BMY / CheckMate 227 — I was and still am flabbergasted that the PFS and OS co-primaries are in different biomarker subpopulations, and that the “non-Yervoy” Opdivo arms (monotherapy or chemo combo, depending on PD-L1 expression) are totally absent from the primary analysis plan, apparently. What a mess.
That said, I’ve also seen chatter on Twitter slamming the trial for being open label, as if that means the company is making all these statistical amendments post hoc after looking at the data, since it’s not blinded. I have a strong feeling that BMS’s data handling was more rigorous than this, and it’s a bit of a red herring to decry a trial like this for being open label. It’s very hard/practically impossible not to have an open label trial when you have infusions given on different schedules between different arms. Worth noting that KEYNOTE-024 was open label, which appropriately caused exactly zero controversy. Many other IO trials are open label, and this isn’t necessarily a design flaw.
Market Data 
Markets 
AI
