Replies to post #160955 on NorthWest Biotherapeutics Inc (NWBO)

[CherryTree1]

Yes, this is similar to the calculation I did in my seeking alpha article several months back. However I focused on PFS which showed even more convincing results.

I would be interested in seeing how you did this with PFS Barunuuk. Is there PFS data NWBO is sharing?

The only comment I have is on your discussion of NWBO waiting longer. In your examples the p value doesn't change the longer you wait, however, the difference in your examples and the P3 trial is that the p value will not be calculated on SOC data, it will be calculated comparing the 331 patients, meaning the SOC data will be the 110 placebo patients vs ~221 vaccinated patients. You recalculate the p value using your assumed results and 110 SOC patients, and the p value falls down dramatically.

Yes agreed and that was part of the reason I spent the time doing this. There are a good number of posts prior to mine that alleged that we will never have enough events to have a statistically significant difference between 221 DCVAX from day 1 patients and the majority of the 110 that started with placebo and switched to DCVAX after progression. I wanted to show that if they simply compared the whole group of 331 patents most of which received DCVAX at some time we could demonstrate statistical significance with SOC patients

Moreover, because it is a cross over, if DCVax-L actually does work, then the 110 patients will perform better than SOC, meaning that will drop the p value even further, SO, it goes without saying (if you are well versed in Statistics), that in this case, the longer the data is matured, the more chance the p-value will drop, meaning the fatter and longer the tail of the 221 patients compared to the 110 patients, the smaller the p value will be.

This was kind of interesting on P values smaller than 0.0001
https://www.ncbi.nlm.nih.gov/pubmed/15080563

"OBJECTIVES:
To explain from hypothesized and published examples why extreme p-values like p>0.95 and p<0.0001 may indicate that sampling was not completely random."

so there are those that would argue one really wouldn't want the P value to be smaller

Hence, why Linda and NWBO is pushing this as long as possible.

I my mind, I have no doubt the PFS endpoint has been met, and is statistically significant, however the OS endpoint may not be as strong statistically if the patients that crossed over are living longer as well.

I had a mistake in my orginal spread sheet. Here is a link to my post with the correction. It didn't really make a difference in the conclusion, but it was a bit smaller difference.
https://investorshub.advfn.com/boards/read_msg.aspx?message_id=138858295
Thanks for the comments and feedback Barunuuk