Replies to post #372516 on Anavex Life Sciences Corp (AVXL)

[boi568]

Just my lay impression: The 2a data is well behaved with minimal random outliers, implying for the purpose displayed on slide 12 that a subsequent properly blinded and controlled study can safely get by with a relatively small n -- which is what Anavex did in the 2b/3.

[Steady_T]

Mayo... I think this will help.

https://en.wikipedia.org/wiki/Box_plot

I'm not sure why they would do the samples of increasing sizes. Perhaps that is a way to show that there are a few extreme values that are impacting the data analysis. By doing the multiple sample sizes they are able to show the outliers which are finally included in the last box plot.

[Investor2014]

Slide 12 is missing context as to what parameter the dispersion index refers to, but we can assume given context that it is A2-73 Exposure and Dose (PK) over time. So yes the top chart on slide 12 is the Boxplot of increasing samples of our small n=32 sample.

From the peer reviewed P2a paper SupplementaryFigures.pdf document

c. Boxplot of mean concentration groups of Blarcamesine (ANAVEX2-73) in the plasma (High: ≥ 4ng/mL; Medium: < 4ng/mL and ≥ 2.06 ng/mL; Low: <2.06 ng/mL) is linked to an improved ADCS-ADL response from baseline (ie. Higher change in ADCS-ADL). Significant p-value (<0.05) of Mann-Whitney U test. (N=24 patients having a ADCS-ADL value at Week 57)

The Bootstrapped data plot refers to this methodology: Introduction to Bootstrapping in Statistics with an Example

Bootstrapping is a statistical procedure that resamples a single dataset to create many simulated samples. This process allows you to calculate standard errors, construct confidence intervals, and perform hypothesis testing for numerous types of sample statistics. Bootstrap methods are alternative approaches to traditional hypothesis testing and are notable for being easier to understand and valid for more conditions.

If one reads and think about the full explanation in this short paper, it should be obvious that n=32, while it may be representative of the overall AD population, could also be skewed in some significant way and still internally with bootstrapping applied look as if it is representative of the overall population. Doing bootstrapping on a larger sample, say even double or certainly on 509 patients would be a more reliable indicator that we understand our patient sample is representative of the overall population distribution.

One extreme analogy I like, is that the cosmos at large is very uniform and boring, where over time entropy (disorder) increases., Looking at it zoomed in on much smaller scales like planetary systems and even smaller just our planet Earth, one finds low entropy pockets vastly different from the mean such as humans and the human brain.

Now turn to slide 13 of same presentation, the poster below and then the intermittent dosing patent for more context. Then contemplate why perhaps another Precision Medicine P3 AD trial might be needed - of course is also might not be needed, but imo less likely:





Anavex2-73 for the treatment of alzheimer's disease

Contemplated dosage regimens include administering to the subject a pharmaceutical composition comprising ANAVEX2-73 according to an intermittent dosing regimen of at least two cycles, each cycle comprising (a) a dosing period during which a therapeutically effective amount of said pharmaceutical composition is administered to said patient and, thereafter, (b) a resting period. In some embodiments the dosing period and the resting period are of the same duration or are of different durations.
Attention is brought to the dosing period and the resting period in the range of a lower limit of about 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days, 8 days, 9 days, 10 days, 11 days, 12 days, 13 days, and 14 days to an upper limit of about 28 days, 27 days, 26 days, 25 days, 24 days, 23 days, 22 days, 21 days, 20 days, 19, days, 18 days, 17 days, 16 days, 15 days, and 14 days. Also noted is the dosing period of between about 1 day and 12 days and the resting period is between about 1 day and 12 days with particular reference to a dosing period is 12 days and resting period is 12 days. Such regimen is usefully employed wherein the therapeutically effective amount of said pharmaceutical composition of ANAVEX2-73 is about 1 mg to about 60 mg and particularly about 30 mg to about 50 mg, and particularly for oral dosage forms. Also contemplated are ANAVEX2-73 dosages of about 3 mg to about 5 mg and particularly with intravenous administration.

[falconer66a]

What Is This “Dispersion Index?"

I’ve been looking at slide 12 of this presentation and can’t make much sense of it, can you help me out? If anyone else likes to reply that’s cool too!
https://www.anavex.com/_files/ugd/7b494c_66761ab69d6e4c2b905d5377eb3005f0.pdf

Page 12 of the Anavex PDF is titled, “Confirmed Reliable Inter-Individual Variability (Dispersion) for the ANAVEX2-73 Phase 2a Study with 32 Patient Cohort.”

Further, it is stated that, “Evaluation of the dispersion index of all the 32 patient[s] of the Phase 2a reveals that above 16 patients, the dispersion index is maintained at a fixed level with the narrowest confidence intervals.”

What might all of this mean? Whatever is “inter-individual variability (dispersion)?”

From another paper, here’s one helpful statement:

“In contrast to individual cognitive performance, the dispersion index is thought to provide an indicator of fairly stable endogenous factors, such as central nervous system (CNS) integrity (MacDonald et al., 2006, 2009), and to be less influenced by situation-dependent factors including fluctuations in stress or sleep (Hultsch et al., 2000).”
[From, De Felice S and Holland CA (2018), Intra-Individual Variability Across Fluid Cognition Can Reveal Qualitatively Different Cognitive Styles of the Aging Brain. Front. Psychol. 9:1973. doi:10.3389/fpsyg.2018.01973]

In essence the dispersion index in this clinical study indicates, at “the narrowest confidence levels,” that blarcamesine maintained “central nervous system (CNS) integrity,” “less influenced by situation-dependent factors including fluctuations in stress or sleep.”

The take away? Even though there were a small number of patients (only 32), the results were not randomly variable, and most importantly, were predictive of results that would emerge from a study with a larger cohort (larger number of patients). The page text makes that claim, that the dispersion index of this small study “allows for meaningful predictions for larger populations.”

As always before, the data and results from every clinical test of blarcamesine, even Phase 1 and Phase 2 trials, are accurately predictive of subsequent Phase 3 trials. The data from the big Phase 3 Alzheimer's clinical trial are going to be exceptionally positive.