Biotech Values

[iwfal]

FGEN

I guess the obvious question for me is, regardless of efficacy, how do you know the trial could have been stopped? We don't know any specific stopping criteria, or even if any exist, right?

Per the same CC from which I excerpted the quote, there is no interim. But if the OS had a strong trend at 150 man-years, and if the same efficacy repeated with 20 or 30x more patient-years the p value would be ridiculously low. The DSMB would have no choice but to stop the trial even if there was no interim and no assigned alpha. This has happened before, although I cannot, offhand, remember the particular drugs. This do no harm aspect is the whole purpose of the DSMB.

Note that, as I suggested in the post to which you are responding, there are lot of other possible explanations. All I can do is rule some out and provisionally suggest some as most likely.

And note, as I said in a different post, if the HIF drugs have a spike in events before driving the event rate down (and that is highly likely in nephropathy - early hyperkalemia, later benefit in GFR) then they won't see a good p value until x months after they stop enrolling.