You beat me to it. They would have missed the 30% LL if they had fully enrolled the 8618 patients. If the efficacy data would have stayed the same, they would have failed with patients enrollment of 8618, or 16,000, or 32,000. They would have failed even if they hadn't done the early look and taken the alpha penalty.
Obviously, we don't know what attack rates would have been seen with an extended trial (a different season). It would have been a crapshoot. The efficacy for that added season would have to have been much higher than the 39/40% seen for even the expanded data.
One could argue that they made a lucky choice when they received the partial data from the early look. If they had obtained the complete topline data for the 4636 patients at the time of the early look, would they have made the decision to continue enrolling to the 8618? I think not. They missed the FDA mark by a wide margin and would have had to get great results in the next 4000 patients.
Since this vaccine has shown it "works", then maybe the original trial design was screwed up (patient numbers, end points, vaccination time, etc.). How would one design a repeat trial? Promoting one of the secondaries as the primary endpoint MIGHT help, but the original endpoint might give the best result in a trial repeat. Surrogate endpoints would be a hard sell. A much higher number of patients would certainly help (funding). The numbers of events, even with the expanded exploratory data, seems very small.
NVAX remarking that the AR was comparable to that of the early look was something that would be unthinkable for well-run companies. Heinous. When you listen to NVAX investor presentations, you can "hear" the problems with the top of the organization.
So what's happening with the EU and expanded data, or has it already happened?