>But note that the proportion of patients whose death was the trigger for PFS is probably a good indicator of excess death due to the chemo itself.< True. This is the reason the FDA asked GPCB to amend the definition of the primary endpoint in 2005 to include the death component.