Great post Jupiter.

Iconoclast wrote "Statistically it is impossible that the Lipitor patients did not respond to Apabetalone." 

1) I caution everyone to be skeptical of anyone who claims something is statistically impossible.

2) Too many variables to look too deeply into the crystal ball.

3) Very few, if any, folks are predicting that apabetalone has zero benefit with Lipitor. I may or may not speak for the majority, but personally I am expecting apabetalone to have reduced MACE with both statins. However, I am also expecting the %RRR to have been greater with Crestor vs. Lipitor. 

4) Also important to clarify magnitude vs. significance. Apabetalone may reduce MACE with both statins, which would indicate a magnitude benefit in both. As described in #3, if crestor+apabetalone had a greater MACE reducing effect than lipitor+apabetalone, it would be likely that the placebo+crestor vs. apabetalone+crestor comparison would be statistically significant. However, if the %RRR benefit is more modest in the lipitor subgroup, the placebo+lipitor vs. apabetalone+lipitor may not reach statistical significance. A trend for apabetalone benefit with lipitor, but not p<0.05 to indicate significance. This is just one of many scenarios.

Iconoclast has made some bold predictions before that I have disagreed with. Regardless of who is right or wrong, we'll know in less than 12 days.

BDAZ