Reading over the posts past couple days, a few comments/responses/perspectives to several random subjects brought up. Don’t look for cohesion here. Thoughts to provoke thoughts. Possibly a sports analogy or two, we’ll see:
In no particular order.
Anavex Life Sciences (main topic)
PDD stat sig results
SIUF PR
Rett PR (the onside kick?)
AVXL PDD vs LLY PDD hmmm
Quickest path to mkt, imo
...well, that should get us started
PDD stat sig results -
Noticed that Gernee and Investor had a nice thread going speculating about results and the need for a ph3 if top line results were not stat sig but the genetic variant subgroup secondary endpoints were proven out.
My two cents: Gernee, although Investor is correct in the trial protocol needing top line to carry the day, let’s look at some math (not a lot, just a little) to see why this is still favorable for Anavex and patients and perhaps why Dr Missling made the statement to a shareholder (reportedly after last non virtual ASM or prior, can’t recall) that, “We were so relieved when we got the (KEM) correlations.” (Paraphrased but that was the gist to my recollection)
I now believe that they were “so relieved” not only because they could account for the super responders to the drug and power up a trial (eventually), but because the genetic variant occurs in 80% (emphasis) of the population. That’s the relief part. No need to power up a trial and know that your drug is in play for only, say, 30% of the general population (assuming a lesser degree of occurrence of the responding genes). That’s great news - why? - because 80% of people (normally distributed with AD/PDD/Rett/ CNS diseases) will respond to the drug at sufficiently high blood concentrations if our correlations hold. That means, or should mean, that within a placebo group, randomized trial, 80% of the dosed group should respond to the drug. Top line, that alone should achieve stat sig. Here’s that math we were referring to:
Originally 32 AD patients took part in a non placebo open label trial in Australia begun 5 years ago. Of that group, we know ad hoc - after the fact, that 20% or 6.4 of those should NEVER have responded to the drug at all due to genetics - assuming normal distribution. The end.
Round down and subtract 6 patients who represent the 20% non-responders. Leaves 26 patients.
The trial had 7 “super responders” at the 57 week point and that was prior to everyone being at max tolerated dose (MTD).
7/32 is greater than 20% and 7/26, 26 being the number who should have had a favorable response, is greater than 25%. Still not at MTD for all patients and not at highest blood concentration for best response. Additionally, other patients besides the “super responders” responded favorably, just not as favorably.
After studying the original patient group in open label and adding gut microbiome data to the mix, one could estimate that the response rate could be higher than the 26.9% 7/26 which occurred without any prior knowledge. How much higher? That’s the question (Missling already knows the answer, btw, after 5 years of observable tweaking). And if the answer is that after 5 years of studying trial and error and testing input/outcome strategies, we could edge that up to 35% - 40% total favorable response in the super responder range: such as drivers exam folks, and the climatologist still blogging very knowledgeably (one company slide stated that a patient had a 2 fold chance of slowing the decline in MMSE score against a historical database of normally declining patients if on the drug at the 147 week mark), so we know others did well enough to keep the average performing significantly better than a “mock placebo” group - if the answer is 35-40% would that be significant to reject the null hypothesis of random chance?
150 patients in PDD trial. 50 placebo, 50 on 30mg, 50 on 50mg A2-73. Total dosed - 100 patients. Expect 80 to respond based on normal distribution of 80% WT genes in general population. Of the 80 patients who should respond based on genetic match, if 35% - 40% got to the necessary concentration and responded, then 28 - 32 patients out of 100 performed favorably on the drug and improved whereas zero on the placebo should improve. 28 - 32 (avg 30) out of 100 vs 0 or possibly 2 or 3 (if placebo effect) out of 50, should statistically allow the company to reject the null hypothesis that this was random chance - and that’s not even taking into account the fact that the denominator should be 80 not 100, as 20 were not expected to respond as prespecified. Still, 28-32% of total population is roughly 30% of the time, a diseased patient who takes a medication will improve.
And here’s the mathematical catch - why it would be nonsensical to do a ph 3 with all genetically matched responders: it doesn’t raise the outcome very much statistically. If you stack the trial with 100% of WT patients instead of only 80%, ONLY 80%, you are not going to achieve 100% improvement over placebo any more than you guarantee 80% improve over placebo with the normal distribution. It’s the percentage of those 80% who carry the gene and to what degree they will respond which determines our stat sig. An extra 20% will not overcome an inefficacy rate which 80% cant overcome. This simply means that if 80% already have “what it takes to be ideal candidates” and they don’t succeed, 100% won’t either - the null hypothesis remains valid in my opinion and the drug does not have efficacy.
The math will have spoken and Dr Missling said let the data speak for themselves. Imo, either the ph 2 will succeed and be stat sig good enough to reject null or they will not find evidence to proceed to a Ph3 trial to simply stack 20% more WT in the trial and run it again. I think that would be futile and a foolish waste of resources amounting to no gain. This is it, imo, all or nothing. This will pass or fail but not inform to design another step - if others disagree, please specify what they can do to meet the primary endpoints if 100% in ph3 where 80% fails in ph2. I’m already allowing for the fact that we are comparing the entire total in the denominator including those not expected to respond. If 30/100 isn’t good enough to reject null/ random occurrence, how does 35-40, so 37/100 get you where 30/100 doesn’t? If spending the time and money to re-screen and enroll again for more precise precision will convince the EMA due to 7 extra responders, I’ll be a monkey’s uncle. That’s so broken it’s obvious. And we can’t declare the gene variant a primary endpoint because all enrollees will have them. Do you see the futility? If we can’t prove it with an 80% head start, it’s a no go, imo.
SIUF PR
3 things I noticed in the PR which cause me to entertain that this is actually a ph4:
The phase is not spelled out in the PR. So, they are funding a trial and announcing it but do not disclose the phase. The blurb which Kentucky123 posted (Thank You!) described Clyde Campbell and his brother as people who can’t just wait but would rather make things happen! As others have described the Aussie “Get er done” spirit and mentality. Why would these “doers” leave that out??
Second, the Anavex PR mentioned that this trial would be an attempt (in conjunction with the 2 foundations) to get the drug to market as quickly as possible. Yeah - a ph2 or ph 3 over 48 weeks duration along with enrollment and possible follow on phases isn’t going to do that. That will truly take years - plural.
Third, the trial states the duration, even though it does not state the specific phase?, the duration is, and I quote, “not less than 48 weeks”.
According to Xena’s research on PA and Ph4’s, the drug is usually monitored between 1-6 years and approval can be granted anytime in that window if the TGA sees convincing results.
I say that, “Not less than 48 weeks” meets that requirement of describing how long this trial could last without full knowledge presently of a more exact timeframe to commit to. Try reading it that way as opposed to a “fluff” piece of distraction, or imaginary “wonderfulness”, so as not to get blindsided by the oncoming train. Sometimes (often) words mean exactly what they say. It’s our own bias against the CEO, the outcome probability, the long wait, too much risk that went awry, these negative biases can prevent us from reading a couple of blokes trying to tell you what they know and they will give more details when they are known. Must be secret fluff conspiracy from a guy with Parkinson’s(!!!) to distract from the fact that more misery is coming / the latest hope for a cure just caught the last train outta town so we decided to shuffle the playlist and randomly sponsor a trial to generate some fluffy excitement to fool people.
Is Rett the onside kick?
Well, if Rett is approved first - in Australia, pediatric through PA or AA, because adult trials won’t do it, for some inexplicable authoritative reason, then, adult Rett and PDD and AD PA are NOT pipe dreams, I would argue. This is because, based on imatinib - a poster child for all leukemias/myelomas, the regs of EMU and FDA both have clear provisions whereby once a drug has been approved for an indication, it can be approved for another indication based SOLELY on extensions of RCT’s. Well, we got those. So, once the ice breaks and A2-73 has approval, if PDD meets secondary but narrowly misses primary endpoint, can still be approved full stop based on whatever additional data the OLE bear out. If approved for PDD in Europe or US, AU has a much higher probability of granting SAS because it has already been approved elsewhere and that’s the main source for the SAS approvals. So, the PDD folks would likely get SAS. And the secondary endpoints of motor function would let Clyde Campbell have his doctor applying for his prescription as well. Much quicker than waiting for the couple years to pan out. Meanwhile, the P4 is ready, funded, and waiting to go for full approval in 48 weeks rather than the full length of the red tape. Get it? Like bootstrapping...And can AD be far behind? Who knows - but I hope Barrie Pittock weighs in at that point.
I believe this would be the quickest path to market for every indication. One approval, then extensions.
Jj was asking about Anavex PDD trial vs LLY PDD trial. As to the Lilly PDD trial, here’s the low down: your friends wont tell you but I will.
Lilly is conducting a mixed trial on 2 types of dementia using the same endpoint for both and it’s the same endpoint we are using for our PDD. Roughly the same trial duration and very close trial completion dates. The differences are: LLY is trialing the drug on PDD patients but also on LBD Lewy Body Dementia patients. It’s a drug which has an MoA of “maximizing the effect of dopamine in the brain”. But, as you know, in Parkinson’s and LBD, dopamine is the neurotransmitter which declines rapidly. Therefore, even if you maximize the effect of what remains resident, it will continue to decline unless you can stop that which they make no claim to. I think they are hopeful they can provide a little more time for patients without changing the outcome in the end.
However, what would be significant would be if LLY, through this trial, showed a correlation between PDD and LBD. That would help Anavex because we have a drug which works upstream from dopamine already A2-73 which, if it works on PDD, could be a very short jump to show that 3-71 would help with LBD (same MoA). LBD is much more aggressive and 3-71 is more potent and in safety trials now. I’m intrigued at the possible intersection of information gathered by these 2 companies simultaneously and how that knowledge might be used strategically, perhaps for both companies? For both companies...that’s a hint. Both companies are running a little late on presenting results. Maybe they need to find a time which works for them mutually?
Nah, that has an inference of “wonderfulness” in it. Scratch that. Always remember, go through life with no strategy or planning. Things will just happen - either by doing them in haphazard fashion and seeing the result, or going to your basement at midnight and performing wizardry and magic. There is nothing in between these two extremes such as planning out a strategy - don’t kid yourselves. Stick with: just try things with no rhyme or reason and succeed or fail or believe in magic (but that’s illogical, just like “strategery”...strategy is just a fancy way of saying wizardry.)
Sorry, there’s no way to tell a kid who saw the immaculate reception live not to believe in hard work and strategy. Accompanied by miracles!
Biostock
