Another point to consider:
According to the trial protocol, bavituximab is administered weekly until progression or toxicity. Whose measure of progression was used to make the determination to stop administering bavi, the (local) investigators, or the (independent) central reader? Given the divergent assessments of PFS between the investigators and the third party readers, the total amount of bavi given would be significantly different depending on whose read was used. How would this affect the MOS?