North, re SD numbers
One would expect that data to be consistent with the TTP data. And that is what we see, the Bavi+S number is 53% while the S number from Sharp was 43%. So progression was somewhat slower in this trial than in Sharp.
I don't see how this really changes the story much. If we consider the trial comparison valid, one has better progression with Bavi added, but OS was worse (and that is true even if adjusted of PS status). If the trial comparison is invalid, then what exactly are we looking at? That single 18 month patient means nothing (Sharp almost certainly had several).
[BTW, I am not claiming that Bavi actually impairs OS.]
The only way Yopp can make this look better than "meh" is to prove his statement at the end that the patient characteristics were poor. But that has not been done, and is fairly hard to get more than "possibly".
BTW, Opdivo had trials stopped early in BOTH melanoma (2nd line, after Yervoy failure) last fall and squamous NCSLC, just now. Perhaps your sleepy mind was trying to put those facts together :-)
There are several upcoming trial readouts for them that directly overlap Bavi trial space. 2nd line non squamous NSCLC (vs doce) and 1st line melanoma (in addition to Yervoy).
The game is fast changing.
Market Data 
Markets 
AI
