Biotech Values

[poorgradstudent]

dndn

>But, as a way of guaging your beliefs, would you take a bet on it's being stat sig using its predefined analysis? I'd give 90% odds of that. And you would give...? (And I'd give virtual >95% odds of being on the right side and p<0.3)<

Not being flippant, but do you seriously think we'll find out what the exact pre-specified analysis is for the cox primary endpoint? They *ahem* misled about the "prespecified" cox analysis in 01 / 02a, so i don't expect any transparency on the 02B endpoint unless it is changed to a log rank as the committee seemed to suggest.

If log rank, i bet HR is less than 1.3. If you want a number, i'll go 1.21. If I'm not entirely mistaken, an HR of 1.21 in this 500 patient trial won't be a guarantee to come in at p < 0.05.

I'll agree with you that it will numerically come out HR > 1, but that is par for the course for many failed therapies.

>What caused you to change your mind?<

Absolutely no immune response to a physiological antigen. No reasonable MOA explanation (i don't concur with David's suggestion that it "picks" away at specific cells in vivo). No objective responses in 01/02A to reproduce the single instance in the phase I/IIs (perhaps they're simply not reported?). No other endpoints ever meeting stat sig to support the finding that the survival finding is a bona fide treatment effect. The continued data mining early on which suggests the company had no expectation for an ITT survival benefit, let alone one at HR 1.7. David was also kind enough to tell me a long time ago that DNDN removed the PAP antigen pre-screen from 02B... at the time it was confusing, but now it makes total sense: there is no point.

I got snowed by small's first jco paper and the sleight of hand he used in discussing the findings. I thought it was a weak immune stimulant, not an inactive one. I should have known to read a clinical journal more intently with regards to their laboratory findings. That was absolutely careless on my part.