Umibe5690, I agree it is always useful to understand the entire data package to most fully comprehend all aspects of any issue, in this case to best understand the cancer's holistic response to the treatment, but in the prioritization of the data here nothing trumps the OS data for therapy performance and approval.
PFS is a logical but an imperfect surrogate, and in this case somewhat compromised, just often the best surrogate available in trial designs. I think the RA reviewers will place a very steep discount on PFS when OS is available and PFS will be viewed in this trial as a minor aspect of the entire data package, more for info consistency only.
All my opinion of course.
GLTA