It would seem common sense to me that the further you go out, the greater the likelihood of cancer recurrence. I.e., much greater odds of cancer recurrence in year 3 than year 1.
Respectfully disagree. The proportional hazards model (equivalent to exponential K-M curves) is not only intuitively appealing, but also the model that’s almost always employed in time-to-event trials.
I would assume you still agree with the uncontrolled data caveat.
Nobody is suggesting otherwise. The ADXS-HPV trial in anal cancer about which we’ve been talking is not a registrational trial, but it’s encouraging enough to generate interest in starting a registrational trial, as noted in #msg-114977668.