Anavex Life Sciences Corp (AVXL)

[Anshu2]

The problem is not entirely with the 0.063 p number. I have heard FDA say in an AdCom (BMRN's Prosensa drug AdCom in 2015-2016, if you want to look it up) that they do not have a hard line of 0.05 and are willing to approve drugs that have p higher than 0.05 (e.g., 0.07 may be acceptable). BUT, the problem with 2-73's drug is that there is NOTHING ELSE that supports clinical efficacy claim -- EXCELLENCE's other endpoints failed outright (perhaps, p > 0.2 and/or the DELTA was insignificant), AVATAR failed (AVXL had to post-result resort to AUC to save its face). So, looking at the evidence (of any credibility -- i.e, you can only look at PRE-SPECIFIED endpoints of placebo-controlled rails) overall -- the evidence is very very thin, and in all likelihood the drug does NOT WORK (in terms of its impact on RSBQ and/or CGI, at 7-12 weeks period).

One can speculate of reasons behind the failures -- even if the reasons are plausible valid, that doesn't prove the drug will work in absence of those reasons (e.g., placebo benefit).

So, there is no way FDA will be convinced of drug's efficacy from the data (nothing else matters to FDA). IN FACT, even if a future trial succeeds -- AVXL will only have a lot of explaining to do, as to why these previous trials failed (may not be hard to do, but AVXL will need to).

FDA doesn't approve based on hunches/speculations. It needs to be convinced of the evidence (even if the p value is over 0.05) that the drug "works". E.g., recent SRP-9001 trial failed on its primary endpoint. I am (and, the SRPT's CEO and many analysts) are still betting that FDA will still approve the drug (i.e., at least expand the label for AA which has a lower bar on functional efficacy, but also do a full approval) --- because the overall evidence is quite CONVINCING (of course, FDA needs to feel that too -- time will tell).