Cel Sci Corporation (CVM)

[lightrock]

SeaAlice,Trustbaby, and Spidey

Ok... but... needle in a haystack though.

For example, in the 10% OAS scenario we are looking for a differential that amounts to approximately THIRTY out of the entire populations of test and control ( minus the auxiliary arm )

So finding futility is about as hard as finding success.

i.e., you would need about 298 events worth of data to decide "futile" just as much as you would need 298 events to decide "success".

Unless, under the microscope, there is no discernible difference in tissue samples after about... what.. 100 ?

But the study is worded OVER ALL SURVIVAL which may or may not show up clearly in tissue samples.

I just don't see how the IDMC can possibly say anything other than "proceed" after all injections are already completed.

That one time when the FDA used the word "futile"... that was interesting... but I would chalk that up to the very same problem of not understanding the statistics. "signal" doesn't even appear until near the end of the study.