~0.36 might be close to maximum for amyloid removal for an MCI/Mild AD sample. Thus, possibly 0.50 CDR-sb difference over 18 months versus placebo in the typical AD clinical trial setup is the most that is achievable; gamma would then be close to this maximum.

The other points ( two purple ones to the bottom) are for the tramiprosate phase 3 conditioned on number of epsilon 4 alleles. It is somewhat less clear how to place these points as tramiprosate actually avoids removing amyloid plaque. We have noted before on thread that removing the plaque with mabs and then risking ARIA really might not be the best strategy. Trami at therapeutic dose can fully block the formation of toxic oligomers which then can have large anti-dementing effects. The purple dot to the bottom is the phase 3 result for the epsilon 44s which is simply massive at -0.79 CDR-sb (though statitiscally insignificant) and p=0.0066 for ADAS-cog; these results were with tiny n of 34. The purple dot above the bottom purple dot is for the epsilon 34s. Trami does add support for the amyloid perspective, though it is not easy to know where to place the dots on the CDR-sb vs SUVR axis. I am not sure whether PET amyloid was ever done with trami; it might not have been considered all that relevant given its mode of action. In fairness, the trami results probably exist beyond the axes given. Potentially this confusion could at some point be resolved when oligomers imaging creates a new set of axes that more fully reflect the toxic amyloid species that drives pathology. As it is now, the SUVR dimension is largely a correlated proxy for the true driver of AD pathology.

Nevertheless, both gamma entrainment and trami add further supporting evidence for the broad amyloid perspective.

The remarkably strong relationship between CDR-sb and SUVR might encourage others to highlight such a connection with their product. The 4 main anti-amyloids (Aducan, Leca, Gante, and Dona) have through their combined efforts established a near law of AD: delta CDR-sb ~ 1.27 delta SUVR. Their clinical results have helped to strongly reinforce each other and make the connection almost beyond argument. Others who have amyloid related products (e.g., gamma entrainment) might see advantage in also adopting amyloid as their biomarker. The current FDA standard is that accelerated approval can be granted for amyloid as a biomarker. That is quite a low barrier to overcome.
Last edited by J11 on Mon Jan 24, 2022 10:15 pm, edited 1 time 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 »

The CMS anti-amyloid position is unsupportable even over the short term. How could they think such a policy stance makes sense? With the overwhelming evidence in favor of anti-amyloids the FDA had no reasonable choice but to approve Aducanumab. The FDA has stayed in business for over hundred years exactly because they have consistently made the right decisions over a long period of time. They showed leadership on Aducanumab, though they will be vindicated within months and will then be seen as visionary.

For the CMS, there does not appear to be any longer term (or even shorter term) victory to anticipate. It is self-apparent that anti-amyloids are effective against anti-Alzheimer. Not covering them then appears merely as an attempt at self-dealing: take people's money and then oh, by the way we will decide whether treatment is covered according to our profit and loss statement. Even now a rationale to end-run their coverage plan seems plausible. For example, send money to an registered Alzheimer coverage plan (e.g, composed of Biogen, Lilly, Roche possibly others).

The companies could then pay "interest" in kind of whatever they liked. They would basically have the ability of a central bank to essentially create currency. I am not very clear what the marginal cost of anti-amyloid treatment is, though as a very rough guess I would go with ~$2,000 for a 700 mg of Aducanumab. It would seem there would be a great deal of latitude to be creative with the financing plans. Such schemes would be all the more popular, as from what I can see, anti-amyloids at some point could be used nearly universally. Almost everyone accumulates amyloid int their brain and this can cause a range of serious medical conditions. For as much as people like questioning the safety of Aducanumab it is important to note the dangers that the placebo arm faced in the trials because they had untreated amyloid deposits. It is quite surprising how toxic amyloid is to the brain even above Alzheimer's. Preventatively removing amyloid would seem a reasonable idea especially when maximal dosage is not given.

Even from a purely financial perspective non-coverage does not seem the best strategy. Alzheimer's already costs hundreds of billions of dollars per year (and exponentially growing) in hard cash. Helping to potentiate earlier treatment into Stage 2 & 3 Alzheimer's would off load treatment from the CMS' budget while also turning off the AD conveyor belt at the pre-MCI stage. Ultimately, the cost savings will be humongous (all those hundreds of billions of dollars per year growing exponentially). My best guess is that I am probably in this Stage 0 - Stage 2 Alzheimer region and that the treatments on the horizon will prevent me from ever progressing. All the tens of millions of others who are at a similar risk level will pay into the plan for the next many years while knowing that they will never actually receive coverage benefit all at the same time noticing others in the dementia community (who often might be their family members) being denied coverage for anti-amyloid treatment that is self-apparently effective.
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 »

Extra post. I will use this post as a placeholder as I all the following figures to be on the same page for easier access,
Last edited by J11 on Mon Jan 24, 2022 8:26 pm, edited 2 times in total.
Post Reply