I'm not suggesting that the following occurred, only providing it for the purpose of discussion. Lets say up front that progression can be completely distinguished from pseudoprogression, so that all calls of progression are in fact progression. For simplicity, lets look at 5 patients who entered the trial together ten years ago.
Patient one was determined to progress after 6 months, patient two at 9 months, patient 3 at 12 months, patient 4 at 15 months, and patient 5 at 18 months. I think we'd all agree that the median time to PFS was 12 months. I don't believe that anything in this data would be viewed as spectacular.
Now lets say that today all 5 remain alive, that would be spectacular, I believe all would agree. I doubt if anyone would disagree the results would still be considered great if the median survival was 7 years for those who passed on, but one remained alive.
What I'm suggesting is that the determination of PFS is meaningless if years after the determination of progression was made the patient is still alive, and perhaps even healthy. Correct me if I'm wrong, but I believe that once progression is called, it cannot be cancelled. If in fact patients improved dramatically after progression was called, there was no way to change that call.
I don't know, but I suspect that such things occurred routinely during our trial. If I'm correct, it's telling everyone that PFS just isn't properly telling the story of what's happening in the trial. I believe the company became aware that basing the trial on PFS would severely limit the impact of the results. It's very possible that PFS was no better than with SOC treatment, yet people just kept living, perhaps improving dramatically, yet they were judged to have progressed, perhaps just months after entering the trial.
This could not have been known before the trial started. Clearly the choice would have been to allow the trial to fail, as PFS wasn't definitively better, or change the acceptance criteria in line with what was being observed, people were living longer, and better lives.
There is no doubt in my mind that the company realized that stressing overall survival would extend the trial out for many years, it would greatly increase it's cost, but no one could deny the strength of the results. Frankly, even today, knowing all they do, I don't know that there is an acceptance criteria they could use that would permit a substantially shorter trial, and still guarantee approval
I've said it before that a look at the Top 100 should tell us all we need to know. Why? Because 100 patients out of 300+ living a median of perhaps 7 years or longer is spectacular. With SOC treatment 15 of 300 may be alive at 5 years. It really shouldn't take the massive pile of paper represented as the BLA to convince the regulators that this should be approved, but the regulators will insist on it, it's what they do.
Gary
