In retrospect IMO one of the biggest mistakes RVX management has ever made was, in ASSURE, not making the differentiation between rosuvastatin + 208 and atorvastatin + 208 a prespecified end point. To me heads should have rolled after that one.

I don't think ASSURE wasn't sufficiently powered to do that analysis.  And they would have needed some a priori hypothesis and prior evidence to justify that subgroup analysis.