Biotech Values

[iwfal]

AGIX - Correction to review of blended ARISE results:

In doing some checks discovered that there is no way that the blended MACE curve that they released is consistent with the fact that they have had 1000 events (out of 6127 randomized).

A summary of the issue is that there are high odds that with the patient population seen in ARISE we will see an event rate from 12 months on of >6.25% per annum in the placebo arm. The blended ARISE curve shows an event rate of <5% per annum. Very promising. However a better guess (based on other data AGIX has released - including the fact that they hit 1000 patients evented) is that the event rate after 12 months is closer to 6% per annum and that isn't too much better than the worst case 6.25% per year for placebo. (Note that all of my worst case comparisons to other trials were always at least 10% too conservative, so this is not a death knell for ARISE, just a lowering of the odds)

The risks of using data from a company known to be sloppy.

Note that I am sharpening my pencil, but an oversharp pencil is not necessarily a good thing. Can cause self injury.

Clark