NorthWest Biotherapeutics Inc (NWBO)

[VuBru]

Evaluate - I agree that the company has made clear that PFS remains the primary endpoint as in the original design, and that OS (in K-M survival analysis) is the a priori key secondary outcome. Being powered for both simply meant that they had planned on running the trial until enough events occurred that results would be significant if they got the effects they expected. They actually went until more events occurred than originally intended, which in part could be an effort to offset decreased power resulting from them not enrolling the full planned sample size (related to FDA hold). All my opinion, but I don't think the company knows for a fact that the trial has failed either PFS or OS - I think they actually are blinded (suicide with the FDA if not and they know this). I think they suspect that PFS may fail, and are unsure about OS due to the crossover confounding issues. In a prior post here (stats summary), I tried to look at different scenarios and how easy/difficult it might be to address crossover. My take on it is that under some scenarios, it is not that hard to demonstrate that DCVAX works even with the crossover. I am sure the company biostats person is thinking along the same lines. I agree that the company is looking at milestone survival % (e.g., 3 year survival) as their hail mary in the event that PFS and OS both fail in formal statistical testing. All I have seen on the new FDA approach to life saving drugs, especially for orphan conditions, makes me think that DCVAX ultimately would be approved based on success of these milestones vs. SOC, even with failed PFS and OS. Despite all the gnashing of teeth here over the last couple of years, I think the company has been smart in trying to maximize the informational value of the trial to cover all scenarios. I am about 90% sure of eventual approval just based on the blinded/blended data numbers vs. SOC, because results will not get worse with unblinding, and may even get quite a bit better (several months).