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

32% to 50%.

Post by J11 »

What I wanted to highlight in these figures is how low the sample sizes will need to be to achieve significance once the responder group is defined. With aducan there were over 3,000 patients in the trials; yet with the next wave of AD clinical the patient numbers could be much smaller. The calculations directly above this post show how small. The first result was using the lowest tercile group of tau.
Only 54 of these patients would be needed to reach statistical significance in the treatment arm. As is seen in the middle figure of the first of today's series of posts to the left, statistical significance was achieved after 36 weeks. This is remarkable. The tempo of AD clinical trials could increase.

The tercile figure also shows the benefit that can be expected in the responder class. The lowest tercile in the intermediate group had a 50% reduction in AD progression. Simply by selecting the responder group the reduction in progression increases 32% to 50%!

Notice that the high tercile with the intermediate tau had close to (or possibly equal to) zero benefit from treatment. Loading these patients into the trial will simply make it more difficult to find a result (These are the non-rowers.).


This basic concept should now apply to all anti-amyloids (e.g., aducan). aducan achieved a 34% reduction in AD decline on iADRS in the 302 study. Thus, it is reasonable to estimate that it would also achieve a ~50% reduction in AD progression on this measure in the low tercile of the intermediate fraction for tau. Fractionating the low tercile might yield an even more responsive sub-group.

Importantly knowing those patients who will not likely respond would prevent exposing these patients to treatment risk.
Last edited by J11 on Thu Apr 01, 2021 7:39 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 »

aducan 302 Results.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 like posting the figures first and then commenting. That way everything is right there to see and compare. The above figures are highlighting how similar the 302 aducan trial was to the recently reported dona trial. As can be seen the CDR-sb and MMSE were almost exactly the same. Interestingly, it seems as though the ADAS-COG13 and ADCS-ADL results were flipped in the two trials. The iADRS as a result were also almost exactly the same.

The dona results seem to be a near clone of 302 aducan. It would be nice to see the ADCOMS. The dona trial can be seen as a confirmatory trial for aducan. They wanted a confirmatory trial -- well, they got it -- there it is.

While it didn't duplicate everything, it duplicated more than it didn't. The figure above shows how there is this undulating feature especially with the CDR-sb. It would be interesting to have insights from those with strong clinical experience, though as a guess the iADRS curve captures more of the true difference between treatment and placebo. The roller coaster effect shown in CDR-sb does not seem plausible. Given that ADAS-COG has 7 flavors of CDR-sb 0, the bottom figure on the bottom left probably captures the progression the best (though ADCS-ADL is also somewhat similar).
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 »

Regress 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 »

Ok, this one is going to need some explaining and a certain amount of hand wavy finesse.

The blue dot to the right of the red dot on the red line is the high dose 302 aducan's expected regression CDR-sb change.
Plausibly we can slide left from the blue dot along the red line to the red dot to 0.372 SUVR which represents the "0 level" of SUVR at 0.467 CDR-sb where traditional treatment with anti-amyloids might be expected to max out using a 78 week treatment and the standard MCI-Mild AD treatment population (assuming that such treatment is possible, even not it is the theoretical limit that could be achieved with a longer dosing period).

The purple dot is the CDR-sb dona phase 2 result. This is the generic non-selected result. The generic iADRS reported a 32% reduction in decline in this trial and the three fold selected low intermediate tercile reported a 50% reduction in decline. 50/34=1.47; 34/50 =0.68. There is a 47% amp up in iADRS results when you perform this selection. The blue bracket to the left of the purple dot to the pink dot shows how this 47% change would possibly amplify the reported dona results if applied to the CDR-sb (assumptions alert!). Pink dot is calculated to be at -0.53 CDR-sb.

The lavender bracket to the left of the red dot to the brown dot shows what would happen if a 47% change would possibly amp up the established regression line for all of the other anti-amyloids (Assumptions Alert x2!). Brown dot is ~-0.70 CDR-sb.

hmm, what is this all supposed to mean?
It's trying to give some indication in user-friendly graphical form how large the CDR-sb benefit of anti-amyloids actually could be.
From the figure above, it is largish (Remember the assumptions alert!).

So staying with the tour group, we slide left along the blue dot to the right of the red dot on the red line to the red dot.
We then teleport on the lavender bracket to the brown dot.
Alternatively, this is a slide and then a rotation.

The CDR-sb benefit has now doubled from the 302 aducan result forced to its place on the regression line.
The slide simply maxxed out the amyloid benefit which was a highly well-confirmed result from the past linear correlations.
The rotation borrowed the dona result with an extrapolation from the iADRS (which is somewhat sus admittedly).
However, this is a knowable fact; the dona phase 2 might have published the CDR-sb result for the low tercile selection,
so this keeps everything honest.

Note the scaling on the vertical axis is showing almost the entire expected cognitive decline on CDR-sb for the 78 week range (expected to be ~1.5 to ~1.7). This provides context to CDR-sb. It is no longer just the reported number of -0.39 top line result. Once you reach the bottom of the figure below ~-1.4 CDR-sb change on treatment-- That's it! AD has been stabilized! This helps add some sense of what the target is.

This is fairly speculative, though it advances the discussion to what now seems plausible for anti-amyloids given the recent dona result. The reason why there is two results given is that the dona result for some not entirely understood reason seemed to underperform the regression. An assumption used here is that the established regression could also pick up a 47% amp with three fold tau selection which is not certain to be true.

One other problem with this thinking is that there is a scaling effect involved. In the dona phase 2, the low tercile actually had lower placebo decline than especially the high tercile. Low tau tercile placebo declined by about 8, High tau tercile placebo by about 12.
So, simply being in the low tercile meant that the placebo (treatment) declined a third less. With CDR-sb this would translate into a 1.2 placebo decline versus 1.7. Less decline is a good thing, isn't it? Mais oui. Yet, it then messes up the scaling. Probably best to convert the scale to percentage decline. When the 50% less decline from the treatment versus placebo is added in, we now have 0.6 CDR-sb decline. This is actually a fairly remarkable result. Typical AD decline is 1.7 CDR-sb in these trials; if everything checks out we are down to 0.6 decline. It is potentially possible that this helps explain the dona result. The result perhaps was an underestimate because there was an unseen change of scale due to the reduction in placebo in the low tercile. I am not completely clear whether or not I have doubled counted something here.

In the figure above when I teleported along the brackets I was assuming that the placebo compression effect had not happened. Nevertheless, the reported dona iADRS was a 50% reduction in progression yet this did not make it clear that there was also a ~20-30 reduction in scale due to placebo progression reduction effect.

It also could be predicted that the actual absolute value of the CDR-sb for the anti-amyloids after a slide and rotate might not be as large as suggested, though the scaling due to the reduced placebo decline could actually restore some of the relative benefit. Thus, it would probably be best to move more to a percentage benefit reporting (though the absolute benefit is also informative).
Last edited by J11 on Thu Apr 01, 2021 8:18 pm, edited 21 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 »

Um, No users welcomed?
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 »

Dona Phase 3 iADRS.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 »

Ok, just to clarify my previous previous post.
Here the top line dona placebo had (unselected was) a 10 point decline on iADRS.
The reported effect for the low tercile was a 50% reduction in decline.

That should bring us down to 5 point decline on iADRS for the low tercile.
The reported result was actually 4.15.
This bumps up the gain to 59% for the unselected group.
There was about another 10% absolute gain due to the low tercile having a slower placebo.

Translated to the anti-amyloids (using 60% reduction in decline), this would give us 1.7 --> 0.7 CDR.
Last edited by J11 on Sun Oct 09, 2022 2:39 pm, edited 1 time in total.
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:Um, No users welcomed?
Hi J11, One of our new interns accidentally signed off for the day on your thread!
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 clarifying that.
I was worried that this thread was going to be silenced (perhaps for being too long and boring).

Yet, I think it is important for the dementia community to talk and think about aducan.
We need to make up our own minds before others make the decision for us.

J11 will not be silenced!
I haven't called for a revolution in at least a week -- so it's about time -- make it a conversation revolution!
Vive le Conversation libre!
Post Reply