AI
Monday, March 06, 2017 7:09:04 AM
I believe your perspective is based on past trial problems with pseudoprogression and an emphasis on the potential limitations from criteria that are being used to determine a progression event in this trial even though pseudos are being removed.
You misunderstand me. I agree psPD are being removed. At least the majority of them. The ones remaining are equally randomized between arms so psPD is not the issue.
When I ask where the missing pseudos from what would be expected to be a higher total from all arms including from the pseudo trial, informational arm and compassionate use, you would say they are in the main arm of the trial and being counted as early eventers which would perhaps place the first red dot in your graph where you have it right?
Yes and no. There are some psPD who make it into the main trial. They are "late" psPD's. But because they are equally randomized any effect from them would be neutral.
What I was trying to demonstrate with the graph below is how a 60% interim might show data that is both futile (therefore stopping enrollment) but not harm (trend towards benefit) therefore allowing dosing to continue.
You and others have criticized my view saying it makes no sense to allow a patient to take a drug that's futile. I am saying the trial is futile, not the drug. And the example below shows an interim result (red dot) where it is reasonable to allow randomized patients to continue on treatment but also to stop enrolling more patients into a futile trial.
And I was trying to amplify my case by noting that even if the "red dot" fell below the z=0 line (placebo trending better for PFS) in some circumstances it would make sense to allow patients to continue dosing.
Berry's comment about this:
"Outcomes that would result in stopping a trial of standard cytotoxic experimental therapy would be less negative and might even be expected in an immunotherapy trial."
And that goes to my debate with RKMatters. I assert it "might even be expected" that the "red dot" fall below the Z=0 line below if DCVax "works". So even if the PFS analysis trended negative at the interim, if the OS result trended positive it might make sense to allow patients to continue dosing (but stop enrollment because the trial is futile, not the drug).
