Replies to post #475819 on Anavex Life Sciences Corp (AVXL)

[12x]

AVXL must tread carefully when presenting post-hoc analyses to avoid overextending the narrative. If the COL24A1 subgroup (70% of the population) shows a 2x improvement over ITT, the implication is that the remaining 30% with the mutation might be performing worse than placebo (+2.6 points). This raises a serious question: Could Blarcamesine potentially harm certain patients? Such a concern might prompt the EMA to require a confirmatory trial.

With the OLE and genetic analysis results seemingly leaked, CM is navigating a narrow window. A press release addressing these findings feels urgent, ideally by tomorrow or early Wednesday morning, ahead of his fireside chat at noon. The Q4 CC and 10-k have to come out next week to leave a window for insiders to purchase stocks before the end of the quarter black-out window starts again.

[Doc328]

Here's some amateur analysis of the new gene discovery on Stocktwits by Mayomobile and Peter Karol (posting as "Formeridiot88"):

Amateur indeed. Mayo needs to understand the difference between WES (whole exome sequencing) and RNA-Seq (RNA transcriptomics analysis). The abstract that 12x posted was WES not RNA-Seq. WES looks at all the exons in the DNA to determine if genes are WT or gene variants (i.e S1R WT vs Q2P variant or in this case COL24A1 WT vs variant) and RNA-Seq looks at mRNA gene expression. So of the 25000 genes in the human genome a couple hundred showed different responses between WT and variant. We don't see the p value threshold used (usually 5 x 10^-7 to handle multiplicity) or definition of 'trait' (in this case response definition). Of course if COL24a1 did better than mixed WT and variant, then the variant would have done worse, maybe even no response to A273.

Peter takes an optimistic approach to imply neurogenesis due to a differential response to A273 based on COL24A1 WT vs variant. Gene expression of COL24A1 can be found in some neurons but is not necessarily linked to AD (though COL25A1 is!). There actually is not too much known about .

Data like this is hypothesis generating. A different cohort will be needed to confirm the association ---- perhaps the phase 3 study will pre-specify the 40% of AD patients who are S1RWT and COL24A1 WT.

[Hoskuld]

Need the basis for that "dosed group" claim instead of ITT.