Agree there are caveats and they have to prove the AAG biomarker response correlation in a prospective fashion and of course no guarantees of that. But, I am also hopeful that ARRY won't really need the AAG biomarker for 520 for success since high AAG levels only occur in 20-30% of MM patients anyways and response so far seems like it may be good enough when including these patients anyways. But, could be a nice bonus if the biomarker does pan out as I'd like to see combo data for 520+kyprolis in low AAG patients.