I tried to post a link, but it was deleted. Post 90899 on the ihub PPHM board shows ECOG status for PPHM 2nd line. My 2 points are simple. If a control arm has 75% ECOG 1 and 2 patients, it is probable to have a lower MOS than a trail with ECOG 0 and 1 patients. That's just basic science.
Also, you cite a p-value of .2 as a failure, but fail to mention the survival curve in SNTA's interim MOS data yielded a p value of .18. So, if you choose to look at one as a failure you can't ignore the other. If anything, a data set of n=80 having a similar p value to a data set of n=43 would favor the likelihood of the smaller sample size have a better chance at success.
Thanks for making the attempt to give me the ECOG data for the PPHM Bavi trial. As far as ECOG scores and MOS, you must look at historical controls in those groups compared to the Bavi trial to get a measure of the corruption of the control groups. That is just basic science as well.
AS far as the Bavi trial, unless I am mistaken, the P=0.22 (not 0.2) mature in the Bavi trials was an estimate from a mature trial. The small number of patients make it unreliable as a predictor of future lager trials. As I recall, the 3m/kg arm has 40 patients in it, how many were left after censoring?
You are missing severals point about the SNTA trial. First and foremost, what matters is the composition of the control Docetaxel group vs the Ganetespsib + Docetaxal group and how these two groups performed. Your focus on what ECOG scores comprised the PPPHM Bavi trial is completely irrelevant to SNTA, so why do you keep bring it up here?
Second, your citing the P=0.18 for the Ganetespib trial is misleading and irrelevant as it is an early interim data report, not a statistical measure from a mature trial as was the P=0.22 from the Bavi trial.The Ganetespib data still contains newly recruited patients so of course it won't achieve statistically significance. I will point out that the HR was still a pretty good 0.68 in that group anyway. A better measure is when they look at patients who had been in the trial more than 6 months prior to that interim data analysis, and that reached P=0.056, on the border of significance with only ~40 paint in the data set, the same number by which Bavi generated a P=0.22 in a mature data set. See the difference? There are other groups that make me very hopeful.
The whole pont is that the interim data set was very promising but the numbers were small. I know we need to see more patients from a more mature data set. I am hopeful but unsure.
That R&D chief leaving when he did ijust reeks of something bad coming. I don't short cancer drug stocks. I think it's evil. But you remind me of myself prior to losing my #$%$ on PPHM. When they pulled out of ESMO for "personal reasons" I should have seen the writing on the wall. But I chose to look at the glass half full when I should have been running for the hills. I expect they'll be some positive survival benefits from ASCO in some biomarker subgroups but I'd expect no real survival benefit overall and I think the market will react violently if that's the case. Caveat emptor.