Replies to post #403849 on Anavex Life Sciences Corp (AVXL)

[Joseph_K]

But I didn't ask you about the Odds Ratio. You wrote about

the unusual application of Odds Ratios and lack of primary endpoint time/response charts.

I asked,

As for the "lack of primary endpoint time/response charts," why do you feel Anavex's claims stated in this earlier post of mine fail to fill that void?

with an embedded link to that earlier post. I'll spell out what was in my "earlier post," which you seemed to have ignored:

•ANAVEX®2-73 treatment slowed cognitive decline by 45% compared to placebo at 48 weeks
•Mean difference in ADAS-Cog score change of -1.85 points

and

•[with respect to CDR-SB scores] ANAVEX 2-73 treatment slowed clinical decline in cognition and function assessed by 27%
compared to placebo
•ANAVEX®2-73 treatment difference in mean score change of -0.42 points

You aren't saying those two claims (about the 45% and 27%) are invalidated by how Anavex handled the totally independent claim about Odds Ratio, are you?

Look at the 12-5-2022 conference call, at about the 22-minute mark, and you'll see, for ADAS-Cog, that Anavex has laid out the Baseline and Week-48 scores, for the Placebo group and for the Anavex 2-73 group; it includes the mean change for each, the relative reduction in decline between the mean changes for the two groups, figures for 95% Cl (which I believe is Confidence Level, yes?), and the P-value (0.033). For CDR-SB, look at that same conference call at about the 19-minute mark.

I'm asking you, in our newly-born "Due Diligence forum," why doesn't this data satisfy your desire for "primary endpoint time/response charts"?

[sumbuysumsell]

Not the market and professional biotech analysts.

This statement doesn't seem to reconcile with the approx 10% increase in Institutional ownership. Roughly 33.5% to current 36.5%.

Would institutions not be considered part of the market and/or professionals?

[scorman1]

Because an Odds Ratio with a threshold is meaningless without the n that met that threshold.
Anavex knows the n as part of setting the threshold and doing the logistic regression, but chose not to publish the n.

I believe it was Richard who gave me detailed example of odds ratio and now your comment tweaked me....it appears that choosing -0.5 as a threshold is an arbitrary setpoint compared to p value which is absolute. Why not choose -0.4 or -0.6 and it will generate a totally different set of "n"s and different reporting of efficacy ?

I too looked for OR in Biogen's presentations which some here claimed that is why Avanex used it for comparison, but I could not find it either. Is this a red herring?

[bas2020]

Not sure why you have such a difficult time with this...
The n in the calculations are ITT: 320 (min) for treated and 165 (min) for placebo.

Regardless of the number of placebo patients experiencing a placebo effect...
The # of treated patients exceeding improvement of -0.5 ADAS-Cog vs. placebo is > 3x.
The # of treated patients exceeding improvement of +3.5 ADCS-ADL vs. placebo is > 4x.

Given that the placebo effect occurs in roughly 10%-20% of patients, then that translates to...
10%: 16 placebo and > 48 treated.
20%: 33 placebo and > 99 treated.

That's rather significant to true investors.