Dendreon Corp fka DNDNQ

[iwfal]

Mouton - Note that in the 9902a document you quoted they didn't use the term "Event Driven". "Event Driven" generally means (and I haven't run across any exceptions) the trial ends when and only when the number of events meets the preset number.

While this is not strictly speaking incompatible with a pre-defined finite follow up period, in general that would seem counter productive or pointless. What happens if the event rate isn't high enough (e.g. Tax slows down the event rate in both arms) to hit the predefined number of events needed to unblind?

You'd either have to:

a) Change the follow up - except that since you had already stopped the follow up that would probably be an unacceptable ammendment to the protocol.

b) Start enrolling again - what company would want to do that?

c) Accept fewer events - and most companies have a target HR so going with fewer events would be pretty unpalatable unless you were dead sure it was the drug being overly beneficial causing the fewer events than planned.


I give it 99.9% odds that the trial protocol states the trial ends when and only when the events are hit - with the only exception being if the DSMB stops the trial early.