I noted a higher PFS (1.7) than TTP (1.6) in the ITT drug group. How is that possible?
By definition, it’s mathematically impossible for this to occur in any given dataset; hence, if you read the numbers correctly, there must be more than one definition of “ITT” being used here.
But their preliminary HR on OS is still very good. Maybe the crossover won't make such a difference because these patients just don't have much time after confirmed progression.
My fear is that HR=0.47 but no statistical significance and no verbiage of "strongly trending" means that not many OS events might have occurred. Since the trial is successful, I assumed all patients with or without PD will be able to crossover (I am not actually sure of this anymore), we might get a higher final HR and still no statistical significance.
I do not question the data and importance of this data for the NSCLC (or CRC) trial. I am not just sure what other investors will do if/when they encounter HR>0.47 at ASCO.