Replies to post #643227 on NorthWest Biotherapeutics Inc (NWBO)

[flipper44]

No, that’s incorrect. If you took time to read and comprehend, you’d realize logic dictates
that non confounded ECA comparisons would need to come first for QALY purpose.

Next, you’d need to know six patients withdrew from trial in the noncrossover group. However, they did stay long enough to get base dosing of Temodar. If those patient deaths are located after the JAMA paper, they will undoubtedly decrease the 35 survival noncrossover subgroup and the 99 control group, because the majority withdrew early, and patients typically withdraw early if they can no longer tolerate chemo, which destroys/damages their bone marrow and thus their platelet and tcell numbers. They were probably very very fragile when they left. Studies say the best way to fix the unfair advantage of arms (like the noncrossover arm) get from ltfu (lost to follow up) is to track their data, particularly date of death. Otherwise, statistics just assume they lived as well in the future as any other people in the trial that got to that point. Which “viscerates” the results of a trial, because ltfu in survival trials are typically very ill when they withdraw. Understand?

So instead of the six patients helping the 99 by combining with the dominant 64 and minority 35, they would actually take the 35 and 99 survival numbers down, down, down. Because that’s reality, not extrapolation. Think Coffer and company can find their data, I do.

Now just put your bias aside and think what else that does. It increases the separation between the 64 and 35. Which when compared to the difference between the treatment arm pre and post recurrence, will probably show a significant delta between the two comparisons. Understand?

Still, you can’t get authentic QALY valuation by comparing the old OS endpoints (you need an ECA arm for that), but you can later show the magnitude of going from No DCVax-l to getting DCVax-l, as I just described.

Finally, you ignore the 2018 curves not only were immature and loaded with censors, but the second half of the trial allowed more immune compromised patients.

Note: Lykiri did do a good job of locating censors in the JAMA 232 arm, and only found 3. One at 6.5 and two between 24 and 48 months.