NorthWest Biotherapeutics Inc (NWBO)

[flipper44]

Fox,

I'm tired, but I promised to write something tonight.

(Again, I refer you to Longusa's recent post demonstrating why NWBO has multiple advantages above and beyond just statistical significance.)

Here is a link providing an example where the DMC (AKA: IDMC) was scrutinized for early termination for efficacy -- where one of two primary endpoints were not met.

This is not anything I expect to see in the DCVAX trial. I expect an 'across the room' obvious result. However, the link below demonstrates that there is more to the IDMC than statistics.

http://seekingalpha.com/article/639811-will-the-fda-dance-to-j-and-js-tune-on-zytiga

Here is a passage from the above link:

In sum, then, there are those who believe the IDMC had good reason for recommending termination, given the study had achieved stat sig PFS and the survival data were reasonably good even if they failed statistical significance.

That said, about 50% of the patients were known to take subsequent therapies such as Taxotere or Provenge by the time the interim analysis was performed. So, there now is a question how these therapies impacted the trial. The 18-month overlap of the survival curves might have been an indication these treatments were as effective as Zytiga during that period. Then, too, even though the curves managed to open up, now that the trial has been unblinded, patients could overwhelmingly start secondary treatments, including crossing over to Zytiga. This would have the effect of causing the curves to close. Put another way, we may ask this question: can the trial achieve statistical significance for OS if many of the patients cross from the placebo arm and take Zytiga? These are but a few of the many issues upon which the debate regarding FDA approval will almost certainly be focused in the months ahead.

It is not cut and dry. Certainly in the case of Zytiga, the DMC decision was highly controversial because the treatment did not achieve both co-primary endpoints. I do not expect anything of the sort with NWBO. Many moons ago I referenced the 33 patients that are still included in this study today. I am still convinced this anomaly is legitimately and ethically advantageous. Why? Because normally in trials they allow a lot of patients to die in a control arm simply to wait to see how long people in the experimental arm live (because if a treatment works, especially if it is terrific….it takes a while to get data points and meanwhile more control patients die). However, by a series of unexpected events, with NWBO, they are able to get very mature data before too many people have to demonstrate disease progression and/or die in the control arm simply to prove statistical significance. I anticipate the DMC will have a much less controversial decision to make with NWBO if the primary endpoint is met.

Also, here are the FDA guidelines for the DMC.

http://www.fda.gov/RegulatoryInformation/Guidances/ucm127069.htm

As you can see, there is a communication filter between the three groups; however, it allows them to cooperatively work toward the truth, without sacrificing statistical analysis, their separate responsibilities and independence.