improvement in bone mets by itself will likely not be an approvable endpoint
I haven't listened to the webcast, but I personally don't see why improvement in bone mets (including reduction in fractures and pain as secondary endpoints) doesn't make a perfectly acceptable primary endpoint. It's objective, measurable and clinically significant. Thus it's not a surrogate endpoint like tumor size, which may or may not be associated with improved survival or clinical benefits.
A caveat posted on here earlier is that if EXEL is going more for approvability based on pain benefit, there will be less room for error with side effects from XL184 as there would be if EXEL was aiming for approvability for XL184 based on survival benefit.
That earlier post was not on point. Even as pain benefit for terminal CRPC patients, XL184 safety profile is more than OK. Compare to OGX011 where it is used in combination with docetxel/prednisone, FDA/EMA gave blessing of their trial to determine if provide durable pain palliation for castrate resistant prostate cancer patients, XL184 safety profile is more than OK.
The strong point of XL184 is it does more than one thing, thus composite endpoint might differentiate it further from other agents.
If XL184 is used in earlier stage of PC, safety profile needs to be take into consideration more. EXEL already stated in R&D, they would find new appropriate dose in earlier stage of PC.