Biotech Values

[iwfal]

Toxicity could also explain the effect on PFS, if it leads to early treatment discontinuation.

Treatment-related deaths would also be counted as an event for PFS, wouldn't they?

Both good questions - based upon the data I've seen (low rate of withdrawal for tox) I don't believe that is what is happening, but I'll have to do some more data mining to have more confidence in that evaluation.

Additional data to support that this is not a pfs due to withdrawal is that in some of these cases the HR is so much greater than 1 that it is greater than the HR benefit lost by early withdrawal - e.g. in MetMab with Tarceva in nsclc the HR in the Met- group was 1.82 while the pfs benefit of Tarceva alone is only about 0.70 or so. The 1.82 appears substantially worse than you'd get if you just stopped treatment.