In the case of an ongoing study - that is to say, one which does not have a pre-defined study termination date - who/how/what (????) legitimatizes the opening of the study for an 'interim' peek? IOW can interim simply be some negotiated, arbitrary point in time or is there some regulatory “if this then that equals thus and so” about looks?
Interim peeks are per protocol. The FDA frowns on peaks that do not have a specific purpose so the protocol peaks are typically done for p value checks although other purposes could be the justification (e.g. Baysian adjustments). Peaks by the company outside of the protocol would be very bad mojo indeed. As for how the protocol is decided upon, that is mostly a decision of the company although I suspect that in some cases the FDA may suggest dates etc.
(Note that there are cases (e.g. 9901 Survival) where interim peaks are not 'per protocol' but I believe that they only occur after the primary endpoint and cessation of active treatment.)
Taking "insider trading" into account and since Bogdan's out of pocket 46K purchase was made on 6-15-06 - is it safe to assume nothing regarding the P-11 was known up to that point in time?
No. Comapanies almost always have lots of blended/blinded data. Never conclusive, but I suspect quite often strongly indicative IF THE PERSON LOOKING AT IT IS SMART ENOUGH NOT TO BE FOOLED. (This is key - too many biotech executives buy their own hype and suck at data mining) E.g. Ocyan's hypothesis that Dendreon knows when people came in for active P-11 boosters. Or the company might have other blood markers for those booster patients vs the 'boosters' for placebo patients (e.g. PSA Velocity). We have access to 10 pieces of data - Dendreon has access to 100's of pieces of data (none conclusive, but still a lot more info than we have)
is it safe to imply that since results for P-11 were expected to be in by sometime 1H06, but the trial was still ongoing at 6-15 –we can say P-11 is doing better than expected?
What is 'expected'? Biotech execs often extrapolate the event rate during active enrollment to the project when the requisite number of events will occur. But as soon as enrollment stops the event rate starts tailing off (think about the fact that the survival curves are asymptotic). If that is the basis of the error in projection then the error in projection means nothing whatsoever. OTOH if the projection was based upon historical times to event then the error is indicative of something - but this is rarely, if ever, the basis of projections for trail completion made at the time of enrollment completion.
As for whether the unblinding in Nov or Dec means something, see my multiple posts on the subject. Bottom line - if completion of 'enrollment' in Jun 05 means all patients randomized, then yes, there is a good chance of stat sig (IMO). If it means just elligible to start LHRH therapy then a Nov/Dec unblinding probably means only that the chance of stat sig is a coin toss.
Is it possible the P-11 study has already reached conclusion and Dendreon is “sitting” on results as part of some timing in a promo campaign? Is a company obligated to publish good news as soon as it becomes material to them?
My operating assumption is that it is rare when companies sit on unblinded data for too long - the longer they sit on it the bigger the chance that someone will leak it and they will get sued.
All just MO.
