Biotech Values

[IgnoranceIsBliss]

So what do you think happened here? Why would the FDA give them a severely stringent endpoint at the second interim? ---- unless of course Kirkman is just lying and their try to buy time.

What I think is that the SPA has a very high interim hurdle. Higher than "Standard" OBF. And it makes sense considering that Merck wants to use this single Phase III trial to file for approval. And that they only would want a trial stopped if there were overwhelming efficacy.

The Merck dude described the look as "really just a futility look" today, apparently (I didn't hear it == waiting for the transcript.)

I think Kirkman is dying to guide on the hurdles, but has said over and over that Merck won't disclose, and so he can't. They're running the show.

Some analysts (e.g. Wedbush) had been assuming a hurdle of p < .001 for the 2nd look. Where did that come from? We know for a fact they're not using Haybittle-Peto. I think someone got wind of a .001 alpha assumption in the OBF boundaries, and didn't realize that would translate to a much lower p-value than .001.

That actually yields a one-sided p-value of .000005 at 50% information and .00014 at 75% information. That sounds ludicrously severe. But this trial is so overpowered that the resulting HRs required aren't crazy -- about .6 at the 1st look and about .715 at the 2nd look. The 2nd seemed tough but attainable to me. (I was wrong, of course.)

Think about it -- and this goes to iwfal's point about the real value of an HR of .8 or more. With the massive overpowering in this trial, standard OBF (p < .009 or so) would translate to an HR boundary of about .8 or so at the 2nd interim? You want the trial to stop at that level? To get approval on a single Phase III trial?

Anyhow, this is what I think, and had been factoring in going into the 2nd look. It still makes the most sense to me.

Regards,
TGW