Replies to post #138352 on Biotech Values

[iwfal]

ONTY -

Also Kirkman gave out enough info on trial design (90% powered 6 mo MST advantage over control arm MST of 20 mos, HR .76-.77, and overall p value .025) to know that Merck designed START with .025 one sided under assumption of PH.

Hopefully that puts to bed the weird assumption that 0.025 was single sided. (BTW - I confirmed your calcs that HR=20/26, 90 percent powering, and 0.05 traditional two-sided alpha align with 700 events.)

[IgnoranceIsBliss]

I'm not following the logic. Are you saying Merck didn't want to stop at the 2nd interim if the HR is .8 or greater because it's not clinically meaningful, but that it would be OK for the final with just an improvement in p value due to higher n?

I don't think Merck ever expected an HR over .8... I think they designed the trial so that it would only stop early if there were overwhelming efficacy.

In a sense, the stat sig hurdle at the final analysis is irrelevant. The FDA isn't going to care whether the outcome is "stat sig" regardless of clinical benefit considerations. If the p value at the final analysis is .0215 (stat sig) and the survival benefit is 4 or 5 months, there is very good chance that Stimuvax wouldn't be approved without a confirmatory trial.

This last point is pretty much iwfal's point.

Merck wanted a trial that would show long-term survival and would have robust enough results for approval with one Phase III trial. Whether it will do those things is still unknown.

Regards,
TGW

[Summer2762]

Can it be OBF with Lan DeMets spending?

[I did not follow ONTY closely to check if the above is a plausible question]