That is the one! That is the counter example to EXEL. I forgot their name. They are the example of promising, too good to be true phase III results from post hac analysis, good enough to obtain a special and abbreviated FDA protocol, only to see such data, that looked amazing, to collapse when put into a randomized trial with pre-determined end points.
YMI is my cautionary cancer drug company. They are still at it. Maybe they will get one this time.