Replies to post #3445 on NorthWest Biotherapeutics Inc (NWBO)

[ou71764]

gpb, I need an advanced statistics class - not to mention more coffee - to follow all that, LOL! But I appreciate you taking the time to explain where your numbers come from.

I'm not always certain if your numbers are bullish or cautious. For example, you estimate that the control group has a PFS of 9 months. In terms of proving the efficacy of DCVax-L, the shorter the better. Not better for the patients, of course. But statistically, the closer the control group is to the historic norm of 7, the better the odds are for DCVax-L.

But you offered your estimate of 9 in contrast to Flipper's estimate of 12. Since 12 would be less favorable than 9 in terms of early approval for DCVax-L, and Flipper is the most optimistic person here about the odds of an early halt, it seems contradictory. I'll let Flipper respond, but I thought he was estimating that the PFS of the control group would be less than 9.

You said the odds would be twice as great for a halt at the 2nd interim over the 1st. Can you give a number - is it 5% at the first and 10% at the second? Or 25 and 50%?

[flipper44]

GPB,




Nice review with many detailed resection studies. The difference between the DCVAX-L study and those studies is this:


1. DCVAX-L Study: "Patients must not have progressive disease at completion of radiation therapy. Patients with suspected pseudo-progression will be enrolled and analyzed separately." Clinical trials.gov.

In other words, while I believe your numbers regarding PFS for control groups are very reasonable if patients were filtered immediately after surgery, I contend that since the baseline selection occurs after chemo/radiation (Stupp's protocol), the PFS in the control group will be higher than one might otherwise expect. Likewise, because the selection is conducted post-chemo radiation, and due to the synergistic effect of DCVAX-L in such cases, the PFS in the experimental group will be much longer. Remember, this study originated in 2005. The reason it was paused -- in addition to economic circumstances -- was to develop an appropriate placebo (much more difficult than one would think). Prior to pausing the study, if not placed in the experimental arm, the non treatment group did not want to continue in the study. Consequently, I believe the 33 carryovers were very likely to be in the experimental arm. That said, their data points for this trial give us a time scale that is rarely achievable in such trials. I contend when the trial started up enrollment again, they had to catch-up the control group enrollment. I think you can take it from there what this implies regarding where the more recent events mostly materialized from.



2.As far as visualizations, I'll leave that homely analogy alone. I will run/play with your numbers, however, it will take me a bit more time than you. I think the key point in my visualization left wanting was the fact that the control group still will curve to a finite endpoint length of survival, where only 2-5% will extend survival beyond 5 years.


In the meantime, today I am obligated to challenge several different curves with crystallized H2O, so I may be a bit slow getting back to you.