The endpoint of OS for 232 vs 99 is still a secondary endpoint of the trial and certainly has meaning. One would expect that if -L works then the 232 live (numerically) longer than the 99 despite the crossover.
Of course the 232 vs 32 means nothing.
OS in early stage cancer is almost always a confounded by subsequent treatment (often including the investigational drug tself as often already approved in later lines). This is just something that trials live with. Nobody keeps pointing out how often all those CI front line trials fail OS with placebo patients using the CI post progression.
There is no reason not to publish the endpoint just because it is confounded.
Further, it would hugely positive. if even close to stat sig. Why would they not be shouting t out?
On the LTFU issue, you still have issues. Despite what you and flipper assert, finding censors does not have to lower the OS. It could go either way depending on how long the lived after the censor. If the censors were unbiased and random they would have no effect. True, there likely is some bias. But you still are talking a fairly minor effect from the few found LTFUs. More on the numbers on a post later.
News 
Market Data 
Discover 
AI
