"Wouldn't this set a dangerous precedent by lowering the bar for future approvals?"
I don't think so. This will be a marginal approval, not so much because of a weak data set, as because the control arm drug is weak. Remember though, drugs get approved based on superiority to placebo on a regular basis. Cabo was approved in medullary thyroid cancer based on PFS with no OS advantage. The first interim for that trial showed an OS HR = 1.0 and FDA scheduled an adcomm to discuss that result. FDA had EXEL do another data lock and the HR dropped to somewhere around .90 in favor of Cabo and the adcomm was cxed. At least Tivo had an active control, albeit a marginal one.
As long as there are no negative surprises hiding in the Tivo 3 data set, I think it will get approved and I think it should get approved. Drugs get approved based on noninferiority on a regular basis. More choices is a good thing and for patients who make it all the way to 4th or 5th line treatment, Tivo may be a good choice.