If I remember from looking earlier in the day, the median number of preceding treatments for the Eribulan P3 was 4. For the CDX-011 BC trial it is a remarkable 7. I'd expect that to have a meaningful effect on background rate of ORR.
A calibration of the difference between the ORR for a drug in refractory and first line:
Tarceva
a) in refractory NSCLC - 13.5% (7/52) b) in first line NSCLC - 22.7% (12/53)