I can’t imagine that they did not have a draft SAP based upon the original terms that, at the time of any efficacy IA they would then file, but the court determined they never did the efficacy determination. It does not mean they “could not decide” to do one, but they’d have had to have filed their SAP and apparently, per Senti, they actively decided not to file one, and the reason to do that would be because one was still not sure that one had the right approach and might want to decide if one wanted to go with a surrogate of survival or survival. And of course, then the partial halt came out of nowhere, I believe unexpectedly and I believe without full explanation. No doubt they were hearing about other trials, they had already run or started their compassionate use arm for pseudoprogression patients, which was not blinded, with excellent results, and then they published that info.
So I think the notion that they intended to do one thing, then changed their mind, being a “lie”, I think it was just a fluid situation, not a lie but something they thought they could do to end the trial early, and then they decided that the better measure should probably change at some point and that they probably needed to go through the full trial and get survival data not just progression.
Plus at some point they likely discovered that their placebo arm was disappearing as so many were crossing over because of the apparent excellent results.
In that context HOW do you do an efficacy analysis? You know from the safety analysis that things are good, you know you have a pseudoprogression “issue” but even those patients are doing really well on survival, and you know your placebo arm is non-existent…. I think you delay until you figure out the best way forward, you consult with your SAB and your regulators. You may have plans and until you decide exactly what you will do you presume your plans remain unchanged, but delayed.
I think they were figuring it out, not lying. This is why there is a future statements safe harbor.
