Anavex Life Sciences Corp (AVXL)

[Doc328]

Missling is the best Texas sharpshooter... I can see him now in his cowboy hat with a gun and bucket of paint. Maybe his hat says p-hack on it.
1. Use GWAS to find a poorly characterized gene best associated with a good outcome (COL24A), then reanalyze the data and say patients with this gene do better.
2. Write a PR as though prodromal is normal aging. Prodromal is CDR-G 0.5 (clearly spelt out in reference 3 of the PR). This is commonly referred to Mild cognitive impairment, not normal aging. In the 2b/3, 65% of the treated patients in the 2b/3 were CDR-G of 0.5 and in the placebo group 57% were 0.5 (just part of the mismatch that placed slightly more advanced patients that don't progress much in the treated group vs placebo). Surprise, a subset of group A that is mostly comprised of subgroup B behaves similarly to subgroup B. How did "Cognitive outcomes observed in the oral blarcamesine 30 mg Precision Medicine cohort move toward normal aging profiles" get past the lawyers
3. So are we combining 30 mg and 50 mg or what -- in order to get the MMRM LSM to be significant (for the primary endpoint that did not sort out genes), Missling needed to combine the 30 and 50 mg and handwaved that they were actually the same as few patients actually stayed on 50 mg. Now we are separating out 30 and 50 mg --- did this additional data worsen the p?

Classic Missling.


Reference 3 of PR: https://www.sciencedirect.com/science/article/pii/S2274580724003844 "clinically defined prodromal populations (CDR-GS = 0.5)"