Replies to post #34891 on NorthWest Biotherapeutics Inc (NWBO)

[flipper44]

Nice find. I think that recent example with Bluebird Bio sounds exactly like what I was discussing that FDA guidance clause allowed for. The question is what goes on between the DMC, LP/NWBO and public communication if such an interim analysis success occurs allowing NWBO to proceed with AA while maintaining a blind. Is a halt still recommended? Is it bypassed? If by passed, is the public told? If the DMC recommends a halt, I think NWBO has to communicate it. So what is the mechanism beyond what you found? Regardless, nice find, I really appreciate that.

[flipper44]

Koman, I think the key to resolving my consternation over the way a blind would be treated under the previously discussed scenario before a recommended halt versus after a recommended halt, is that I must keep in mind the halt recommendation represents reaching the primary endpoint, if the primary endpoint is reached, there is no mystery regarding this because it must be announced if the halt is recommended. That we all know the trial is a success no longer must be hidden because the statistical endpoint is reached. Therefore, even if the blind is continued, it is done for the separate purpose of approval.

In other words, because, under this scenario, the treatment group's pfs advantage was proven over placebo, the only thing left to prove after the primary endpoint was reached and by the time accelerated approval arrives is how large that advantage proved to be. The blind normally does not remain in such a situation, but under one specific clause, the (less strict) blind can remain in order to arrive at/reach the median pfs (in this case) in the treatment group by the time accelerated approval happens.

In other words, a very strict blind is required to disprove the null hypothesis, but a less strict blind (aka: one where everyone knows the treatment has some efficacy) is required to demonstrate the full extent of that advantage.

In layman's term. The blind is more strict when one is initially trying to prove it works; after that, if you need to keep the blind for/until accelerated approval, the blind is less strict.

If this reasoning is sound, then a company can announce a halt recommendation from the DMC and simultaneously move forward to accelerated approval while keeping the blind -- albeit a less formal blind.

[flipper44]

Thanks again Koman. You wrote this in May of 2015, long before Dr. Liau's October presentation.

Hey Flipper, this quote is for you in regards to your concern about what happens at the IA for the dcvaxL pIII trial if there is a halt for efficacy but the trial may continue and what happens to the blind etc that you posted a short time ago where I told you it shouldn't be a concern- at least to me. Remember?

"The phase III study will enroll 2,500 patients with NASH and liver fibrosis. The patients will be treated with either a low and high dose of OCA or a placebo. After 72 weeks of treatment, an interim analysis on 1,400 patients will be performed on co-primary endpoints: first, a decrease in liver fibrosis by at least one stage; and second, resolution of NASH with no worsening of fibrosis.

The efficacy endpoints to be used in the phase III study are similar to those used in Intercept's successful phase II study. In that older study, OCA demonstrated a statistically significant reduction in liver fibrosis compared to placebo. OCA also improved NASH resolution compared to placebo, but the benefit was not statistically significant.

Intercept is not being asked to conduct a cardiovascular outcomes study of OCA, alleviating a concern raised by investors.

If the interim analysis is successful, Intercept will be able to seek accelerated approval in Europe and the U.S. The company, however, will continue to follow patients in the study on a blinded basis until a pre-specified number of liver-related clinical events, including progression to cirrhosis, occur. The phase III study is expected to begin enrolling patients in the third quarter."


I don't know if DCVAXL has something like this predesigned into their pIII trial but that is how I see it being played out IF an interim analysis shows meeting of a prespecified data point that would allow AA until the rest of the trial when finished confirms or not the OS advantage. I don't think LP ever mentioned anything like this, though.

In other words, under that design, Dr. Liau's October speech might be interpreted more your way, and it would explain our timeline flip flop with each other. Maybe. That language in red above reminds me very much of the language she slightly struggled with regarding "predetermined number of events."